Reviews and Daily News with a  Difference!
Recent Updates  | Mac Upgrades/Mods  | CPU Upgrades  | Storage  | Video  | Audio/HT  | Apps/OS/Network  | Search
News, Tips, Reviews or Questions to News at Xlr8YourMac.com
Mac Radeon 9200 PCI user feedbackReturn to News Page


Reader Reports on Mac Radeon 9200 PCI Card
Performance Tests & Notes/Tips on Problems

Reports Last Updated: 7/5/2007

A reader sent a note and some screenshots from the Mac Radeon 9200 PCI card (sold here) docs on high performance PCI and Versavision (display rotation) options in the control panel . Also included are some initial game tests (Call of Duty, Q3, RTCW, Jedi II, Halo) with the card from another reader using it in a Quicksilver Mac.
Note: See below for later reports with other systems like the Yikes G4/PCI, B&W G3 and PowerMac 7500 (w/CPU Upgrades).

" Two new options, which the (previous) ATI 7000 PCI did not have:

Rotate Screen (Versavision):

PCI-Boost/Timing option:

Out of the Manual, i don't have the card.
Thanks, Mitch "

I wonder if the High Performance PCI option also enables PCI Quartz Extreme. (Which some readers have done by hand or using the free utilities since 10.2 was released.)

Here's some initial performance tests from a reader post:

" I've finished up most of my benchmarking. Here's some results. I didn't run identical tests in all of the games because, well, I didn't want to... I'd cry.

I apologize for the poor formatting. The spacing works fine in some forums/browsers but not others. Someday I'll reorganize the data and make some graphs. Actually, I'll probably make this a web page tonight and just link to it rather than worry about formatting in various forums.

One thing to note. The 9200 has a High Performance setting and a Normal Performance setting. When possible, I recommend using High Performance for obvious reasons. However, if there are issues setting the card to Normal Performance may fix the problems.

I still have to finish up tests on my platinum(beige) G3/G4 (only a 533 G4). I unfortunately do not have access to a B&W G3 or a Yikes! G4.

These tests were all done on the same machine. While the Quicksilvers don't have 66 MHz PCI slots the system should be sufficient enough to push the card given the limites of the 33 MHz PCI slots.

Quake 3 Arena

Radeon 9200 Quicksilver Dual 1.2GHz G4 9.2.2

Max Detail
640     800     1024    1152    1280    1600
69.8    64.2    55.8    51.2    51.1    51.1    NoFSAA
72.0    67.7    60.2    55.7    55.6    55.6    NoFSAA Bilinear
72.4    65.6    56.4    51.7    51.6    51.6    NoAudio
-Simple Items on 
-Morks on Walls off
-Ejecting Brass off
-Dynamic Lights off
-Identify Target off
-High Quality Sky off

640     800     1024    1152    1280    1600
77.6    70.4    60.0    54.8    54.6    54.6    NoFSAA


Quake 3 Arena
Normal Performance
Radeon 9200 Quicksilver Dual1.2 G4 10.3.4

Max Detail
640     800     1024    1152    1280    1600
70.0    64.0    55.8    50.9    45.1    37.5    NoFSAA
58.4    49.9    39.5    34.4    23.6    00.0    2xFSAA
45.4    36.9    21.8    00.0    00.0    00.0    4xFSAA
71.9    65.7    57.0    52.1    46.1    38.2    NoAudio
-Simple Items on 
-Morks on Walls off
-Ejecting Brass off
-Dynamic Lights off
-Identify Target off
-High Quality Sky off

640     800     1024    1152    1280    1600
82.2    74.1    63.0    57.0    49.9    40.5    NoFSAA


Quake 3 Arena
High Performance
Radeon 9200 Quicksilver Dual1.2 G4 10.3.4

Max Detail
640     800     1024    1152    1280    1600
93.0    81.9    68.2    61.1    52.7    42.3    NoFSAA
73.1    59.7    45.0    38.4    25.3    00.0    2xFSAA
53.2    41.5    23.2    00.0    00.0    00.0    4xFSAA
95.8    84.2    69.7    62.3    53.7    43.2    NoAudio

-Simple Items on 
-Morks on Walls off
-Ejecting Brass off
-Dynamic Lights off
-Identify Target off
-High Quality Sky off

640     800     1024    1152    1280    1600
105.9    91.8    74.8    66.0    56.5    44.4  NoFSAA

Mac OS 9.2.2
Jedi Knight II

Max Detail
640     800     1024    1152    1280    1600
43.3    41.8    39.2    37.4    34.9    30.9
34.1    29.4    19.9    00.0    00.0    00.0    4xFSAA
39.2    35.7    30.8    28.0    21.3    00.0    2xFSAA

Geometry Detail Low
640     800     1024    1152    1280    1600
64.7    63.9    61.4    58.5    53.5    44.8


Mac OS 10.3.4
Normal Performance
Jedi Knight II

Max Detail
640     800     1024    1152    1280    1600
37.3    35.7    33.1    31.4    29.2    25.4
30.8    26.7    18.4    00.0    00.0    00.0    4xFSAA
34.3    31.4    27.1    24.8    19.2    00.0    2xFSAA

Geometry Detail Low
640     800     1024    1152    1280    1600
59.6    56.0    49.8    46.1    41.3    34.7


Mac OS 10.3.4
High Performance
Jedi Knight II

Max Detail
640     800     1024    1152    1280    1600
51.5    48.6    43.7    40.7    36.9    30.9
30.8    26.7    18.4    NA  NA  NA  4xFSAA
34.3    31.4    27.1    24.8    19.2    NA  2xFSAA

Geometry Detail Low
640     800     1024    1152    1280    1600
71.2    69.9    62.8    57.1    49.7    39.5

Normal Performance
Call of Duty

Max Detail
640     800     1024    1152    1280    1600
16.4    15.5    14.7    14.3    13.5    12.0
19.1    18.0    16.9    16.3    15.3    13.3    Peformance min - model
detail normal
23.7    21.9    20.4    19.6    18.3    15.7    Peformance min - model
detail minimum
24.5    23.9    22.8    22.0    20.8    18.8    Everything off or low


High Performance
Call of Duty

Max Detail
640     800     1024    1152    1280    1600
24.2    22.5    20.5    19.6    17.8    15.2
25.4    23.6    21.4    20.3    18.6    15.7    Peformance min - model
detail normal
34.5    31.3    27.9    26.2    23.5    19.6    Peformance min - model
detail minimum
36.5    35.0    32.2    30.4    28.0    24.4    Everything off or low

Return to Castle Wolfenstein

Radeon 9200 Quicksilver Dual1.2 G4 9.2.2

Max Detail
640     800     1024    1152    1280    1600
29.8    28.5    26.6    25.2    23.0    20.0    NoFSAA


Return to Castle Wolfenstein
Normal Performance
Radeon 9200 Quicksilver Dual1.2 G4 10.3.4

Max Detail
640     800     1024    1152    1280    1600
31.4    29.6    26.7    24.7    22.7    19.8    NoFSAA


Return to Castle Wolfenstein
High Performance
Radeon 9200 Quicksilver Dual1.2 G4 10.3.4

Max Detail
640     800     1024    1152    1280    1600
40.3    38.0    33.5    30.5    27.5    23.3    NoFSAA

Radeon 9200 PCI
Dual 1200MHz G4 1152MB
Mac OS X 10.3.4

Max (highest quality everything but lens flare and FSAA)

###Video Options###
Resolution= 640 x 480
Refresh rate= 0 Hz
Framerate throttle= No Vsync
Specular= No
Shadows= No
Decals= Yes
Particles= High
Texture Quality= High

Min (everything off or low. Particles low because of plasma grenades)

###Video Options###
Resolution= 640 x 480
Refresh rate= 0 Hz
Framerate throttle= No Vsync
Specular= No
Shadows= No
Decals= No
Particles= Low
Texture Quality= Low

Vertex Shaders Only Normal Performance PCI
Average frame rate=12.15fps

Average frame rate=14.15fps

Vertex Shaders Only High Performance PCI
Average frame rate=21.72fps

Average frame rate=26.66fps

No Shaders Normal Performance PCI
Average frame rate=21.23fps

Average frame rate=24.76fps

No Shaders High Performance PCI
Average frame rate=29.29fps

Average frame rate=34.86fps

(He later wrote with links to pages he'll update with more test results

  • B&W G3 test results page
  • Quicksilver G4 test results page "
  • With the original Radeon PCI card in the past, the B&W G3 66MHz PCI slot (vs the 33MHz slot) added appx 10% in performance in some tests in the same system. (granted the B&W G3 has less than expected 33MHz PCI slot performance though - and the older cards didn't have the high performance settings option.)
    If any reader with a B&W G3 or Yikes G4/PCI (w/66MHz PCI slot) or a G5 (w/PCI X slots - which will run the card at 66MHz) test the 9200 in the future, let me know. Thanks

    Updates/Later 9200 PCI Owner Reports: (most recent first)
    (Note: If you have an older mac w/G4 CPU upgrade (beige G3 or older for instance), for years there's been reports of conflicts (boot hangs, etc.) with ATI OS 9.x extensions (graphics accelerator extension IIRC) and cache enablers. Renaming extensions (space or tilde first letter) to change load order may help. For OS 9.x users, install the 9200 OS 9 drivers before installing the card. Also there's been several problem reports from older macs w/PCI IDE cards or SCSI PCI cards and the 9200 PCI card, although one Beige G3/9200/Sonnet Trio card owner said he's not seen any problems._
    Please include notes on OS version/system used, any upgrades/other PCI cards installed and if you applied the latest radeon firmware update when you send a report. Thanks.

    ATI 9200 PCI Card not loading OS X Driver (ATIRadeon8500.kext): (not common)

    (from 7/5/2007 news page)
    "Hi Mike, I'm dealing with ATI via their web-ticketing process, but am curious if you could help me also.

    I have a 9200 PCI Mac Edition that's not loading any kext files on a G3 (B&W) or G4 (Digital Audio). And yet my G5 (DP 2.5GHz) loads kexts for the card just fine.
    The result is that there is no 2D graphics acceleration on the G3 or G4. (needless to say: very painful) I have tested with the ATI Displays installed to no avail (about to test again)

    I have tested using two B&W G3s and two Digital Audio G4's so I'm steering towards a software issue.

