Accelerate Your Mac! - Railgun 466/233/1MB ZIF Review

A Click shows your site support to my Sponsors

The Source for Mac Performance News and Reviews

Review: BottomLine's Railgun G3/466/233/1MB ZIF
Published: 6/28/99

Software Cache Control Features

Intro | Benchmarks | Appl. Tests | Software Controls | Docs/Installation | Specs/Design | Summary

Cache Control Software

NOTE: Bottom Line and Mactell are now out of the CPU upgrade business. The last source for Manuals and software for Railguns is listed in the CPU Upgrades section of the FAQ. For cache control software, Powerlogix.com's free cache profiler usually works also (but disable any existing Railgun extensions/controls before switching to a different cache enabler.)

Bottom Line's Railgun comes with Mactell's new PowerJolt cache control utility and an extension for cache enabling. However in the Apple Beige and B&W G3 the Railgun ZIF, like the XLR8 and Vimage ZIFs, does not need any software to enable the backside cache. For tests in the Apple G3s, I did not use the PowerJolt software, as I prefer run lean and mean (without any unnecessary software when possible). I did use the PowerJolt software when running the Railgun ZIF in the PowerCenter Pro with the CarrierZIF card since software was required to enable the cache in pre-G3 Macs. I also noted as shown below that the PowerJolt control panel did not accurately display the bus speed in the PowerCenter Pro. This was verified by checking speeds with XLR8's and Powerlogix's utilities. Both agreed with the card's actual speed, and simply dividing the CPU speed by 8X ratio (max) proves the Powerjolt reported bus speed was not correct. Not a big deal, but along with the cluttered interface, it was another reason I prefer to use XLR8's cache control (which works fine with the Railgun).
From what I have seen in real world applications tests, backside cache speed is not a primary factor in performance so don't push the bleeding edge on cache speed as it can affect reliability.

Installing the Software:

The XLR8 software installer automatically places the cache control and extension in the system folder. After a reboot, it tests the backside cache automatically and reports what it suggests is the maximum reliable cache speed (but offering you the option to manually set the speed to another setting). Often this speed may not be reliable in my experience, as it reported cache speeds of up to 296MHz were OK on the review ZIF, which I found later to be too fast. Only the expensive (and rare now) IBM cache can run reliably near 300MHz speeds in my experience (Some reviews like the PF 466/311 ZIF and older cards like the XLR8 G3/266 show pictures of this cache if you're interested.). I manually set the cache to a 2:1 ratio (half the CPU speed), which ran perfectly at speeds of up to 225Mhz.

PowerJolt Cache Control
(PowerCenter Pro speeds shown)

Advanced settings page

Note the 67C temperature using the supplied heatsink, which was the highest temperature I've ever seen from a G3 desktop CPU, yet there no errors of any kind during the PowerCenter Pro tests. The stock heatsink would not let the PowerCenter Pro's fan/bracket fit, so I'd suggest using the CarrierZIF's stock plate heatsink which does allow using the fan/bracket and should dramatically help cool the CPU. In the B&W G3, at 550MHz a 39C reading caused finder errors so as I've said before, I don't believe the temperatures reported in the B&W G3s or with most copper CPUs is with 50% of the actual reading. My guess is the above reading was accurate, and note the CPU is rated for 65C. (Powerbook G3 CPUs are rated at 85C and run 80C at least here on my PB G3/250).

The PowerJolt control does have an option to disable motherboard cache, but I did not need to test this (my PCPro cache was removed). Note that this may not fix a case where your motherboard cache is too slow to allow reliable booting, as the extension must load to disable the cache.

Note: As I've mentioned many times in the past in the front page news, I don't believe the CPU junction temperatures are accurate (literally), especially with the new Copper CPUs (IBM 300Mhz and higher are all copper based now - get our exclusive PPC Checker to test yours). In some cases I've seen 400+ MHz copper G3s reported as running at 8 degrees C - which cannot be accurate as that is below the ambient room temperature. Since the software just reads a register on the CPU unless a company does some tests and applies a 'calibration factor' to the reading the temperatures are of little use in my opinion. What is baffling is that the same CPU, running at the same speed will report different temperatures in different Macs. Perhaps supply voltage has something to do with the variance.

Update: As reported in the early June front page news, a 7300 owner with an overclocked 366MHz ZIF reported that although it ran fine at 400/200/50 speeds (overclocked), he saw SMI file checksum errors and had to back off to 375/187.5 (50Mhz bus). I suggest all owners that are overclocking their CPU verify that Apple's SMI files mount without checksum errors. I cannot recommend overclocking to anyone, but if you do (cache or CPU) always verify that you do not have data errors. Mounting an Apple SMI file is one way to test for disk errors (as noted in news recently from B&W G3s with IDE addon drives).

The PowerJolt software control has some unusual features like the motherboard cache disable feature but was not accurate in the PowerCenter Pro, has a cluttered layout and is not needed in Apple G3s. If you're getting this for an Apple G3 you can skip the software install. For these reasons I rated software controls a 6.

The next page describes the documentation and installation instructions supplied with the card. Or you may use the links below to jump to a specific page.

Index of Railgun G3 466/233/1MB Review Pages
Intro | Benchmarks | Appl. Tests | Software Controls | Docs/Installation | Specs/Design | Summary

- or -
Back to WWW.XLR8YOURMAC.COM

Copyright © Mike, 1999.

No part of this sites content or images are to be reproduced or distributed in any form without written permission.
All brand or product names mentioned here are properties of their respective companies.
Users of the web site must read and are bound by the terms and conditions of use.

