As mentioned on the first page of the review, I received one of the first cards off the production line, with no documentation included (to get this review to you as soon as possible). Therefore I can't rate documentation. I could try to do so based on past PowerForce manuals, but without having one specifically for this card I can't rate how accurate the speeds settings listed are or if there were any errors present in the installation guide. Therefore I cannot return a score on documentation at this time. PowerLogix will send a printed manual soon and I'll update this page when it arrives.

Installation was of course identical to other CPU cards. Again without a printed manual I can't say how clear or accurate the instructions are. I'm familiar with the procedure and it was a snap. As with all adjustable speed cards, most of the initial effort will be adjusting the speed of the card and finding the maximum reliable speed. If you're not the tweaker type - just set it to a safe speed (at or near the stock speed). If you want to get the maximum performance out of the card, then additional testing and trial and error will be required.

Although I rarely follow this advice, it's always best to back up your data before boldly going where no man had gone before.

Speed Adjustments:
The CPU speed of the card is set by two small rotary hexadecimal (0-F) switches on the back of the card, mounted at a right angle so that they are accessible when the card is installed in the system. You can leave the card set to its default CPU speed of 220 Mhz or you can try and run faster speeds by adjusting the two switches to the desired speed according to the table in the manual. Of course you adjust the switches with the power off.

Cache speeds have proven to be much less a factor than CPU speeds in overall performance, a good thing since once I exceeded about 280 Mhz CPU speeds I had no choice but to use the 2.5:1 cache ratio for reliable operation. This is the reason for the low (116 Mhz) cache speed.

I don't think the combinations I used were necessarily the best ones that are possible but I was very satisfied with the results. I ran the card continuously for almost two days to prove that the speeds noted here were reliable in this system. There might be a different combination of bus/cache/cpu speed that would deliver better performance, but I did not want to include speeds in the review that I have not proven reliable. Since I spend at least 24 hours of burn-in testing to verify reliablility, doing this for all settings was not possible in the review timeframe.

After the XLR8 G3/266 test, I was surprised that the card ran fine with the PowerTower Pro motherboard cache installed. Although they are based on the same design, it just goes to show that no two cards are alike.

Regardless of the relatively low cache speeds, the performance is pretty amazing for the money.

Verify Reliability!
When running the card above the rated speed, always verify reliable operation at the new speed by running with CPU intensive applications for many hours. If there are any errors or failures, reduce the speed setting to a lower one as shown in the table. When errors are seen I like to reduce speeds by at least two clicks backwards to ensure I'm not running the card on the "ragged edge".

Use common sense, always start low and work up, not the other way around, which can be a disaster leading to frustration at best, and corrupted files at worst. It is foolish to have the fastest machine around if it crashes every other hour. Speed without reliability is nothing, but you don't have to sacrifice one for the other if you are careful, use quality components and spend a little time up-front in setup and testing.

I tried really hard to prove that 291/116 speeds were not reliable. After almost an entire day of testing in benchmarks and applications, I ran Newer's RAMometer continiously for 24 straight hours with the cover on, and saw zero errors.

The PowerTower Pro, like most PowerComputing models, does have a fan blowing on the CPU heatsink at all times. That could be a factor. I'm going to run it overnight tonight without the fan running just to see if that did make a difference. Do you get those types of tests anywhere else?

Below is a photo of a PowerForce CPU card installed in the PowerTower Pro. Although it's a close fit with the PTP CPU bracket/fan assembly, I could still adjust the switches with the card installed.

Speed Adjustment Switches (PowerTower Pro installation shown)

Since I did not have any printed documentation to judge, I can't rate documentation at this time. When PowerLogix sends a printed manual I'll update this page with a score.

Installation was no more difficult than any CPU card, but if you want to explore accelerated speeds, count on spending some time up front to verify that settings you choose are really reliable.

The next page describes the hardware features and specifications of the card. Or you may use the links below to jump to a specific page.