    Could you please post this this on your site and ask other owners of 9200 cards with G3/G4 macs, if any kexts load for them.
    This can be seen via ASP and looking at the Extensions section.
    On my G5 here, "/System/Library/Extensions/ATIRadeon8500.kext" loads for the 9200 PCI card.
    (Just for the record, I asked what OS X version was installed)
    All macs I've tested with were running 10.4.9.
    I've since tried 10.4.9 Server and 10.4.10 on the G5 as well.
    Anyway, if you could help me out, I would be very appreciative.
    Thanks again, Raoul C. "

    About 1/2 dozen 9200 PCI owners wrote they didn't have this problem with their G3/G4 Macs - the ATIRadeon8500.kext is shown as loaded. I wrote Raoul again to ask if he'd rebooted a few times and maybe tried clearing caches/kext cache. If all else fails I'd be tempted to try a reinstall of the last "Combo" OS X update (10.4.9 PPC Combo Update since he has 10.4.9 on the G3/G4 Macs) or a reinstall of ATI drivers from OS X using Pacifist. (Assuming his G3 and G4 OS X installs are not both corrupted - but can't hurt to run a disk utility on them, maybe even a RP and sometimes reinstalling a Combo updater can fix odd problems.)
    I also sent him a FYI on the Hack to Enable Quartz Extreme w/PCI Radeons, but regardless the ATI driver should be loading for it. I don't recall anyone else ever having a problem with no OS X driver loading.

    A reply to a previous post below (complaint on DVD playback in 10.4.9 from an upgraded B&W owner)

    (added 5/1/2007)
    "I have a B&W G3 with a ATI 9200 PCI and I haven't seen any issues with the DVD playback.

    The specifics for the system are as follows:
    B&W G3 - 900MHz Processor Upgraded (Powerlogix)
    1 Gig Ram, 60 Gig HD
    OSX 10.4.9 - Build 8P135
    ATI Displays 4.5.7 update
    -Harry B."

    10.4.9/LCD Projector user complaint on DVD Playback Performance:

    (added 4/30/2007)
    "I would like to describe my experience with an ATI 9200 PCI Mac Edition.
    Here's my rig:
    Mac OS X 10.4.9
    B&W G3, upgraded with Sonnet 500MHz G4 ZIF
    1GB RAM, 160GB HD (only 128GB formatted) (no onboard IDE 48bit addressing)
    80GB HD (both HDs on the internal ATA bus)
    SCSI PCI card (not used, but still installed)
    (I'd remove the card if not used.-Mike)
    Opteron (sp?) LCD projector
    (only the projector used?)

    This machine is exclusively used to play DVDs and MPEG2 videos. When using the ATI Rage 128 that came with the machine, DVDs play reasonably well. They're smooth, but the de-interlacing is not all that great.

    When I plugged in the ATI Radeon 9200 Mac Edition, I noticed a considerable decrease in performance. Both in the OS and DVD playback. Video response was sluggish across the board. Windows would open very choppy, and DVD playback was unwatchable. I installed the software that came with the card as well as an update from ATI's (now AMD) web site, but it did not improve matters. I turned on the Performance option on the ATI utility, but it did not help. Turning on Deinterlacing made it even worse.

    Finally, I decided to install the hack that enables Quartz Extreme on the 9200. (Reader FYI - see the Mods to Enable Quartz Extreme on PCI Radeons-Mike) That made it better... DVDs would play well enough and the windows were definitely opening faster, but it still isn't as good as my original Rage 128 card. Funny that the DVD menus are still choppy, even though the actual movie isn't. Some MPEG2 videos are a little choppy, but still more-or-less watchable.

    The support I received via a submitted trouble ticket was insufficient. On two separate replies, the "solution" began with "Click on the Start menu". They continue to assert that the problem is in the Apple DVD player, not in their hardware (of course).

    I read one response on your 9200 page from somebody who had almost exactly my config. That person said that DVD playback was just fine, but I didn't see anything significantly different between what he did and what I did.

    (I assume the card is in the 66MHz PCI slot? What resolution are you running? It's not a great solution but sometimes setting color depth to 16bit (thousands) vs millions helps performance on older Machines. (I used that in the past on some older Macs and the movies still looked ok to me personally.) I also asked if Spotlight was indexing a drive or if he checked Activity monitor for anything taking up resources/cpu time)
    I disabled dashboard, and let the system run for a couple days so Spotlight should be all indexed up. (Spotlight will also immediately start to index any drive you connect)

    When watching a DVD with Apple DVD Player, Activity monitor CPU load hovers around 30-50%. When watching a DVD with VLC, it's at a solid 100%. And VLC is much more choppy than Apple DVD Player. (I suspect apple's DVD player is more efficient than VLC, assuming the CPU readings were done playing the same MPEG2/DVD) It's funny though, the movie plays well enough with Apple DVD Player, but the menus are unwatchable. With VLC, they're about the same level of choppyness. Also, it seems that VLC is having a hard time de- interlacing the video while Apple DVD Player does a pretty good job.

    I'm beginning to think that hardware decoding is actually working with Apple DVD Player, but not with VLC. (Apple's DVD player is software encoding since version 2 for OS 9.x. V1.1 of the DVD player did use the MPEG decoder (piggyback card for the OEM Rage128) and other MPEG hardware options for other macs like the Lombard, but since v2 (and all OS X players) the Apple DVD player does not use the hardware decoders.-Mike)
    All in all, the card APPEARS to be working, but I'm not noticing an improvement over my Rage 128. I assumed it would be night-and-day. Particularly since my Rage 128 is missing its DVD decoder daughter-card.

    Thanks, Shawn"

    A reply to this post (above) from another 9200/10.4.9 user said he didn't have any DVD Playback issues.

    Beige G3/OS 9.2 report (OK)

    (added 10/13/2005)
    "Mike, It's been a while, but I had a fully successful configuration with a beige G3 running 9.2.2 with a Radeon 9200, Sonnet 1GHz G4 ZIF, and ATI TempoTrio all at the same time. I had some trouble getting the cards in the right slots to prevent Open Firmware problems at boot time, but once I worked it out, the system was pretty much rock solid.

    It's been a while since I had the machine in this config (all of the equipment moved into a Blue & White), but if I recall correctly, I used the latest drivers from ATI's site and I updated the video card with the latest firmware as well.

    For what it's worth, relying on ATI's tech support is about as useful as asking your dog to fix your computer. When I was having OF problems at boot time, ATI responded with a "we'll get back to you." Sonnet replied promptly and accurately. After a short email conversation with the Sonnet tech, I was up and running without trouble.

    One more thing: That tech note you linked from ATI (see below) is related to Open Firmware problems. It doesn't sound related to what the original writer was emailing about. (previous report below) The OF problems will cause the Mac to hang during the boot while loading extensions. This is the problem I was having - it's related to how much free memory is available in OF for cards to use for their specific variables and firmware. If you've got several cards (especially complicated cards like dual function cards or PCI bridge chips), there can be OF conflicts that cause the Mac to freeze at the Mac OS splash screen. (this is an issue that was also seen many years ago with legacy PCI Macs that had several PCI/controller cards installed for instance and there was not enough space in the NVRAM/PRAM IIRC.-Mike) The Radeon 9200 is a common culprit, so moving cards around or removing other PCI cards often solves this problem.
    Author of ClassicStumbler
    http://www.alksoft.com/ "

    G4 Upgraded Beige G3 rev A/OS 9.2 problem report:

    (added 10/12/2005)
    " Last spring I bought an ATI Radeon 9200 graphics card for use in my brother's PowerMac Beige G3 running OS 9.2.2 and a Fast Mac 400MHz CPU. (G4 ZIF I assume-Mike) The 9200 couldn't be made to work. The symptom was that with a monitor attached to the VGA port on the 9200, the system would boot and the desktop would appear on the screen, but as soon as my brother moved the mouse, the computer would freeze. I e-mailed ATI Tech Support about this, and here is their reply.

      " We have actually tested the Radeon 9200 on OS 9.2.2 and on a Beige G3, and have found that the system does hang. We have queued this issue to our Engineers for further investigation."

    Users of the Beige G3 might want to use a different card.
    (For years there's been notes on problems here (hangs at boot usually) from G4 upgraded PCI mac owners with ATI cards running OS 9.2 due to a conflict with the ATI graphics accelerator extension and cache enablers for G4 CPU upgrades (renaming the extension to change load order doesn't always help). I asked if he had checked that and if the same problem happened with extensions diabled, if he had tried using OS X and what rev Beige G3 he had.-Mike)

    We did try renaming the cache control extension so that it would load after the ATI extensions, but this didn't change anything.
    The system boots up OK with the monitor on the Radeon 9200, but as soon as you move the mouse it hangs.
    If you have the Radeon 9200 installed, but the monitor connected to the motherboard graphics port and nothing on the 9200, then all is well. I have read many reports of users fiddling with the extension load order and with which slots the various cards are installed in, but none of this has worked. I am waiting to see what ATI has to say, if anything...

    This is a G3 233MHz Rev A. The computer has been tried with no other cards in it with the same result. If a monitor adapter is connected to the motherboard graphics port, but no monitor is attached to it, and a monitor is attached to the ATI 9200, then the screen stays dark - the computer wants to put the image on the motherboard graphics port by default, even if you use the Monitors control panel to put the desktop and Menu on the Radeon 9200 card. Very frustrating. (as mentioned years ago here in the FAQ - do not leave any Mac/VGA adapters connected to the onboard video port w/o a monitor attached, as the Mac will think a monitor is attached there and result in a blank screen on the PCI card.-Mike)
    If you boot with extensions off, the motherboard graphics port works, but the ATI port does not. The 9200 can't seem to start up in a default mode.
    My brother has a genuine OS X disk, but he has not installed it. (OS X v10.2.6 plus the 10.2.8 updater.)
    -Victor "

    There's other reports here (below and above) from Beige G3 owners running 9.2.2 with a 9200 card OK and another Beige G3 owner report said you should install the later ATI 9200 OS 9.x drivers (includes a later graphics accelerator extension) before installing the card.
    BTW - ATi's 9200 Mac Edition Specs page has a note at the bottom that links to one of their kbase articles (unfortunately requires cookies enabled in the browser for the link to work) that the card has no OS 9 support when used in some G4 models (and the linked article also mentioned the Beige G3, B&W G3, etc.). Here's a copy of that kbase article:

    " ATI Customer Care > Solve a Problem > Lockups and Hangs > Operating System >
    737-468: Problems with Mac OS 9 and RADEON 9200 MAC EDITION on G3 and some G4 systems

    The information in this article applies to the following configurations:

    • "Gigabit Ethernet" Power Mac G4 model
    • "Digital Audio" Power Mac G4 model
    • "Sawtooth" Power Mac G4 model
    • Blue and White G3
    • Beige G3
    • Mac OS 9.2.2

    The RADEON 9200 MAC EDITION may cause the system to hang when booting into Mac OS 9. Moving the board to a different PCI slot or rearranging the existing PCI cards in your system may remedy the problem.
    This is not an issue on the MACOS X platform. "

    This reportedly is due to space problems in OF/NVRAM (see later report above).
    Since it's been awhile since this topic came up, if any other Beige G3 owner is running a 9200 (with OS 9.x or OS X) let me know. (Include system details in reports - thanks)

    Beige G3/OS 9.x Success Story: Although several Beige G3 owners reported problems with the 9200 PCI card in OS 9.x, this reader didn't:

    (added 1/7/2005)
    "Hey Mike,
    Just letting you know I got a Radeon 9200 PCI from OWC for my tricked out Beige G3.

