Click for Data Doubler kits!
Click for Data Doubler kits!


One Click on the Banner shows support to my Sponsors

Accelerate Your Mac!  - the source for performance news and reviews
The Source for Mac Performance News and Reviews
Don't forget to check out all the other site features!

Photoshop 5.5 Performance:
Apple G4/500 & Dual G4/450 vs Pentium III 1GHz

PSBench's 21 Filter Test Results

Published: 6/20/2000
(updated 9/5/2000 for G4/450 MP System Results)


Introduction:
Is it fair to test a 1GHz PC against a 500MHz G4 Mac? Of course not. Imagine the comments from PC owners if it were the other way around. Granted 1GHz PCs are available for about the price of a 500MHz G4 system (depending on configurations). As of the original date of this review, a bare-bones G4/500 (64MB RAM/10GB drive/DVD ROM) costs about $2299 at the Apple store, the HP Pavilion 1G (1GHz PIII/128MB Rambus RAM/40GB drive/Geforce 32MB DDR/10X DVD/4x CDRW/100BaseT NIC) sold for $2799 as of June 2000. The standard G4/500 configuration however costs nearly $3500 with 256MB of RAM, DVD RAM drive and 40GB hard drive. (Update: At Macworld NY in July 2000, Apple revised the G4 systems so that dual processor 450MHz and 500MHz G4 systems are now priced the same as the previous single processor systems were, and include Gigabit Ethernet as standard.)

I was able to test a friend's Pavilion 1G (upgraded with 128MB more PC800 RAMBUS RAM) against my G4/500 and am posting the results only to satisfy those that wanted to see Photoshop 5.5 PSBench 21 filter tests between these systems. Previous Mac vs PC tests on the web or in print used only a few filters. This is just a test of Photoshop performance - games performance or other application performance would show different results. (The PC is a far faster gaming machine for instance - see the table of game results below.)

NOTE: All tests here were run with a Photoshop 5.5 History setting of 1 history state and 'create first snapshot automatically' disabled. The defaults are 20 history states and creat snapshot enabled. (The history settings changes were done per a tip from a Photoshop user to improve PS 5.5 performance). The revised settings did make for more consistent filter tests - each filter is run 3 times; with these history settings there was less variation in the times of the repeated tests.) As per the PSBench instructions, 1024x768, millions colors with Photoshop interpolation preferences set to 'bicubic (better)'. For these tests, both machines had Photoshop set to use approximately 50% of available RAM (256MB RAM installed in each machine, PS allocated 120MB). For best results with Photoshop, you should allocate up to 5 times the image size in RAM to avoid any swap file activity which reduces performance.

The results below are with the 10MB test image size. I tried to run the 50MB version but the PC repeatedly reported page faults at Plugin.dll when running the Polar Coordinates filter, even when memory was boosted to 75% of available RAM. (Norton Utilities reported the RAM in the PC was ok.) If I can get access to the PC again I'll try tests with the maximum possible Photoshop RAM allocation.


Update: G4/450 Dual Processor System Results: A reader sent the results of his new (summer 2000) Dual G4/450 system with PSBench. Although I did not witness these results, I did remind him of the system settings, etc. to use. His tests used the latest 7/19/2000 Adobe Altivec Core Plugins update (available here) and only a 128KB Cache (my G4/500 system used the previous Altivec plugin versions and used OS 9's detault disk cache settings).

Configurations:

  • Apple G4/AGP
  • 500MHz G4 CPU (speculative processing enabled)
  • 1MB of L2 Cache at 250MHz
  • 100MHz system bus
  • OS 9.04 with Firmware update
  • Virtual Memory Off
  • 256MB PC100 RAM (222)
  • 45GB IBM 75GXP IDE Hard Drive
  • 6X DVD ROM (IDE)
  • ATI Rage128 PRO AGP w/16MB SDRAM

  • Apple G4/AGP (Reader's system)
  • Dual 450MHz G4 CPUs
  • 1MB of L2 Cache at 225MHz (x2)
  • 100MHz system bus
  • OS 9.04 with Firmware update
  • Virtual Memory Off, 128K Disk Cache
  • 256MB PC100 RAM (one 222 dimm, one 322 dimm)
  • ATI Rage128 PRO AGP w/16MB SDRAM

  • HP Pavilion 1G
  • 1GHz Pentium III (Coppermine)
  • 256KB of on-die L2 Cache at 1GHz
  • 133Mhz system bus
  • Windows 98 Second Edition
  • Virtual Memory Enabled
  • Intel VC820 Motherboard
  • 256MB PC800 RAMBUS RAM
  • 40GB Maxtor IDE Hard Drive
  • 10X Pioneer DVD ROM (IDE)
  • 4x/4x/24x CDRW (IDE)
  • Hercules Prophet II 64MB DDR (Nvidia GeForce 2 GTS)
    (owner replaced original GeForce 32MB DDR card)
  • Intel 10/100 PCI network card
  • SoundBlaster Live PCI audio card

All systems were running Photoshop 5.5 (G4 had the latest version of the Adobe Altivec Extensions). Windows background tasks like McAfee's anti-virus and the task scheduler were disabled. The Windows swap file was set to the same min/max size (384MB) - a common tip to improve performance by preventing Windows from resizing the swap file during apps use).

The PSBench scores below are the times in seconds that each machine took to perform each of the 21 filter operations on a 10MB image. Photoshop was allocated 120MB of RAM on each machine.