The Source for Mac Performance News and Reviews
Review: BottomLine's Railgun G3/466/233/1MB ZIF Published: 6/28/99
Software Cache Control Features
Intro \| Benchmarks \| Appl. Tests \| Software Controls \| Docs/Installation \| Specs/Design \| Summary
Cache Control Software
NOTE: Bottom Line and Mactell are now out of the CPU upgrade business. The last source for Manuals and software for Railguns is listed in the CPU Upgrades section of the FAQ. For cache control software, Powerlogix.com's free cache profiler usually works also (but disable any existing Railgun extensions/controls before switching to a different cache enabler.) Bottom Line's Railgun comes with Mactell's new PowerJolt cache control utility and an extension for cache enabling. However in the Apple Beige and B&W G3 the Railgun ZIF, like the XLR8 and Vimage ZIFs, does not need any software to enable the backside cache. For tests in the Apple G3s, I did not use the PowerJolt software, as I prefer run lean and mean (without any unnecessary software when possible). I did use the PowerJolt software when running the Railgun ZIF in the PowerCenter Pro with the CarrierZIF card since software was required to enable the cache in pre-G3 Macs. I also noted as shown below that the PowerJolt control panel did not accurately display the bus speed in the PowerCenter Pro. This was verified by checking speeds with XLR8's and Powerlogix's utilities. Both agreed with the card's actual speed, and simply dividing the CPU speed by 8X ratio (max) proves the Powerjolt reported bus speed was not correct. Not a big deal, but along with the cluttered interface, it was another reason I prefer to use XLR8's cache control (which works fine with the Railgun). From what I have seen in real world applications tests, backside cache speed is not a primary factor in performance so don't push the bleeding edge on cache speed as it can affect reliability. Installing the Software: The XLR8 software installer automatically places the cache control and extension in the system folder. After a reboot, it tests the backside cache automatically and reports what it suggests is the maximum reliable cache speed (but offering you the option to manually set the speed to another setting). Often this speed may not be reliable in my experience, as it reported cache speeds of up to 296MHz were OK on the review ZIF, which I found later to be too fast. Only the expensive (and rare now) IBM cache can run reliably near 300MHz speeds in my experience (Some reviews like the PF 466/311 ZIF and older cards like the XLR8 G3/266 show pictures of this cache if you're interested.). I manually set the cache to a 2:1 ratio (half the CPU speed), which ran perfectly at speeds of up to 225Mhz. PowerJolt Cache Control (PowerCenter Pro speeds shown) Note the 67C temperature using the supplied heatsink, which was the highest temperature I've ever seen from a G3 desktop CPU, yet there no errors of any kind during the PowerCenter Pro tests. The stock heatsink would not let the PowerCenter Pro's fan/bracket fit, so I'd suggest using the CarrierZIF's stock plate heatsink which does allow using the fan/bracket and should dramatically help cool the CPU. In the B&W G3, at 550MHz a 39C reading caused finder errors so as I've said before, I don't believe the temperatures reported in the B&W G3s or with most copper CPUs is with 50% of the actual reading. My guess is the above reading was accurate, and note the CPU is rated for 65C. (Powerbook G3 CPUs are rated at 85C and run 80C at least here on my PB G3/250). The PowerJolt control does have an option to disable motherboard cache, but I did not need to test this (my PCPro cache was removed). Note that this may not fix a case where your motherboard cache is too slow to allow reliable booting, as the extension must load to disable the cache. Note: As I've mentioned many times in the past in the front page news, I don't believe the CPU junction temperatures are accurate (literally), especially with the new Copper CPUs (IBM 300Mhz and higher are all copper based now - get our exclusive PPC Checker to test yours). In some cases I've seen 400+ MHz copper G3s reported as running at 8 degrees C - which cannot be accurate as that is below the ambient room temperature. Since the software just reads a register on the CPU unless a company does some tests and applies a 'calibration factor' to the reading the temperatures are of little use in my opinion. What is baffling is that the same CPU, running at the same speed will report different temperatures in different Macs. Perhaps supply voltage has something to do with the variance. Update: As reported in the early June front page news, a 7300 owner with an overclocked 366MHz ZIF reported that although it ran fine at 400/200/50 speeds (overclocked), he saw SMI file checksum errors and had to back off to 375/187.5 (50Mhz bus). I suggest all owners that are overclocking their CPU verify that Apple's SMI files mount without checksum errors. I cannot recommend overclocking to anyone, but if you do (cache or CPU) always verify that you do not have data errors. Mounting an Apple SMI file is one way to test for disk errors (as noted in news recently from B&W G3s with IDE addon drives).
The PowerJolt software control has some unusual features like the motherboard cache disable feature but was not accurate in the PowerCenter Pro, has a cluttered layout and is not needed in Apple G3s. If you're getting this for an Apple G3 you can skip the software install. For these reasons I rated software controls a 6.
The next page describes the documentation and installation instructions supplied with the card. Or you may use the links below to jump to a specific page.

Index of Railgun G3 466/233/1MB Review Pages Intro \| Benchmarks \| Appl. Tests \| Software Controls \| Docs/Installation \| Specs/Design \| Summary - or - Back to WWW.XLR8YOURMAC.COM
Copyright © Mike, 1999. No part of this sites content or images are to be reproduced or distributed in any form without written permission. All brand or product names mentioned here are properties of their respective companies. Users of the web site must read and are bound by the terms and conditions of use.