    Not only did it work flawlessly in Mac OS X 10.3.7 and OS 9.2.2 (no freezing or artifact issues as of this writing), but I was able to copy over a few of the OS 9 drivers from the CD to a Mac OS 8.6 installation (which resides on the same partition as OS 9). I was extraordinarily happy to find that graphic acceleration was enabled! I could tell before/after because moving windows had painfully slow refresh times before I got the acceleration to work.

    If you copy all the OS 9 drivers over to 8.6, you'll get a system crash on startup, but just moving a few of the key components worked like a charm for me. Believe it or not, I still have to use OS 8.6 for some old applications that doesn't work so well in 9.2.2, so having fast refresh rates on the windows is a big help.

    And of course, it's all about testing your limits on the classic machines. I know most people don't use 8.6 anymore, but we do it and try it because we can. :) Thanks for the backwards compatibility ATI!
    (I asked Joey for more details on his system (i.e. any other PCI cards installed and just curious if it was a rev C rom dimm system.) Here's an answer added in April 2007.-Mike)
    I run a Beige G3 with a 500MHz Sonnet G4 and 768MB installed. I do have a Rev C ROM, and the order of my PCI cards are: OrangeLink+ USB2/Firewire, ATI Radeon 9200, ATTO Express PSC SCSI (which powers my OS X installation), and the A/V Personality card.
    I never use 8.6 and 9.2.2 anymore. I've had this machine going on 6 years and it's getting really pokey on 10.3.9. I'll probably invest in a Mac Pro after the next revision the end of this year.
    Joey K."

    Radeon 9200 PCI in B&W G3 w/G4 Upgrade:

    (added 1/7/2005)
    "Hi Mike, Love your site - I've been coming here since it's inception. Here's some info about using the VersaVision (screen rotation option) in the PCI Radeon 9200 Mac Edition. The Radeon 9200 readme file says that frequently (!) when you rotate the screen orientation, that the screen resolution will change (instead of 1024x768 to 768x1024 it can go to 1024x1280). Well I can confirm that this is definitely so: I was running at 1024x768 on my 15" LCD monitor and selected the 90° clockwise option in the ATI Displays app and the screen went dark, giving me an "out of range" error. And naturally there is no confirmation dialog box so I'm stuck with a dark screen. This is just crappy programming, especially when they say in the readme: just select the correct resolution if you don't get the same one after rotation. Well, I can't if my screen is dark, you idiots!

    I can start up in safe mode and get my screen back but there is no going back to a normal screen when not in safe mode unless I remove the ATI kexts from the System folder. Yes, you read right, it has nothing to do with any preferences in the User folder because the screen goes dark *before* login- it's set that way before any user prefs are read. And there are no hidden or other prefs file to delete to recover a normal screen function. My install disk is in another location right now so I need to get it back before reinstalling fresh drivers and getting back my 2D and 3D acceleration.

    Allow me to point out again that it is sheer idiocy to *knowingly* ship (documented in the readme) a product that fails to perform a feature correctly (screen rotation) and not leave a simple way to recover from a failure (industry standard 15 sec "OK" dialog box after screen rotation).
    - Tom
    (I asked Tom if his card firmware and ATI displays/software was up to date)
    I did not apply the latest firmware patch (downloaded it after the black screen problem) but I did have the latest Radeon v.4.4.1 drivers (latest public release of ATI displays control panel for OS X is 4.4.2) and I was using OS 10.3.7. I didn't check the firmware version so I don't know which it has. The computer is at home so I'll find out tonight before running the latest FW patch and reinstalling the drivers. "

    I asked he let me know if the firmware update/displays update helped.

    Another Beige G3/OS 9.x problem report: (trying to catch up on mails in the last few weeks - still some 9200 PCI mails not posted here yet.)

    (added 12/22/2004 from Dec. 11th mail)
    I am Scott C. I wrote my original e-mail from my work e-mail address. It's Saturday morning and I see the edit on the web page asking how I installed the card. (a reader sent a note that installing the card before the 9200 drivers in OS 9 can cause freezes due to older ATI extensions)

    I have been using the MAC platform now for 20 years as a graphic artist. I am very familiar with the platform.

    In trying to make the 9200 work, I tried numerous iterations of install methods. The 333MHz Beige G3 is a "new" machine to me, with nothing on the disk I had to save. Getting this card to function was a challenge, so I would wipe the hard drive clean and install a fresh OS for each iteration of what I did to ensure I had not forgotten anything while switching extensions on and off. (OS 9.2.1 install disk, then upgrade via 9.2.2 updater) NOTHING worked.

    The card was installed, and then the drivers. Then the drivers were installed, and then the card. I tried switching PCI slots, removing the on-board RAM module, (Each of these machines has 2+4 MB video ram) (Note - if not driving a monitor also make sure there's no mac/vga adapters on the onboard video port - otherwise the system will think there's a monitor attached and cause problems) installing in the 300MHz machine with no other PCI cards present. All to no avail.

    The absolute hardest part of this whole nightmare has been ATIs complete and total lack of support. I would call ATI, be put on hold for extended amounts of time while they figured out that yes, there is such a product as a 9200 PCI for MAC.

    Twice I was cut off by tech support reps, and later given the excuse that Bell Canada had been performing digging operations outside of ATIs offices, and that is the reason for being cut off. (Why they couldn't e-mail me after being cut off, having all my info on their screen during the call, is a mystery to me.)

    A support rep finally called me back one day at work, and told me that OS9 was dead, they couldn't figure out why the card wasn't working, (after being sent ASP reports from both machines) and that maybe I should just upgrade to OSX, that that seemed to work without a hitch with this card. I laid into him, telling him that Panther wasn't supported on a Beige G3, that I'm not the type of user that tries to make a cat bark, and that "Gee, while others seem to have had success running OSX with XPF on a beige G3, why couldn't I have success in using this card for what ATI claims is a fully supported application of their product?"

    Unfortunately, that call was a death nell. This whole situation has put a very bad taste in my mouth in regards to ATI, the marketing of MAC products in general and, unfortunately, the MAC platform itself. Having used the MAC for so long, I do not consider OSX a worthy replacement. It is slow, the user interface is cumbersome. It lacks that ease of use and easy integration that made a MAC a MAC. OSX is just another flavor of linux. IMHO, X marks the spot where the MAC platform died.

    The 9200 will be put up for sale on eBay this weekend. Good riddence to bad rubbish! Shame on ATI for hoisting this mound to market.

    Thanks for a great web-site. I so appreciate the no BS, no marketing hype attitude of the information it contains. Keep up the outstanding work! Best wishes for a great holiday season to all the xlr8yourmac team.
    Scott C. "

    Comments on Older ATI Extensions Problems w/9200:

    (added 12/8/2004 from 12/7 email)
    " Hi Mike,
    I was reading Scott C's report (below) regarding OS 9 freezing with the Radeon 9200. Could you ask him if he installed the 9200 drivers before installing the card as the instructions mentioned (or should have mentioned)?
    In my experience, older versions of the 2D ATI Graphics Accelerator driver would freeze OS 9 with a 9200. If the card is already installed, then one should be able to disable the ATI Graphics Accelerator extension (I don't recall the exact name) (that's it. The FAQ's Video cards section has an item from some years back on the names and functions of all the ATI OS 9.x extensions.-Mike) with the Extension Manager by either holding down the spacebar during startup to bring up the EM or by holding down shift to start up with Extensions disabled. Then, once the Finder is loaded, disable the extension via Extension Manager. That should allow one to get around the Finder freezing problem. Then they should install the OS 9 9200 drivers.
    A.J. "

    Conflict w/Kingston NIC Card extension:

    (added 12/7/2004)
    "I had the ATI extensions that came with my '9200 mac edition 128MB' conflict with Kingston's network card extensions, and hang the system under os 9.2 on my powermac 9600.

    I had to re order the extensions in order to fix it, or disable the kingston extension.
    Ben "

    For older macs running OS 9.x and a CPU upgrade (i.e. beige g3s w/G4 CPU upgrade), there's been report (for years) of freezes at boot due to a conflict with some cache enabler extensions (and the ATI graphics accel. extension as I recall - renaming to chg load order may help).

    9500 Mac/OS 9.x/Sonnet Trio OK:
    Several Beige G3/Sonnet ATA/133 card owners reported problems with the 9200 card installed, but not this Trio card owner:

    (added 12/6/2004)
    "I have a 9500 with the Sonnet Trio, SIIG 2-port USB 1.1, AsanteFast690, Miglia DVR and 9200 mac edition 128MB.

    I have not experienced any bus mastering issues with my Sonnet G4 700 Crescendo installed while using OS9.1 with 1GB of 128MB EDO interleaved memory
    Bus mastering is a hardware only issue? or is it related to the OS as well?
    Kevin M. "

    Most other problem reports were from OS 9.2.2, but not sure that is a factor (some claim too many busmastering cards in older macs, some thing it's due to the cards and nvram usage.) The 9500/9600 macs have 2 PCI busses (upper 3 PCI slots/lower 3 PCI slots) which can help.

    Another Beige G3/OS 9.x problem report:

    (added 12/6/2004)
    "Nothing but problems with the 9200 on two different Beige G3s running OS 9.2.2.
    I purchased this card back in September when it was released. In two pretty much stock systems, the mouse freezes when the desktop comes up. Countless time and energy has been expended trying to get this card to function, with no help or direction from ATI as to the problem and/or a solution. The card was sent back and replaced, the exact same things happen with the replacement.

    My systems:
    Beige G3/333MHz MT - OS 9.2.2
    ATTO SCSI card
    768MB OWC RAM
    ROM rev "C"
    Only PCIs installed: 9200 and ATTO

    Beige G3/300MHz MT - OS 9.2.2
    Stock IDE Drive/port
    768 MB RAM
    ROM rev "C"
    All other PCI slots empty

    A frozen desktop is what appears after boot up on BOTH systems with this card installed. ATI has shown that is it unprepared/unwilling to provide technical service for this card. Their answer in a nutshell: "Sorry, we don't know what’s wrong, OS 9 is a dead OS and we won't be releasing any updates for this item. You should try another solution."