PS5Bench Test
(10MB Image File)

HP Pavilion 1G
1GHz PIII

256KB L2 at 1GHz
133MHz bus
256MB PC800 RAMBUS
Geforce 2 AGP (64MB)
Win98 SE VM On
Apple G4/500 AGP
500MHz G4 CPU
1MB L2 at 250MHz
100MHz Bus
256MB SDRAM (222)
Rage128 Pro AGP (16MB)
OS 9.04 VM OFF
Apple G4/450 MP
Dual 450MHz G4 CPU
1MB L2 at 225MHz x2
100MHz Bus
256MB SDRAM (222/322)
Rage128 Pro AGP (16MB)
OS 9.04 VM OFF

Rotate 90

0.2

0.2

0.2

Rotate 9

1.6

2.0

1.3

Rotate .9

1.5

1.9

1.2

Gaussian Blur 1

0.5

0.5

0.4

Gaussian Blur 3.7

1.3

1.4

0.9

Gaussian Blur 85

3.2

1.6

1.1

Unsharp 50/1/0

0.6

0.7

0.5

Unsharp 50/3/7/0

1.4

1.7

1.0

Unsharp 50/10/5

1.6

1.7

1.1

Despeckle

1.7

0.8

0.5

RGB-CMYK

2.6

4.0

4.4

Reduce Size 60%

0.6

0.4

0.3

Lens Flare

2.3

3.3

2.2

Color Halftone

3.5

3.0

3.1

NTSC Colors

3.2

3.4

3.7

Accented Edges

7.4

8.9

9.6

Pointillize

8.7

12.5

7.9

Water Color

15.8

18.7

20.2

Polar Coordinates

5.7

2.9

2.3

Radial Blur

21.2

29.9

16.7

Lighting Effects

1.2

1.7

1.2

PS5Bench Index
(time to complete)

85.8
Seconds

101.2
Seconds

79.8
Seconds

System

HP Pavilion 1G
1GHz PIII

256KB L2 at 1GHz
133MHz bus
256MB PC800 RAMBUS
Geforce 2 AGP (64MB)
Win98 SE VM On
Apple G4/500 AGP
500MHz G4 CPU
1MB L2 at 250MHz
100MHz Bus
256MB SDRAM (222)
Rage128 Pro AGP (16MB)
OS 9.04 VM OFF
Apple G4/450 MP
Dual 450MHz G4 CPU
1MB L2 at 225MHz x2
100MHz Bus
256MB SDRAM (222/322)
Rage128 Pro AGP (16MB)
OS 9.04 VM OFF

Game Tests: Since many wrote to ask about 3D Game performance, here's a (depressing) comparison. I'm in no way saying these tests are fair - twice the clock speed PC, with 4X AGP vs 33Mhz PCI, more mature drivers for the GeForce card, etc. etc. The point is to show just how far behind the TNT is to the current GeForce2 cards. (TNT2 scored under 27FPS at 1024x768/32bit HQ in a 1GHz PC.) Most 1GHz PC buyers would want a GeForce card (many like the HP 1GHz shipped with one).

1GHz P3/GeForce2 64MB DDR vs G4/500 Voodoo5
Card Tested 16-Bit 32-Bit
(P3 1GHz) GeForce2 4xAGP 1024x768 HQ 113.9 88.6
(G4/500) Voodoo5 PCI 1024x768 HQ 50.2 49.1
(G4/500) Voodoo5 PCI 1024x768 2xFSAA 45.7 23
(P3 1GHz) GeForce2 4xAGP 1024x768 FSAA 69 38

Driver v1.0 used for Voodoo5
Driver v5.22 used for Geforce2, has only 1 FSAA setting

For the latest Mac video card reviews/tests, see the Video articles page.




Summary:
Personally, I was impressed at the Photoshop performance of the G4/500 compared to a 500MHz faster PC running a 33% faster memory bus and Rambus RAM. (Yes, I've read the web reports that PC133 RAM is as fast or perhaps faster in some tests with less latency than Rambus RAM, which is only 16bits wide, but at higher clock speeds.) Of course for games performance the results would have been far different and other applications would show different results (see the ZDnet review in the related links below for instance, but the G4/500 they used had only 128MB of RAM). The G4 dual processor systems performance should improve under OS X, which has a fully SMP core not found in the current MacOS which should take better advantage of dual processor systems.

And remember that Photoshop 5.5 supports the PIII's SSE extensions (128Bit), which shows that Altivec performance (in Photoshop at least) is impressive. Imagine if the G4 had been running at the same clock speed of the Pentium III. Hopefully Motorola or IBM will be able to produce faster G4s in quantities soon.

I fully realize that in other applications (and especially 3D Games) the PC would show a much larger advantage. This article was just to satisfy reader requests for a Photoshop 5.5 comparison.

Related Links:

  • (New) Apple Dual G4/500 Review by PC Magazine. The review includes applications test results comparing the dual G4/500 to a dual 1GHz Pentium III in several cross-platform apps.

  • PCMag's G4/500 (single CPU) vs Fast PCs Includes many cross-platform applications tests. The G4/500 had only 128MB of RAM and doesn't have dynamic memory allocation like the PC. For Photoshop on the Mac, the RAM allocation can have a dramatic effect on performance. (Macaddict noted this in their very limited tests of G4/500 vs 1GHz Athlon.) As I noted before, for best results with Photoshop, you should allocate up to 5 times the image size in RAM to avoid any swap file activity which reduces performance.

  • Photoshop 5.02 Tests - Macs vs PIII 500Mhz (Win98) and Dual PII 450 (NT 4) Systems

  • Many G4 CPU Upgrade Reviews have Photoshop 5.5 Comparisons on the Apps tests page of the reviews. (See CPU Upgrades page for list of reviews)


Back to XLR8YOURMAC.COM
Your Source for the best in CPU/SCSI/VIDEO card reviews, daily news, and more!

Copyright © Michael Breeden, 2000.
No part of this site's content is to be reproduced in any form without permission.
All brand or product names mentioned here are properties of their respective companies.
Disclaimer: Users must read and are bound by the Site Terms & Conditions of Use.