Asus P4C800-E: Stress Testing

We performed stress tests on the P4C800-E in these areas and configurations:

1. Chipset and motherboard stress testing, which was conducted by running the FSB at 288 MHz with OCZ 3700 GOLD at 5:4 memory ratio (DDR461) with 2 x 256MB double-bank dimms at 2-3-4-6 timings.
2. Memory stress testing, which was conducted by running Corsair 3200LL RAM at 400 MHz with all 4 DIMM slots filled. Two pairs of Corsair TwinX ver.1.2 was used for this test at the lowest memory timings possible. We tested the Corsair because it was one of the memory brands that was not completely stable in the earlier P4C800 Deluxe.

Front Side Bus Stress Test Results:

We ran a full range of stress tests and benchmarks to ensure that the Asus P4C800-E was absolutely stable at each overclocked FSB speed. These stress tests included Prime95 torture tests, which were run in the background for a total of 24 hours.

In addition, we ran several other tasks — data compression, various DX8 and DX9 games, and apps like Word and Excel — while Prime95 was running in the background. Finally, we ran our benchmark suite, which includes ZD Winstone suite, Unreal Tournament 2003, SPECViewperf 7.0, and Gun Metal Benchmark 2.

In addition to tests with 2 DS dimms, we ran tests with 4 DS OCZ 3700 GOLD (4X256) at 288 FSB. With 4 Dimms, we needed to reduce the FSB to 284 to achieve stability at default voltage. In addition, we discovered that 4 double-bank dimms apparently turn-off PAT above 200FSB. This was confirmed by both CPUZ 1.18C and Sandra tests results. At 800FSB (200) 4DS dimms tested faster than 2 DS dimms, while at 1136FSB (284) 4 DS dimms tested slower than 2 DS dimms. CPUZ 1.18C shows Performance Mode “enabled” with 2 DS dimms and with 4 DS dimms at 800FSB. With 4 DS dimms at 1136FSB (284), CPUZ shows Performance Mode “disabled”. 4 SS dimms show Performance Mode “enabled” at all speeds, and Memory Tests confirm PAT is working.

Asus has been contacted for a fix or comments on the high-speed 4 DS dimm issue. We are also investigating whether this is an issue with all 875 boards, or whether it is unique to the P4C800-E. Running 4 DS dimms at 284 FSB is something most users of the P4C800-E will never do, so it is not a huge issue. We will post additional information on this issue here as we have it.

Memory Stress Test Results:

The memory stress test is very simple, as it tests the ability of the P4C800-E to operate at its officially supported memory frequency (400MHz DDR) at the lowest supported memory timings that our Corsair TwinX LL ver. 1.2 can achieve:


Stable Dual DDR400 Timings
(4/4 banks/2 dual-channel populated)
Clock Speed: 200MHz
Timing Mode: Auto
CAS Latency: 2.0
Bank Interleave: N/A
RAS to CAS Delay: 3T
RAS Precharge: 5T
Precharge Delay: 2T
Command Rate: 1T


There have been many reports of problems with Corsair 3200LL on the P4C800 Deluxe, though later BIOS revisions appear to work much better with Corsair memory. We had no problems at all running 4 dimms of 3200LL at 2-2-3-5, which is a bit faster than the Corsair SPD 2-2-3-6 timings.

Since we just used the Asus P4C800-E as our test board for “Searching for the Memory Holy Grail — Part 2”, we were also able to test this Asus with Adata PC4000, Adata PC3500, Corsair XMS4000, Geil Platinum 4000, Kingston HyperX4000, OCZ 4000, and OCZ 3700 GOLD. We had no problem at all with any of the memory tested. The P4C800-E was completely compatible with all the test dimms at both SPD and more aggressive memory timings and voltages.

Asus P4C800-E: BIOS and Overclocking Performance Test Configuration
Comments Locked

30 Comments

View All Comments

  • Anonymous User - Sunday, August 31, 2003 - link

    So, then running 4 DIMMs of Corsair XMS LL Platinum 512 MB would be slower than just 2 DIMMs of that memory? Is it just PAT that is affected? I've heard that this board will slow the latency timings and drop out of dual-channel mode with 4 DIMMS. BTW, I heard this from a rich friend who bought a system from Falcon Northwest and asked them about using 4 DIMMs on this exact board. They said that it would slow the latency timings and drop out of dual-channel. Was this experienced during testing? Is it that Falcon Northwest doesn't know what they're talking about? Now I am very confused about this particular board.
  • Wesley Fink - Sunday, August 31, 2003 - link

    Asus SAYS the P4C800-E is compatible with Prescott, but without a Prescott chip to test with we can not give a definitive answer.

    We did in fact run several benchmarks comparing the old and new when we began the memory testing for our DDR500 review, but found no real difference in performance between the old and new with both running 1.010 BIOS. That is why we did not publish performance tests comparing the two, and we stated there was no performance difference in our update. We did find memory compatibility better on the "E" and also reported that as well.
  • DrMrLordX - Sunday, August 31, 2003 - link

    I second the request for information about this board's compatibility with Prescott.
  • Anonymous User - Sunday, August 31, 2003 - link

    Jesus Christ, how can Anandtech readers be so stupid. Both the old and new ASUS revision are based on the EXACT SAME CHIPSET! It would be ridiculous to compare them, they both use the 875P chipset, the only difference is a couple onboard features that make the newer revision a better overall board and value.

    Christ, READ the reviews for god sakes.
  • PrinceGaz - Sunday, August 31, 2003 - link

    I feel #3 had a very good point. Since it is a review which begins by comparing the old revision with the new one, not including both in the benchmarks was a mistake (assuming the figures are available).

    @#4: Telling everyone to go and compare the results from an older review is crazy; for the time it would take for them to be included, asking everyone who reads the review to go and read another simultaneously is stupid.
  • Anonymous User - Sunday, August 31, 2003 - link

    #4
    Catch a clue. Go read the old review and compare them.
  • Anonymous User - Sunday, August 31, 2003 - link

    So, why the heck did you not bench the old against the new. That is after all what all the bloody writing ia about. That should include a lan test. Get on it, do it. What a waste, waste, waste. Oh yea....what kind of memory can you use with the old revision?

    Useless information is everywhere.
  • Anonymous User - Sunday, August 31, 2003 - link

    Outstanding review! What I'm missing though is information about the support for the Intel Prescott processor. Looking on the ASUS website and in the manual, it is stated that it is compatible but such claims have been made for other motherboards only to be withdrawn later. Is it compatible with FMB 1.5 and VIN 1.5 required for the Prescott?
    I'm also quite surprized that PAT is disabled if you have 4 DS dimms running at FSB bus speeds higher than 200Mhz. After reading "Searching for the Memory Holy Grail: Part 1", I had planned to go for 4 256MB Cosair TwinMX dimms instead of 2 512MB ones. I'm eagerly awaiting the answer you get from ASUS.
    Cheers,
    Manuel
  • Wesley Fink - Sunday, August 31, 2003 - link

    Since this was an update to an earlier review, we did not cover features that would be discussed in a full review. However, I have tested AI overclocking on the P4C800-E and find it works very well.

    You set the desired overclock, and the boards adjusts all parameters for the most stable overclcoking performance. Overclocks up to 30% (around 1066 FSB) are possible using AI.
  • Anonymous User - Sunday, August 31, 2003 - link

    Would Like to know if you tested the much vaunted "Artificial Intelligence" overclocking ability of this board. This is something that mere mortals are very likely to use, rather than all the tinkering with FSB speeds and memory timing.

Log in

Don't have an account? Sign up now