    Never before have I received such lame service/support from a component manufacturer. MACs are not a priority at ATI, that's is so very obvious from the experiences I have had with them concerning this product.
    Scott C. "

    A later reader comment above has some notes on problems with older versions of ATI's graphics accelerator extension (I'll try to find scott's original mail and ask if he had updated the ati drivers for os 9 before installing the card.)

    MDD Owner Problem Report: (Not sure this is common, but could be related to other PCI cards installed.)

    (added 11/26/2004)
    " My setup:
    Apple PowerMac (MDD) G4 Dual 1.25 GHz (with power supply replaced via: Power Mac G4 Power Supply Exchange Program)
    1.5GB DDR SDRAM (3 x 512 MB PC2600U-25330)
    ATI Radeon 8500 64 MB AGP Mac Edition
    ATTO ExpressPCI UL3D SCSI Card
    4 U160 10K SCSI Drives
    Pioneer DVD-RW DVR-104 ATA Drive
    QPS CRD-BP1600P ATA Drive (CD-RW)
    Mac OS X 10.3.6 w/all updates
    All device firmwares updated and latest drivers installed

    The Problem:
    Introduction of ATI Radeon 9200 Mac Edition caused my stable system to become unstable.

    With the 9200 installed, after a few minutes, my screen would freeze completely, and the only way to resolve the situation was to hold down the power button on the front of the tower to shut the machine down.

    To attempt to solve the problem, I tried zapping the PRAM, updating the 9200's firmware, updating ATI drivers, repairing permissions, switching monitors, switching PCI slots, hardware diagnostics, hard drive media scans, and probably some other things also.

    The freeze occurs 100% of the time with the 9200 installed. Without it installed, my system is completely stable.
    I am petitioning OWC for a refund. The exact OWC Description is:
    Item Number: ATI100436011OB
    Description: (*) ATI Technologies RADEON 9200 128MB MacEdition PCI Graphics Card. New, Open Box, 3yr ATI Warranty.
    You have a fantastic site.
    Thank you,
    Kenneth "TigerKR" Roberts "

    If any other MDD owner is running a 9200 PCI Radeon let me know if you see any problems. (also include notes on any other PCI cards installed - some Beige G3 owners for instance have reported problems after installing a 9200 PCI card in a system w/sonnet ATA/133 card installed.)

    Beige G3/OS 9.2.2 problem report:

    (added 11/26/2004)
    " Have G3 Minitower OS 9.2.2 384 MB Memory Radeon (PCI) 9200 Mac Edition 128MB
    Pointer (Cursor?) has 8 rows of eight lines slightly staggered to the lower and right of the pointer it has this corruption with DVI, ADC (NOTE: the card does not have an ADC port.-Mike) & VGA. Installed (124) Mod (firmware update I assume) no difference, but have noticed when the pointer turns to a hand it dissapears but returns when the pointer returns. Unable to take a screen shot it does not appear. So I have drawn it. Attached.
    Regards, Robin P.
    (he later wrote with more info)
    I have looked at a few things eg; its there at start up with extensions off, all nvidia drivers removed. The hand changing on the browser loses it but returns when pointer returns but for example when the hand comes up over the control strip with option key held down it does not lose the corruption also the lines described are shimmering on the left hand sides. I noticed there was a similar problem with some older 64mb cards as well.

    The rom mod version is 113-A27502-124 original rom was 113-A27502-120. I am unaware of any newer versions.

    (and his most recent reply:) Hello again just reading updates and noticed a few people had problems with freezing after installing pci radeon 9200 mac edition 128MB. Yes I am the same guy with the mouse corruption and yes My G3 MT did also freeze after the first installation, but I have found that by moving the card around sometimes it can be overcome.
    What I did was try every slot and finally it worked with the usb card in bottom slot, middle slot empty radeon in top slot just below the oem personality card for video in / out. I also switched the ide channels over at the mother board (? what do you mean by that) it worked but only to find the pointer corruption legacy.
    Regards, Robin P. "

    There were some reports of shuffling PCI card order in Beiges that helped with some other card mixes in the past (like the wired4DVD card as I recall), but it seems like nothing is a 100% cure in some configurations.

    If any other Beige G3 owners running 9.2.2 see this problem or find a fix for it in their system config, let me know.

    More Beige G3/9200 PCI Problem reports

    (from Oct. 18th email)
    "I am having a problem with the ATI 9200 ME PCI video card, and would like to be able to post a message to Bruno about it....

    The main issue I have is in stand-alone Sys 9.2.2 running on a Beige G3 MT with G4/500 upgrade. The mouse freezes (just after the desktop appears) if the ATI extension ATI Graphics Accelerator is enabled. Everything works OK in Classic mode or OS 10.3.5 just as it should.
    (I asked if he tried disabling any Cache Enabler software for the G4 Upgrade, as this has been a known issue in the past from some beige G3/ATI card/G4 upgrade users. Sometimes renaming the extensions to change the load order might help-Mike)

    The other issue in Sys 9 is about what the ATI Config Menu should do. ATI Help only beeps.
    Chris B. "

    I asked if he would try disabling any cache enabler extension in OS 9.x as a test. I also forwarded his mail to Bruno at ATI.

    Beige G3/Sonnet ATA/133 PCI/9200 PCI Problems:

    (from an Oct. 16th email sent to bruno of ATI and CC'd to me)
    "I've been through quite a bit with the 9200 PCI and my Beige mac and I just wanted to share, hopefully for the collective good. Just for the heck of it, I called ATI support to get their opinion. They told me to "give up" and find another kind of card. I'm not upset about that, but I was a little disappointed in their lack of interest. Bruno, I know you suggested I join your beta forum and I will, I've just been really busy (buying a house and moving) and I've had to try and solve the card issues in time to meet the deadline for returning it if I "give up" on it.

    So, I sent the first 9200 back to OWC and got another one to try. OWC confirmed that the first card I had was defective (as I had suspected). This second card did work in my system, but only if my Sonnet Tempo 133 ATA card was not installed; otherwise the machine would hang on startup. I didn't understand this and asked OWC for help. They spent a couple days researching and then said they thought it was a PCI bus mastering issue. My understanding, which is not the greatest, is that an older Mac like mine can't handle too many cards that are PCI bus mastering cards. The 9200 is and so is the Sonnet Tempo and that at computer boot, they fight and cause the computer to lock at startup. I called ATI to confirm this. Their response was that it was true and to "give up." They said the issue could not be resolved and would not be in the future. I asked about why if the original pci radeon worked how was it the 9200 doesn't. They basically parroted back that the 9200 was a pci bus mastering card and that the original radeon pci is/was not. That was what I suspected, but I wanted to hear the ATI guy say that. I'm not sure I'm totally convinced that completely explains the situation, but I've got nothing else to go on.

    I have to either get the card in the mail by Monday to return it to OWC (and pay their restocking fee) or sell it to somebody else now as I can't go without my Sonnet Tempo card. I'm going to have to score myself an original macintosh radeon pci since I gave mine away.

    I'm sending this email in hopes that if this is new info, you can use it to help others avoid spending the ridiculous amount of time that I have trying to figure this out. If there is anything you can add to help me better understand the whole PCI bus mastering business or angles that I've not considered, I'm all ears.
    Thanks, Greg T. "

    Some have suspected this problem is related to NVRAM/PRAM but not verified. If anyone is running a Beige G3 w/PCI IDE card and a 9200 PCI OK - let me know the details. Thanks

    (from Oct. 15th email)
    " Mike, I have an odd issue concerning the ATI drivers that shipped with the Radeon ME 9200 PCI card and the later releases of the drivers and flashers.

    Equipment: PowerMac Beige G3 mini-tower with Sonnet G3/500 ZIF processor; 60+80GB hard drives; 640MB RAM; USB; CD-RW. Video card was an original Radeon Mac edition 32MB (not the later Radeon ME 7000). OS is 10.2.8. The upgrade to 10.2.8 was done shortly before adding the CD-RW and the Radeon 9200 (added CD-RW first and tested before adding vidcard). ATI drivers powering the orig Radeon were the ones installed with 10.2 retail CD. Monitor is a NEC 19" 97FS CRT. Quartz Extreme is not enabled.

    Issue: X-Bench graphics scores dropped from 44 to 34 after installing the 9200 drivers from the CD WITH the orig Radeon 32MB in place. Also noted choppiness in the Dock and minor "type-behind" stalls in some apps requiring keyboard input.

    Installed the new 9200, ran the 121 ROM updater, then the October 2004 Hotfix (ROM rev now at 124). Still zero change in the X-Bench scores, frame rates in Return to Castle Wolfenstein, or rates in X-Plane. Apple System Profiler and the ATI software report seeing 128MB video RAM. Using GL Overrides or High-performance settings has no effect on the scores. The boot partition as plenty of free space for virtual memory swap files.

    The bottom line is that, after adding the ATI drivers, I now have poorer performance than before with the original Radeon card AND no change after installing the 9200.

    I contacted ATI nearly a week ago, received a ticket number and the usual promise of a 1-2 business day turnaround, but, as of this morning, have heard nothing. I know the Beige G3 is old, but X-Plane is a VRAM hungry beast and I expected to see some improvement from my investment in a 9200.

    Are you aware of any similar issues after the newer drivers were installed? My Mac shows no artifacts, and still works for routine functions. However, I had it where it was a passable flight sim and gaming platform. That seemed to have changed for the worse and I can't fix it.
    Thanks again for a great site
    Allan J. "

    I really don't care much for Xbench's graphics test (the opengl test is only a spinning squares test) but if anyone else has noticed a similar performance drop - let me know.

    Beige G3/OS 9.x/ATI Card/G4 Cache Enabler Conflict (and another note on sonnet ATA/133 card issue): (This has been a problem in the past with cache enabler software in os 9.x and ATI cards)

    (from 9/29/2004 email)
    I installed this card in my beige G3 DT frankenmac with OWC G4/500 zif (7400), running OS 9.2.2 (I'm going to try Panther with XPostFacto shortly.) I *thought* I did everything right: diabled previous graphics card drivers, installed ATI drivers, shut down, installed card, reset cuda, boot. First boot seemed to go o.k. On my second boot I lost both Maxtor drives attached to my Sonnet ATA/133 PCI controller. (see above for another report on problems from a Beige G3 owner w/Sonnet ATA/133 card installed-Mike)

    (I'm not sure the ATI card directly had anything to do with it. Drives attached to the Sonnet card have never been entirely stable for me as boot drives, even with the latest firmware, and I had randomly lost boot partitions 3 times before I installed Disk Warrior and used its protection, which basically makes disk cache write-through. Boots are faster through the Sonnet card -- but so are shutdowns, maybe *too* fast.)

