Intro:

Speed without Compromise. That's what comes to mind when I think of the this card. Nothing else out there currently comes close to the range of settings or the ease of adjustment the PowerBoost Pro offers. This card will allow you to get the maximum performance out of your PowerMac with a minimum of work, period.

To fully appreciate the ease and range at which this card can be adjusted you would have had to have used all the other cards on the market. Most are a compromise, either providing 3 or 4 settings (usually two are actually usable), providing a lot of settings but making adjustments inconvenient and/or exhibiting cache sensitivity requiring a cache upgrade (even then many times not reaching the full potential). Not so with the PowerBoost Pro, which delivers on all counts - speed, adjustments, ease of use, and cache tolerance. Nothing else I've seen delivers that total package.

Not sensitive to cache: I was glad to see that the Pro was very tolerant of cache types. Using the stock Apple/Micron 256k cache it ran a 55mhz bus speed at the 4:1 ratio setting. Ditto with the standard Powerlogix 1 meg RapidCache! This card is not cache sensitive, so you should have plenty of range even with your stock cache. For best performance with a fast CPU like this, it is strongly recommended that you use a 1 meg cache. The PowerLogix cache in all models has proven superior to the IDT 1 meg cache, which seems limited to a 53mhz maximum bus speed in most machines. From a long series of tests I have determined that this particular 8500 is limited to 55mhz bus speeds regardless of cache. Both the Ultra and standard PowerLogix RapidCache can exceed this speed it seems. I am working now with sources to obtain a Power Computing machine that will allow much faster bus speeds to really see what this card can do. Many 8500's will be able to exceed my 55mhz limit, with the proper PowerLogix cache and this card.

Bus Speeds, Ratios and Cache explained:
The bus speed of your Mac determines how fast the the main Input/Output systems communicate with each other (ram/PCI/scsi), and is the main bottleneck in the current Macintosh motherboard designs. Think of it as a pipe size, limiting flow of information to other areas from the CPU. In general the faster the bus speed you can run, the better overall performance you will obtain. There is a fly in the ointment however, as the 604 lacks a much needed 4.5:1 CPU/Bus speed ratio setting. Why? Because in many cases the 4:1 ratio does not allow a high enough CPU speed with cards like a 233mhz. For example, we could run 55mhz bus speeds, but at 4:1 we only have a 220mhz cpu. Performance was better with the 52.67mhz bus speed at 5:1 which ran a 263mhz CPU.

Confused? Don't be, just be aware that if the 5:1 ratio allows a much higher CPU speed at the cost of a few mhz reduction in bus speed you will probably be better off with the higher CPU speed. You can run your own tests to prove this, and the Pro makes it very easy to do. Unlike other cards, the range of adjustment allows maximum efficiency.

To Summarize: If you can run a 50mhz+ bus speed at 5:1, it is better than a 56mhz bus at 4:1 with most cards. If 5:1 produces 20mhz or so more CPU speed at the cost of a few mhz in bus speed it is the better ratio to use. This is especially true if you can run 5:1 at 50mhz or above bus speeds.

Cache - why bigger is better
The bus speed bottleneck is why cache is so important, as cache buffers the CPU from this slower "pipe". Cache contains very fast specialized ram that allows the CPU to run at its full speed without waiting on slower devices like Ram and disks. The larger the cache the better able it is to keep a fast CPU "fed" so to speak. When data is not in the cache, the CPU must wait on slower ram or disk access to process data.

Finding the Max Speed - Easier to do than to describe:
I know it may sound confusing or complicated - this search for the maximum speed of your Mac. However with this card it is very simple. You start at a default setting (say 50mhz bus 4:1 ratio setting). That's a 200mhz cpu speed, and it is always best to start low and work up. Now you simply boot the Mac, and test for reliable operation. We followed a standard procedure of running the Newer Clockometer, Newer Ram test{run 5 times}, then a full MacBench 4.0 test series. If no errors were encountered, we raised the bus speed. If errors were reported, we reduced the bus speed one click and retested. Once the maximum speed is determined, then we run the Mac for at least 24 hours to determine if it is still reliable. Since shutting down and moving the remote control knob one click is all it takes per cycle (no removing the card to adjust speeds), this exercise is fast and painless.

I recommend you start with the Pro set for the 4:1 bus ratio setting (adjusted by this knob) during the search for the maximum bus speed. If your system or cache does not allow bus speeds of 57mhz or greater, then you may get better overall performance with the 5:1 ratio setting. Simply rotate the bus ratio knob to the 5:1 setting, then start your tests with the remote control set at 47mhz bus speed. Repeat the previously mentioned boot/test cycle, increasing bus speeds until you determine the maximum speed for you Mac. In our 8500, at 5:1 ratio were found that 263.35 was the maximum CPU speed we could reliably attain (52.67 bus speed times 5 ratio). If you have questions about this procedure, feel free to contact me.

Here's the routine:

1. Set processor card to 4:1 ratio using this knob
2. Insert card in Mac
3. Set remote control bus speed to 50mhz to start
4. Boot Mac
5. Runs OK? Increase bus speed knob one setting
6. Errors? Decrease bus speed knob one setting.
7. Goto step 4, verify reliability in normal use.

As with all PowerLogix PowerBoost cards, this product carries a 30 day money back guarantee and a 3 year limited warranty. From my experience you will not want to return this card - ever. They can have this card back when the pry it from my cold, dead 8500 (just kidding, PowerLogix!).

Base System Notes:

The base system used to test the PowerBoost^TM was our standard PowerMac 8500, with 128megs of matched memory (60ns), standard built-in video with 4 megs Vram, running OS 7.6. Disk cache was 512k, Speed doubler 2 and LibMoto were active (Speed Doublers faster disk access was disabled). Quicktime 2.5 and Quickdraw 3d extensions were also enabled (as done in prev. testing).

Since this card offered such ease and range of adjustments, we tested it with the following caches:

1. Stock Apple 256k cache (Micron chip based) (Image)
2. PowerLogix 1 meg RapidCache (p/n: RC1MB02)
3. PowerLogix 1 meg Ultra RapidCache (p/n: UC1MB)(Image)
4. IDT 1 meg cache (sold under many names) (Image)

All tests were run using MacBench 4.0 . All graphics tests were run at 1024x768, thousands of colors , not the lower resolutions and color settings used by most of the magazine reviews. We feel our settings are more in tune with actual users. Consider this when evaluating the video scores.
This was the most detailed test so far, as the ease in which settings could be changed, and the incredible range the Pro provides make searching for our system limit a pleasure in comparison to the competition. We tested all our available caches, and ran tests with 16 and 64 meg dimms, in interleaved and non-interleaved configurations. The Dimm size/interleaving made no difference in the performance scores in general (interleaving did help the Video scores a bit). The remote control was a godsend during this effort. For a complete Excel Spreadsheet showing all the test results and conclusions - download this Excel 4.0 Pro_test file ( Mac users - Hold down the mouse button and choose to save this link/file to disk, otherwise your browser will try to display it. Win95 browsers - right click on this link and choose save link ). (It is not compressed - as the WinDoze pc was used to keep the records during the Mac testing).

• Selectable Bus speeds from 40mhz to 70mhz in 0.3mhz increments.
• Remote Control for easy adjustment of bus speeds.
• CPU/Bus ratios of 3:1., 3.5:1, 4:1, 5:1, 6:1.
  (This card can drive the CPU to well over 300mhz, but our maximum reliable CPU speed was found to be at 263.35mhz.)