    Fortunately, I had backed up everything 2 days earlier to an external firewire drive attached through a MacAlley Firewire/USB(1.1) combo card.

    The painful rebuilding process: I had to boot from 9 from the CD (this works with the 9200), reinitialize both disks though the onboard ATA/3 controller (the finder would just hang if I left them attached to the Sonnet card), and incrementally upgrade software to a working system.

    Good news: OS 9 and OS 9.1 ATI graphics drivers are not troublesome, but provide no acceleration.

    Bad news: OS 9.2.1 and 9.2.2 ATI Video Accelerator extension cause my box to hang with the 9200. (try disabling any cache enabler extension - or renaming it to change the load order. but if the ATA/133 card problem is common, then there could be other problems.-Mike)

    Bad news: 9200 drivers make my box hang if the G4 ZIF's backside cache is not enabled.

    Even more bad news, determined through Conflict Catcher: Powerlogix' cache enabling software conflicts with the 9200 drivers.

    BUT -- if you make the Powerlogix extensions load before the 9200 drivers, either through Conflict Catcher or by renaming the cache enabler (e.g., prepending a space to the name) everything seems to work fine.

    My box will now run Warcraft III. Yippee!

    I'm going to continue using a drive attached to the ATA/3 controller for OS 9 boots, but I intend to try Panther on a drive attached to the Sonnet card. Wish me luck.
    Best Regards, John"

    9200 PCI ROM Hotfix Feedback: (see link below)

    (added 10/12/2004)
    "Hey Mike, Just wanted to let you know that I have downloaded the new ROM update (Hotfix) for the ATi 9200ME and seems to be working very well. I have been collecting several components over the last few months in order to upgrade a B&W G3 rev.1 that I've had. I started putting a few things into it, among which was the video card. I installed the card a few days ago and it seemed to be doing fine until last evening. After the G3 had been running for several hours, I began to notice some of the screen artifacts that had previously been discussed on your site. (see screenshot samples in reports below) I downloaded and installed the ATi update and it seems to be working very well after running for about 8 hours. Thanks for providing the link.

    BTW, I'd be happy to send you a report on my upgrades if you like. I'm still installing things and working out some unexpected issues, but if you think that your readers would be interested, I'd be glad to share some info after I've finished. Among the upgrades I've chosen for this B&W G3 project are:

    • Sonnet Encore/ZIF G4 1000Mhz (see G4 reviews page for full review/comparison)
    • Sonnet Tempo Serial ATA PCI card
    • Two (2) Maxtor Diamondmax 300GB SATA HD
    • Sonnet Aria Extreme PCI
    • Pioneer DVR-107D
    • ATi RADEON 9200 Mac Edition Video Card
    • RAM upgrade
    • PRAM battery replacement

    Thanks for providing such a great site. I've gotten a LOT of help from your "G3-zone" and just thought I might give a little back. You really helped to make this project a lot easier than it might have otherwise been... Thanks again for all your help. Have a great week.
    Charles C. "

    (added 10/12/2004)
    "Mike, I have recently seen some of the original problem with "green pixels" on my replacement 9200ME, but only 5-7 pixels are showing up compared to what was originally happening with the card I returned to OWC and reported to you earlier.

    It was a surprise to see any at all, but since there were so few pixels showing up, I decided to simply monitor the situation and see what happens, then let you and Bruno know about it.

    In the meantime, the hotfix was released and I installed the hot fix yesterday. The few "green pixels" are now gone. So, the fix worked for my second 9200ME (v121 hotfix to v124). I'll let you know if the 9200ME exhibits any further "Green Pixels" or other oddities.
    -Rod "

    See below for Rod's original report.

    Mac Radeon 9200 PCI ROM HotFix for Display Corruption: (10/7/2004)

    " Hi Mike
    Just wanted to let you know that we've posted a new ROM flash specifically for the RADEON 9200 PCI. This HotFix was created to eliminate the corruption some people have seen on the finder (memory corruption to be specific). (some example screenshots in reports below.-Mike) This problem seems to only affect a very small percentage of cards and some people have simply opted to have their cards exchanged. With this update there is no need to send the card back to the original retail nor ATI. There is a flasher program for both Mac OS X and Mac OS 9.

    The ROM will be rolled into future production runs and will also be incorporated into an update of the Universal RADEON ROM installer first released a couple of weeks ago.
    Bruno (ATI) "

    If you saw this problem and this update solves it, let me know. thanks.

    ATI Radeon Firmware Updates/Displays v4.4 Update: (9/20/2004) ATI has released a September 2004 ATI RADEON Universal ROM Update for Retail Mac Cards and a ATI Displays version 4.4 update. (The firmware update hopefully may help with the artifacts some 9200 owners reported although it is not mentioned in the release notes)

    9200 owner feedback on Firmware Update:

    (added 9/21/2004)
    "Hi Mike,
    After all I did notice those green screen artifacts although to a much lesser extend then some people had.
    I downloaded and applied the new firmware and the old firmware was:

    ROM-revision: 113-A27502-120
    The new one:
    ROM-revision: 113-A27502-121

    The screen artifacts remain, and it wasnt mentioned in the firmware update that it would address this issue on the ATI Radeon 9200 Mac edition.

    So this still seems to be a remaining issue.
    I hope this helps, Jacob "

    For example screenshots of display artifacts some (but not all) 9200 owners have reported, see older reports below.

    (added 9/21/2004)
    The Radeon 9/20/04 update was applied to my replacement 9200ME, which I received from OWC today. This new replacement had ROM Version 113-A27502-120, which became ROM Version 113-A27502-121 after the update.

    This 9200ME replacement has been running about 6 hours with no sign of the pixel artifacts I originally reported with the first 9200ME we received. This is via the VGA port only, as my DVI cable was mistakenly returned with the defective 9200ME to OWC, which was then supposed to be sent to Bruno. I've asked Bruno to look for it in the original package. Hopefully, it is still there.

    In the meantime, this replacement card still displays fonts as if they are semi-bold, (a bit fatter) just as the original 9200ME did, compared to the older Apple/ATI Rage 128 card, which is much sharper.

    More later, when I have my DVI cable and can see if the fonts appear normal, as they did with the earlier 9200ME.
    My best, -Rod "

    (added 9/21/2004)
    "i can tell you that the screen artifacts arent gone, even after the ROM and driver updates. (not a driver update - just the Displays Control panel update-Mike)
    Jacob S. "

    NOTE: Reports below are from *before* the 9/20/2004 released Radeon 9200 Firmware update

    MDD Dual G4/1.25GHz: Follow-up from a reader that reported severe display artifacts in an earlier report:

    (added 9/20/2004)
    "The DVI in of my LCD monitor is connected to the DVI out of ATI 9800 while the VGA in of my same monitor is to the DVI out of ATI 9200 with a DVI-to-Analog adapter.

    The artifact only appears on the DVI main screen and supposingly only from the video buffer of the ATI 9800. So it is the ATI 9800 RAM buffer that's corrupted. When I switch over to the analog source which receives signal from the DVI out of ATI 9200, it performs well without any problem.

    It appears the presence of the ATI 9200 is interfering with the ATI 9800. Not sure if this is a graphic driver problem or the BIOS problem as it doesn't seem to be a wide-spread problem.

    BTW, I also happen to have an Alchemy TV DVR PCI card installed on the uppermost slot as one of the users with graphic artifact.
    -Wong Che Yung Kenny "

    B&W G3 w/PL G3 1GHz Upgrade: (follow-up report on using OS 9.x instead of 10.3.5.)

    (added 9/17/2004)
    "Hi Mike
    Just a final update on the 9200.
    I neglected to mention that the artifacts are happening in 10.3.5. There seems to be no artifact problem in 9.2.2. (interesting, most reports are from OS X users. If anyone else that reported artifacts can test in OS 9.x let me know-Mike)

    The artifact problem seems to get much worse after waking from sleep in Panther. On a fresh reboot the problem does not happen like the screenshot I sent that you posted. It gets that bad after waking from sleep.

    The fact there is no problem in 9.2.2 might point to a Panther conflict with the 9200.
    Thanks again, John
    On a side note the PL 1Ghz Firewire extension problem in OS 9 we discussed in the past seemed to be cleared up after I upgraded to 9.2.1 so the 9200 could run under the legacy. (I never noticed this in a review of their G3/800 ZIF in my B&W G3 in early 2003, running OS 9.2.2 on a B&W G3.-Mike) The FW extension conflict looks to happen if you are running 9.1 which I had been. Thought I would mention this for the record. "

    See below for John's previous comments on 10.3.5 w/screenshots.

    B&W G3:

    (added 9/17/2004)
    "Hi Mike, After reading the reports of other people I have tested my own ATI Radeon 9200 PCI.
    PCI Extreme disabled.

    The ATI Radeon 9200 PCI reports 128MB of RAM in both the 33mhz and 66mhz slot. (one previous B&W G3 owner oddly reported that the card was only listed as 64MB of ram if used in a 33MHz slot.)
    The profile is in Dutch but you can see that in the ATI Radeon 9200 functions with the full complement of 128MB in either slot. After each move I reset the CUDA so the results are reliable.
    The 33mhz slot next to the 66mhz slot:

      Type:    display
      Bus:    PCI
      Sleuf:    J11
      VRAM (totaal):    128 MB
      Fabrikant:    ATI (0x1002)
      Apparaatcode:    0x5961
      Revisiecode:    0x0001
      ROM-revisie:    113-A27502-120
      Type:    display
      Beeldschermtype:    CRT
      VRAM (in gebruik):    128 MB
      Resolutie:    1024 x 768 @ 96 Hz
      Diepte:    32-bits kleur
      Hoofdbeeldscherm:    Ja
      Synchroon scherm:    Uit
      On line:    Ja
      Status:    Geen beeldscherm aangesloten
    The 66mhz slot:
      Type:    display
      Bus:    PCI
      Sleuf:    J12
      VRAM (totaal):    128 MB
      Fabrikant:    ATI (0x1002)
      Apparaatcode:    0x5961
      Revisiecode:    0x0001
      ROM-revisie:    113-A27502-120
      Type:    display
      Beeldschermtype:    CRT
      VRAM (in gebruik):    128 MB
      Resolutie:    1024 x 768 @ 96 Hz
      Diepte:    32-bits kleur
      Hoofdbeeldscherm:    Ja
      Synchroon scherm:    Uit
      On line:    Ja
      Status:    Geen beeldscherm aangesloten

    I had no artifacts whatsoever in either 33 or 66 mhz slot, although after installing the lastest ATI accelerator 1.0.3 software I noticed that the card produces a lot of screen garbage when overclocked more then 10% in either memory or processor speed. (it's common to see artifacts if overclocking too far)

    I didnt notice any artifacts after playing Quake 3 Arena or Wolfenstein. I hope this helps. Regards, Jacob "

    Sawtooth G4/AGP (update w/tip on display artifacts he mentioned in his previous report)

    10/12/2004 Update: This reader wrote the firmware update solved the problem)
    Thanks for your note. I did plan to check in to confirm. We have success. The hotfix solved the artifacts problems.
    The monitor on the left is connected to a Beige G3 with a Raedon Mac Edition PCI card. The Cinema Display HD on the right is connected to a GForce4 MX, in my Sawtooth. The Mitsubishi Diamond Pro in the middle--and the small TV--are connected to the affected Raedon 9200. Before the hotfix, the artifacts issue was so chronic, I was a heartbeat away from pulling it and sending it back for a replacement. Now all these guys are working in harmony. The hotfixed Raedon 9200 has been running the center monitor for about 100 hours with no problem.
    (previous report follows, before the firmware update/fix)
    "update on artifacts issue:
    I followed Tom's (OWC) instructions and did the triple reset on my Sawtooth. Zapped the PRAM, did a PMU reset, and a NVRAM reset. When I booted after completing the series, the artifacts disappeared. The Radeon 9200 PCI card appears to be accepted and performing without obvious errors. I'm pleased to see the artifact problem appears to have been solved on my system, however, I'm not yet convinced that it's performing to it's full capacity.

    Also, I'm booting slow. When I do a command-v during restart, I see some PCI-card errors and confusion being chewed on, but the language exceeds my understanding. I'd be interested in knowing where to look for information to help me decode the startup logs.

    I should add that besides the Radeon 9200 PCI card and the GForce4MX AGP card, I also have an AlchemyTV DVR card. I'm a video hog. 3 of my 4 slots have video cards installed. I wonder if there are some inherent or unknown conflicts to this configuration.

    Each of my PCI slots is full. The remaining slot is a Sonnet ATA 133 (recognized in system profiler as UltraTek100Root) card, which drives my two IDE internal hard drives.

    In all other respects, before this latest video-card install (the new Radeon 9200 PCI card) my system has been performing well. I'm still hoping for news that could confirm whether or not there's a hardware or memory problem with the new Radeons. Otherwise, I may have compatibility-related or other unknown conflicts.
    Michael D.
    I asked him if he was using a DVI monitor or a VGA (analog) one since several artifact reports noted using the DVI port. (see next report below for example). Michael however later wrote he was using a CRT (VGA).)
    To confirm: I'm not using DVI ports. (though I wish I had a nice new flat monitor) Both are good-quality 22" CRT's, attached via VGA (analog) port, one monitor attached to the GForce4MX AGP card, one to the Radeon 9200 PCI card. My DVI ports remain unused.
    Hope this helps clarify things, as we try to determine the source of the problems with the 9200

    MDD Dual G4/1.25GHz: (5th report of display artifacts/image problems when using the *DVI* port, but not the VGA (analog) port.)

    (added 9/16/2004)
    "I can confirm that there is a graphic glltch with the new PCI card.
    Mine is a PowerMac 1.25GHz DP MDD that can boot into OS9, and currently running OSX 10.3.5.
    It was shipped with an OEM nVidia Ti4600 AGP card, 1.5GB of RAM and Combo DVD drive.
    My display is the SamSung SyncMaster 193T LCD monitor with both a DVI-I and an analog input.

    I upgraded the graphic card to the retail ATI 9800 Pro 128MB AGP and connected it to the DVI-I in of my monitor.
    Since the ATI 9800 doesn't support OS9, in which I still have some games and applications to run, I hoped to get the latest retail ATI 9200 PCI 128MB for better OS9 performance. It is placed well away from the ATI 9800 Pro 128MB AGP with an empty PCI slot in between so as to have more breathing room for heat dissipation, and connected to the analog in of the same monitor with a DVI to analog video adapter. I do not enable mirroring, so the ATI 9200 acts as an extension of my desktop when I press the "Source" button on my LCD to switch the input.

    The first time I switched on the Mac with the new PCI card in, I saw blocks of artifacts on the main DVI display (ATI 9800) but not on the extended desktop on the analog input (ATI 9200). I thought it is a driver issue and installed the latest software from the accompanied Driver CD of ATI 9200. (ATI Displays 4.30, not publicly available, only on the supplied CD) Indeed the problem went away upon restart.
    However, I got the same recurring display problem upon some prolonged overnight computer uptime. I have attached a screen shot. The artifact doesn't appear on the analog output from the ATI 9200. In fact I can move all these windows from the DVI main screen onto the analog extended desktop to continue working with the windows all normally displayed.

    Initially I suspect it may be a heat issue, and I check that the fan temperature is 53 degree C with Thermograph (I have actually added in 6 fans of different sizes to the chasis for better ventilation but that is another story. ) and the heat sink on both the 9800 and 9200 are very hot to touch. So I open the chasis and run the system overnight, and the same artifact appears on the main DVI screen though the temperature now drops to 49 degree C and all the heat sinks and graphic RAM chips feel just barely warm.

    I have also tried to tweak the ATI PreferencePane control panel settings, including the PCI performance tab but the artifact is still there upon restart no matter what the settings are. I haven't installed the ATIccelerator or any other hacks, including the PCI Extreme Quartz enabling modification for the older Jaguar system with PCI display cards. The only way to get the artifacts away is by shutting down the system for a short while and cold reboot again, then the artifact wouldn't reappear for quite a long time.

    My impression is that this is a driver issue. Whether this is related to the BIOS of the cards I 'm not certain.
    The ROM version of my ATI 9800 is 113-A07525-114 of FCode version 1.90 and NDRV version 1.0.0fDD.
    The ROM version of my ATI 9200 is 113-A27502-120 of FCode version 1.94 and NDRV version 1.0.1f34.
    Hope these information are useful to Bruno.

    And BTW, the manual of ATI 9200 showed more updated BIOS version than mine and I wonder where and how I can update to it. Perhaps Bruno can show me the light ;-)
    -Wong Che Yung Kenny "

    I asked him if Apple system profiler reported the card as 64MB vram or 128MB, as a B&W G3 owner reported earlier it ASP listed it as 128MB if in the 66MHz slot, but only 64MB if in a 33MHz PCI slot. (just wondering if that is common with other systems.)
    Note: if you see display artifacts and have both a DVI and Analog monitor, see if it appears on both interfaces or just the DVI one.

    B&W G3 systems w/PL G3 900 CPU upgrade:

    (added 9/16/2004)
    "B&W G3 w/PL G3 900 at (950MHz) No Fan or extra cooling, 640MB Ram 160GB/8MB cache WD + 80GB/8MB cache Maxtor, SIIG ATA133 PCI card, ATI 9200 PCI card
    B&W G4 w/PL G4/550 at 650MHz, Fan On CPU Heatsink, 512MB Ram, 80GB/8MB cache Maxtor + 40GB IBM GXP, ATI Radeon 7000 PCI
    Both Machines have multiple Partitions / 4 seperate OS 10.35 installs + two 10.3.4 + 4 OS9.22 partitions.
    Sorry this is not organized I am at work now.

    I first put the Card in the G4 650 Machine and did extensive testing mainly usuing Halo. Results were inline with those posted already for Halo. (G4 650/1MB got wiped by the G3 950/512MB)
    Halo is still playable in real life on the G4 but it likes CPU speed to push the 9200. Best Score for the G4 average 14.75 everything off 640/480 about 50% better then the 7000
    Best Score for the G3 average 21.9 everything off or low 800/600 (same for 640/480) Most of the Candy at med or High about 18.5 -19.5 avg (640/480) G3 950 of course.

    PCI Hi performance has almost Zero effect , sometimes a few points lower. (this setting may affect other PCI card performance like an IDE controller and there's probably some variation in run/run even on the same system/config.) The G3 1000/1MB and G3 1100/1MB cards should get another couple FPS esp with Eye candy on.

    Real World performance is all my Games that slugged on the 7000 or were unplayable now are very playable with almost all candy on and Higher Resolutions. (Its like the bench marks are not reading correctly )

    Before heading to the G3 though I thought I'd try ATIccelerator II 1.03 (overclocking utility for ATI cards) on the G4 650 When I first opened the control panel it read the memory at 400. I had it set at about 10% GPU and Memory on the 7000 Card previously and just left it. Set to default - Then tried about 5 % GPU then about 5% memory. I got immeadiate Artifacts like thouse illustrated all over the screen I quicky tried to lower the memory setting but Clutzed out and Bumped it up to 220 by accident - Locked the Machine.- Rebooted to Desktop. Artifacts were all over the screen But the desktop was not frozen. I then went quicky to the Control panel and set to default. No Change. REbooted to another Partition - Same thing.
    Pulled the Card and put it in the other BW with The G3 900 - No Artifacts - Same dismal perforamce in Open GL bench Marks.
    7000 Card usually gave 99 - 108 Open GL / 9200 was showing after repeated tests on both Machines 50-59 open GL X bench 1.0 or 1.3? (not sure what he's asking - xbench 1.0 is very old, xbench 1.1.3 is latest to date. still not a very useful opengl benchmark as it only does one 'spinning squares' test.-Mike)

    Had not tested any Games Yet (After I Did) on the G3. And Did not Try ATIccelerator II 1.03 again. The 7000 Card would go to 280+ GPU and about 230+ Memory before going Bonkers. Back to the store exchanged the Card for another new one. Said the Card was Faulty. Same Xbench Scores But MacBench in OS 9 gave me Hope.

    ( 2D ) Test Let 1000WindowsBloom seems to have some type of Barrier of 40 without QE on both Machines ( OS X Block ?) But when I tried PCI Extreme Later (to Enable Quartz Extreme on PCI Radeon cards) Near the end of Testing. ( 2Days )
    Prev Best G4 650 ATI 7000 WO / QE about 41 - 43 With 36 - 38 ( Have not tried yet with QE on with the G4 650 + 9200 )
    Prev Best G3 900 - 950 ( No change) ATI 7000 WO / QE steady 40 ( Barrier) With QE average 33 with a low of 32. ATI 9200 G3 950 WO/ QE Same 40 ( Barrier) With QE WoW 26 low of 25 (All with 3 trys )
    Something is happening here now.
    I tested 2D in OS 9.22 with Macbench 5.0 on the G3 machine Highest Graphics score for 7000/64MB (PC/flashed) = 5669 , (G4 650 about 3700- 3900 )
    Highest Graphics score 9200 = 6459 Both @ 950Mhz
    So The 2D is a little faster with the 9200 Card ( One of the few Strengths with the 7000 ) (I think he meant to say 2D was faster on the 7000 card per macbench scores-Mike)

    Bottom Line - I am pleased with the Card but would like to try a Faster G3 ( Yes More Mhz is the answer for Games ) (PCI bus and graphics chip can also be limits depending on game settings, but many mac games are CPU bound even w/high end graphics cards in faster AGP slot macs. See tests on video cards page here.-Mike)
    I am surprized in OS 10 the Benchmarks do not show the Difference between the Cards. BUT there is a difference.
    And I am dissapointed the PCI High performance option did not do much for My Machine like it Did the Quick silver ?

    All Test done in The 66Mhz slot
    PS. After reading all the info and my experience - Vram problem on some cards ? maybe.
    Hoping the Guys at ATI can improve the drivers for this Card.
    John S.
    Programmer Analyst
    (he later wrote)
    Further to my review yesterday. I mentioned that the Hi Performance PCI option had virtually no effect.
    Last night I was trying Halo At Higher resolutions to see how the performance degraded.
    Thought I'd check PCI option again to see if it had any effect. It being pretty late so I only tried the 2 combinations and verified them.

    BW G3/900 at 950MHz

    Halo 1.05
    Normal Performance PCI
    1024x768 60Hz Everything off or on Low
    average score 20.21 FPS ( almost the same score as 640 * 480)

    Normal Performance PCI
    1024x768 60Hz Model Textures on Med - Particles on High - Textures on Med
    average score 17.28

    Hi Performance PCI
    1024x768 60Hz Model Textures on Med - Particles on High - Textures on Med
    average score 19.45

    Hi Performance PCI
    1024x768 60Hz Everything off or on Low average score 20.85 FPS

    This indicated to me that as the Data load increased the Hi Performance PCI started to finally work.
    Over 2 FPS increase ( Before I was only getting .1 or maybe .2 or - .1 ) It looks to me as if My CPU is part of the bottleneck here.

    I know the Card will go faster based on the Quick silver scores.
    It would be nice if another BW owner with G3 1000/1MB or G3 1100/1MB would test this to see if the Hi Performance PCI can push the Card higher at this resolutions with extras on. "

    Sawtooth G4/AGP: (the 4th report on display artifacts - see his later report w/tip above.)

    (added 9/15/2004)
    "Hi Mike,
    Just installed the ATI 9200, and I'm having problems, too. More details if interested, but since a user asked for feedback on this specific thing, here's an attached image file illustrating artifact issues.

    My system is a Powermac G4 APG Graphics (Sawtooth) with 1.38 gigs RAM, 1.2 GHz (Gigadesigns) processor, NVIDIA 32 MB (geforce2MX? 4MX?-Mike) in APG slot, new ATI 9200 in PCI slot. Running Mac OS X 10.3.5 (7M34)

    Is this bad ram? Do I need to consider sending the new graphics card back to be looked at?
    (he later wrote)

    GeForce4 MX: 
      Type: display 
      Bus: AGP 
      Slot: SLOT-A 
      VRAM (Total): 32 MB 
      Vendor: nVIDIA (0x10de) 
      Device ID: 0x0172 
      Revision ID: 0x00a5 
      ROM Revision: 1121 

    Sorry--my earlier note was rushed. Above is my primary card, in the AGP slot, a GForce4MX. I should also add that my computer...

      Machine Model: Power Mac G4 (AGP graphics) 
      CPU Type: PowerPC G4  (3.3) 
      Number Of CPUs: 1 
      CPU Speed: 1.2 GHz 
      L2 Cache (per CPU): 256 KB 
      L3 Cache (per CPU): 2 MB 
      Memory: 1.38 GB 
      Bus Speed: 100 MHz 
      Boot ROM Version: 4.2.8f1 

    ...choked and was disoriented for a while after installing the new PCI Radeon 9200 card. I zapped the PRAM, but couldn't boot successfully for a while, had to boot from another disk, then back again. I endured long boot times until it finally settled down. For about two days the newly-installed PCI card performed normally, then on the third day, artifacts began appearing. I'm not sure it's ever really performed correctly, it seems no faster or better than the 4-year-old PCI (32 meg) Radeon Mac Edition that this new card replaced.

    Based on this description, OWC's Tom, in a recent note, recommended additional resets--PMU reset, and an NVRAM reset--suggesting my system might not be recognizing the card properly. (it might still be oriented toward the older Radeon Mac Edition card in that slot) which I've not done yet. I plan to do these additional resets at the end of business day. I'll report back if I get good results.
    Having said that, here's how my system currently recognizes the Radeon 9200

      Type: display 
      Bus: PCI 
      Slot: SLOT-B 
      VRAM (Total): 128 MB 
      Vendor: ATI (0x1002) 
      Device ID: 0x5961 
      Revision ID: 0x0001 
      ROM Revision: 113-A27502-120 

    Hope this helps clarify. "

    I forwarded this mail to Bruno of ATI who's investigating this problem from previous reports here (below). I've heard of problems in the past with some mixes of Nvidia and ATI cards, but from the looks of this I'd suspect a card/vram problem (image corruption/dots, etc.), however I'm still waiting to see if ATI suspects other causes.
    Note: If any users have PCI Quartz Extreme enabled - try disabling it to see if that matters. Also toggle the PCI High Performance setting in the ATI control panel just as a test.
    If you see display artifacts and have both a DVI and Analog monitor, see if it appears on both interfaces or just the DVI one.

    B&W G3 w/PL G3 1GHz Upgrade: (the 3rd report on display artifacts - see his later report above dated 9/17/2004 where he said he does not see the problem in OS 9.2, but does in OS X 10.3.5)

    (added 9/15/2004) "Hi Mike, I am having the same type of artifacts as the screenshots sent in by Rod. (see Rod's report and screenshots below in 9/13/2004 dated report) It does not happen all the time. Only in certain applications. Suitecase and Photoshop seem to leave these green artifacts behind only the left edge of windows when you move them around almost like a ghost edge is left behind. When I quit the apps the artifacts go away. Other apps do not have this problem.

    Does this look like a straight vram problem or is it something that ATI can fix driver wise.
    Thanks for any feedback.
    B&W G3/1GHz Radeon 9200 in the 66mhz slot.
    (he later wrote) Let me first add that my monitor is the original 17" Apple blue pod monitor that shipped with the B&W's.

    PCI Quartz Extreme is not enabled. I never installed it.
    I toggled the high performance settings but it did not seem to make a difference.

    The odd thing is it happens differently for the apps that are effected. For Suitcase bringing up the Suitcase window I get the trailer affect coming from the left edge of the window. For Photoshop it seems to only happen for save dialogues and action windows like Levels, channel mixer etc.

    I will attach a screen shot for an example. In the image the artifact bands are created only moving the save dialogue from left to right. When I move the window back over the artifacts they disappear.
    The card really performs great aside from this problem.
    Just another follow up. It seems as though the artifacts have gotten worse since emailing you this morning. It is now happening with just about every app and even Finder windows. It only occurs as stated before from the left side of a window.

    I have tried changing settings in ATI display but non seem to make any difference.
    I have attached another screen shot to give a clear picture of the problem.


    B&W G3 w/PL 1GHz ZIF Upgrade (another mail noting display artifacts - maybe there's a bad batch of cards out there?)

    (added 9/14/2004)
    Mike, Just received the 9200PCI and installed it, along with a 1GHz PowerLogix ZIF, in my B&W rev.2. Bad news is, when installed in the 66MHz slot, there is continuous vertical pink banding on the monitor and what looks like electrical artifacts on the screen. The artifacts worsen when playing Warcraft III. I tried both the VGA port and the DVI port with the included adaptor, but to no effect. I removed the card, blew out the slot, reinstalled the card - no effect. I installed the card in a 33MHz slot - all symptoms disappeared. I reinstalled the card in the 66MHz slot - symptoms reappeared.

    The new ZIF is not a factor, as replacing it with the old ZIF altered nothing.

    So, although I now have a radically faster machine overall, I'm still bummed because I feel like I should have even better performance but can't make use of the faster slot.
    (I asked for more details like os version used, if he installed the 4.30 displays software from the cd, monitor used, etc. and if what he saw looked like the images in the report below)
    Not really. It was giving me vertical pink bands, subdued but consistent. The artifacting looked like regular, rectangular bricks of noise, that more or less fuzzed in and out, but were all over the screen pretty much all the time. Since I sent you mail, I downloaded and installed all the latest 10.3.5 updates, and the situation has changed a bit.

    There are no longer pink bands, but there are almost regularly spaced, ghosted columns of horizontal bricks, all across the screen when on the regular desktop. Gaming produces more static-appearing artifacts. Weird. Again, this only occurs in the 66MHz slot.

    Monitor is a crt ViewSonic 17GA -- not new, but perfectly servicable. Normally is bright, crisp, good color.
    I'm now running 10.3.5, all updates. Installed the driver from the CD. It's a pretty recent format & re-install of Panther, due to massive drive corrupting (I believe from crashes resulting from an overheating ZIF). I've been using it as a gaming machine for my kids, no critical files on it at all. I've tried several different screen resolutions and frequencies with no noticeable effect, finally leaving it at 1024x768, 85MHz. No other PCI cards at the moment. 640MB RAM, 30GB DeskStar drive.

    Rudely, if I leave the card in the 66MHz slot, ATI Displays registers the full 128MB VRAM, but when I move it to the 33MHz slot it only registers 64MB VRAM. (Odd...)

    I'll probably get in touch with ATI tomorrow, and will pass on whatever I find out.
    (I asked if he had PCI Quartz Extreme enabled and if he'd tried toggling the ATI Control Panel PCI Hi-Perf. settings)
    I've never used the PCI Quartz Extreme hack, so that doesn't apply. I had already toggled the HP setting with no noticeable result. "

    I forwarded Jim's mail to Bruno of ATI also.

    B&W G3 w/PL G3 800 CPU Upgrade:

    (added 9/13/2004)
    "Mike, Received a 9200ME from OWC yesterday, for use in a B&W G3 with 800MHz CPU, to replace the original Apple/ATI Rage 128 card in the 66MHz PCI slot.

    The reason for examining the card is considerably different from what most users are interested in, which is game or motion video speed. I want to determine just what kind of improvement can it make to 2D display speed, for things like displaying data via Intranet access, initial inspection of digital photos, searching large archived text docs and other tasks associated with content creation editing and repurposing. The parties interested have a large number of B&W G3's, which serve them well, but they desire a single improvement... greater video display speed to improve productivity with the tasks they are performing.

    The ATI 9200ME appears to support this need, unfortunately my unit appears to be malfunctioning, as the video from the DVI-I port disappears at random times for several seconds (the screen goes black) and video from both ports starts generating random pixels trails when windows are moved. If I restart the B&W G3, the random pixel generation stops, but is back again in 5-20 minutes, depending on how busy the card has been. It appears to be memory related, not heat related.

    Further, there is a considerable difference in font display characteristics, between the two ports, when viewed on the same LCD display. The 9200ME's DVI-I port looks like the original ATI Rage 128 output, but the 9200ME's VGA port makes the fonts look as if they are semi-bold. I've tried this on three different brands of LCD's, and the effect is the same. Clearly, this will bother some experienced font users, who can tell what the fonts is on site and expect accurate font display quality to assist in doing this.

    I'll get back to you with some speed numbers, showing the difference between the ATI Rage 128 and the 9200ME doing the kind of tasks I've described above, as soon as I can get a replacement 9200ME.
    (he later wrote)
    A follow-up to my earlier E-Mail...
    I've attached several pictures of the artifacts that the card creates, when objects are moved in a horizontal plane on the desktop after 15 minutes or so. You can't actually do a screen capture, because that causes the screen to refresh and the artifacts are replaced. Restarting the Mac corrects the problem, which returns in another 15 minutes. The problem exists with OS X 10.3.5, 10.3.4 and 10.3.3 with and without the ATI Control Panel installed.

    The card's DVI-I output goes to black at random times for about two seconds. The VGA output does not do this.

    My B&W Config is:
    B&W G3 Rev. 2 with PowerLogix 800MHz G3 ZIF CPU, 1 GB RAM, dual 80GB WD 8MB cache drives via SIIG PCI/ATA card, ZYNX 100Mbps Ethernet PCI.

    jpg of OS X drive window with artifacts caused by moving the window horizontally.

    (images were 300+KB and 1.3MB so I trimmed/compressed them down for bandwidth)

    jpg of Eudora window with artifacts caused by moving the address book window horizontally.

    The primary display is a Sony 19-inch LCD SDM-S94/B, 1280x1024 with DVI-I (28-64KHz x 60Hz) and VGA (28-80KHz x 48-75 Hz).

    Two other displays were attached (VGA only) and they both showed the same pixel artifacts. An HP 15-inch LCD Pavilion vf51 at 1024x768 and Samsung 15-inch LCD SyncMaster 152v at 1024x768. The HP also showed the "fattening" of the screen fonts, making them appear semi-bold. The Samsung looks like this all the time, so it is incapable of showing small display changes in screen object appearance.

    I've seen this same type of artifact on earlier model Twin-Turbo 128 and SuperMac video cards with faulty VRAM, so I suspect this is the problem with the Radeon 9200ME.

    The HP and the Sony displays appear identical when attached to the Apple/ATI Rage 128 card in the B&W G3.
    -Rod "

    I suspect this is a bad card (bad vram, etc.) but sent a note to Bruno at ATI. Rod has an RMA number from OWC and Bruno of ATI also said they're going to examine the card when it's returned to ATI. If any other 9200 PCI owner sees a similar problem let me know.

    Yikes G4/PCI w/G4 ZIF at 600MHz:

    (added 9/8/2004)
    "...I ordered from OWC on the 7th... though the only game I own with a viable benchmark is Q3. It will be interesting to see how improved it is over my current Radeon 7000 (G4/600, 1GB RAM)

    Just got done with some preliminary testing and WOW, this card screams! My machine is a Yikes G4/600MHz (100MHz bus), 1GB RAM, and the Radeon 9200 in the 66MHz PCI slot.

    Quake 3 (sound on)
    "High Quality" (800x600)
    Radeon 7000: 36.1 fps
    Radeon 9200: 61.3 fps
    "High Quality" (1024x768)
    Radeon 7000: 27.4 fps
    Radeon 9200: 60.8 fps
    "High Quality" (1280x768)
    Radeon 9200: 57.4
    I didn't even bother to test the 7000 at this resolution.

    In Warcraft III, the 9200 plays the game absolutely beautifully at 1280x1024 with all settings on high! Totally smooth framerates, no hiccups at all. Even the menus drop down perfectly smooth, something I never saw with the Radeon 7000! With the Radeon 7000, I was forced to play at 800x600 with settings on low. I never knew how beatiful this game actually was.

    I must mention the 9200 install CD comes with ATI Display utilities version 4.3.0. As far as I know, the most current downloadable version is still 4.2.5.
    Mark "

    B&W G3 with G4 ZIF at 600MHz:

    (added 9/10/2004)
    "Hi Mike, This is my impression of the ATI Radeon 9200, first let me explain the setup I am using:

    B & W G3 Rev.1 with a 600 mhz G4 (overclocked 500mhz OWC G4 / active cooling for the CPU with a fan). 640 MB RAM PC100, 2 x 80 GB S-ATA Maxtor 6Y080M0 connected to a Seri-Tek1S2 S-ATA PCI adapter. NEC ND-2510A DVD R/RW writer. ATI Radeon 9200 PCI. M-Audio Revolution 5-1 PCI hooked up to a Trust 5-1 speaker set.

    Mac OS X 10.3.5 latest updates/patches.
    ATI supplied Radeon 9200 drivers High performance PCI enabled.
    PCI (Quartz) Extreme enabled.

    When I installed the ATI Radeon 9200 PCI I noticed a better and crisper image then provided by the ATI Rage 128. With Quartz extreme and High Performance PCI enabled I did not have the artifacts (yet) of which other people report when using the PCI extreme hack.
    Quicktime video is smooth, preview of Quicktime in the Finder is normal. No strange dock behavior.
    No system slowdowns after long time intense use (gaming Quake 3, etc.) When I played a DVD video (full screen), no dropped frames or out of sync.

    There was little or no difference in the performance of Quake 3 Arena (1.32) either running with or without Quartz extreme enabled. (Quartz Extreme is for the 2D desktop - it should not affect quake3/3d games unless you left a bunch of 2D windows open in the finder maybe)
    I maxed out all the settings under 1024 x 768 and gameplay was very smooth, with no dropped frames.
    I did not measurements, but I surely can recommend the ATI Radeon 9200. Very good 2D and 3D performance in my case.
    Jacob "

    Some have noted that PCI enabled Q.E. can hog the PCI bus bandwidth if you're really stressing the card (i.e. with lots of 2d windows open, etc. - the PCI bus can get totally saturated, starving other PCI cards if you were using them for instance) But with 128MB of VRAM, that's a lot better than having some of the other 32MB VRAM Radeons so the problem would not show up as quickly/easily I bet. See the comments on the older page on Feedback on Hack to Enable Quartz Extreme for Radeon PCI cards where Tim Seufert comments on Why Apple did not enable PCI QE by default.

    PowerMac 7500 w/PL G3 900:

    (added 9/8/2004)
    "Just wanted to inform your readers that I have gotten the 9200 running in my 7500 w/PL g3/900 ZIF. (in CarrierZIF or ZForce card I assume-Mike) I don't have any benchmarks to share right now but I can confirm that with PCI Extreme, Quartz Extreme is enabled and it works perfectly.
    I did have one problem following installing the card. After installing the drivers on the CD I lost the ability to use any menu items anywhere (menu bar, dock, drop downs). I had to reinstall 10.3.5 and then the Radeon drivers and worked fine. Not sure what happened there but all is good now. I can now play DiVx/SVCDs at full screen, full frame rate in QT that previously was not possible. Definitely satisfied with this upgrade.
    (he later wrote)
    Just one thing to add to my previous note about having the 9200 work w/PCI Extreme in my 7500. I did have to use the custom installal and choose the PCI only option. I initially tried the PCI and AGP as suggested in the documentation but that did not enable Quartz Extreme. -Corey
    (System Details) PL G3/900, Radeon PCI 9200, 304MB RAM, Sonnet Tempo ATA Card 12GB WD IDE, Khypermedia 48x24x48 IDE CD-RW, 3GB IBM SCSI "

    G5 owner general comments:

    (added 9/8/2004)
    "Hey Mike, I just wanted to let you know that i got my Mac 9200PCI in today. I popped it into my G5 and booted it up. One interesting thing was that it came with the manual for the 9800Pro. They must've rushed it through, so they didn't get to print a manual. The CD however was specific to the 9200PCI and includes both an OS X and OS 9 installer. It installs version 4.3.0 of ATI Displays.
    I can confirm some specs of the 9200PCI. It has a RV280 GPU running at 250Mhz, and then nice Hynix RAM running at 200Mhz. On the box it claims to have 4 pipelines as opposed to that of the 9200 i heard about in some rumors with only 1.
    There is an option of PAL of NTSC TV output. I have created a site where i have all of this data as well as eventually benchmarks comparing it to the flashed 9100PCI (i can't do them right now because i only have an ADC monitor for my G5's 9800Pro). The site is here:
    people.vanderbilt.edu/~michael.j.hannick/mac9200.html. -Michael "

    Return to Main News Page

    (Ad/Sale items)

    = UPGRADES by Mac =
    Upgrades just for
    YOUR Mac!

    = Refurb Mac Pros =
    (Click for List)

    SSDs up to 4TB
    Fast SSDs for Most Macs/PCs

    = ThunderBolt =
    Drives, Docks & More

    Up to 12TB HDDs
    HGST, WD, Seagate, Toshiba

    = 2.5in HDs & SSDs =
    Notebook Hard Drives and DIY drive/case kit bundles

    = MEMORY =
    Lifetime warranty RAM Upgrades!

    Internal and External Superdrives/Blu-Ray drives

    Graphics cards, Displays, Adapters, Cables & more

    Interfaces, Cables, Software, Speakers, Headphones & more

    = SOFTWARE =
    Apps, Utilities, OS, VM, Games and more

    = WIRELESS =
    WiFi and Bluetooth Devices/Adapters/More

    = Repair Service =
    for iPhone, iPad, Macs

    = iPad/iPhone/iPod =
    Accessories, Cases, Repairs & More

    NuGuard KX Cases
    *Extreme* Drop Tested!

    XLR8YourMac T-Shirts

    = back to www.XLR8YOURMAC.com =

    = Other Site Topic Areas =
    Mac Mods/Upgrades | CPU Upgrades | Storage | Video | Audio/HT | OS Updates/Network | Recent

    Copyright © 1997-2018. All Rights Reserved.
    All brand or product names mentioned are properties of their respective companies.

    Legal: Site Privacy and terms/conditions of use.