DFI nForce4: SLI and Ultra for Mad Overclockers
by Wesley Fink on February 5, 2005 9:30 AM EST- Posted in
- Motherboards
Memory Stress Testing
Optimum tRAS
In past reviews, memory bandwidth tests established that a tRAS of 10 was optimal for the nForce3 chipset and a tRAS setting of 11 or 12 was generally best for nForce2. Since this was the first memory stress test of a production nForce4 board, tRAS timings were first tested with memtest86, a free diagnostic program with its own boot OS that will boot from either a floppy disk or optical disk. Bandwidth was measured from tRAS 5 to tRAS 11 to determine the best setting.
| Memtest86 Bandwidth DFI nForce4 with Athlon64 4000+ |
|
| 5 tRAS | 2191 |
| 6 tRAS | 2242 |
| 7 tRAS | 2242 |
| 8 tRAS | 2242 |
| 9 tRAS | 2141 |
| 10 tRAS | 2141 |
| 11 tRAS | 2092 |
The best bandwidth was achieved in the 6 to 8 range with this combination of nForce4 and the 4000+, so a mid-value tRAS of 7 was chosen for all tests.
Memory Stress Tests
Our memory stress test measures the ability of the DFI to operate at its officially supported memory frequency (400MHz DDR), at the lowest memory timings that OCZ PC3200 Platinum Rev. 2 modules will support. All DIMMs used for stress testing were 512MB double-sided (or double-bank) memory. To make sure that memory performed properly in Dual-Channel mode, memory was only tested using either one dual-channel (2 DIMMs) or 2 dual-channels (4 DIMMs).
| Stable DDR400 Timings - One Dual-Channel (2/4 DIMMs populated) |
|
| Clock Speed: | 200MHz |
| CAS Latency: | 2 |
| RAS to CAS Delay: | 2T |
| RAS Precharge: | 7T |
| Precharge Delay: | 2T |
| Command Rate: | 1T |
Using two DIMMs in Dual-Channel 128-bit mode, the memory performed in all benchmarks at the fastest 2-2-2-7 timings, at default 2.6V voltage.
Tests with 4 DS DIMMs on an AMD Athlon 64 system are more demanding, since AMD specifies DDR333 for this combination. However, most AMD Athlon 64 motherboards combined with recent AMD processors (the memory controller is on the AMD CPU) have been able to handle 4 DIMMs at DDR400.
| Stable DDR400 Timings - 4 DIMMs (4/4 DIMMs populated) |
|
| Clock Speed: | 200MHz |
| CAS Latency: | 2.0 |
| RAS to CAS Delay: | 2T |
| RAS Precharge: | 7T |
| Precharge Delay: | 2T |
| Command Rate: | 2T |
Tests with all four DIMM slots populated on the DFI nForce4 boards required a 2T Command Rate with 4 DS DIMMs in two dual channels. This is the pattern seen on other top-performing Socket 939 boards, but we hoped that higher voltage might allow us to eek out 4 DS DIMM 1T performance. However, additional voltage did not help and DDR400 with 4DS DIMMs still required a 2T Command Rate on the DFI nForce4 boards.
Facebook
Twitter
RSS
114 Comments
View All Comments
Avalon - Saturday, February 5, 2005 - link
If you don't like the exclusive, wait for the NF4 roundup. It's that simple. Just because you feel enough data wasn't included in an ARTICLE, doesn't mean that Anandtech does not have such data in their LABS waiting to be released. Again, wait for the roundup.overclockingoodness - Saturday, February 5, 2005 - link
Justimann: Like others pointed out, Wesley is one of the most respected journalists in this industry. And AnandTech is one of the best sites as well, so I doubt they will be putting their integrity on line just to get a damn exclusive. AnandTech is just like any business but I am proud to say that I have hardly seen any bias on this site (the only bias may be disagreement between my interpretion of the reviwer's words).AnandTech would never recommend anything unless they are confident about it and even if the data is not on the graph, we trust AnandTech's editors.
DEMO24 - Saturday, February 5, 2005 - link
I dont remeber if people have asked this yet(having a feeling they have, but I dont want to read 5 pages of replys;)) but do you guys at Anandtech still have some Winchestor CPUs lieing around? Would be nice to see how far you could get one :)bersl2 - Saturday, February 5, 2005 - link
#79, #80: It's got nothing to do with any of that. SLI nForce4 chips have a price premium on them. Every Ultra sold instead of an SLI means less money for nVidia.joe4324 - Saturday, February 5, 2005 - link
Its more likely that nvidia doesnt want to deal with supporting hardware running out of spec AND beyond its intentions, There not stupid, they know how people are. Even if this board preformed perfectly in all ways, UNTILL you modded it. then it caught fire. They would be slamed for being unreliable and unsafe, and thus not worth buying. even though it was the best mobo out at the time, when NOT modded.rjm55 - Saturday, February 5, 2005 - link
I don't get it. Why did nVidia do this? If anybody did a mod to SLI they had to buy two nVidia video cards to use SLI. So nVidia got more profit. Looks like they couldn't lose with SLI only working with nV cards.Guess nVidia has now totally forgotten their gaming and enthusiast roots and they're just another big company throwing around their weight. Maybe ATI will do better.
Wesley Fink - Saturday, February 5, 2005 - link
The following Update has just been added to the front page of the review:"UPDATE 2/05/2005: nVidia has acted to prevent, or at least make it more difficult, to mod the Ultra board to SLI. First, DFI has advised us, and posted on their website, that they will NOT sell the SLI bridge to buyers of the Ultra board. Second, nVidia has advised us that future shipments of the Ultra chipset have been modified so that the mod to SLI will no longer be possible. An additional side effect of this second action is that the "Dual Video" mode, which performs at about 90% of SLI performance levels, will only work with early nVidia drivers 66.75 or earlier. If you do a quick check of web driver postings you will see it is now very difficult to find 66.75 drivers. With a chipset modded to SLI the "Dual Video" mode worked through 70.xx versions of the nVidia driver. nVidia also made it clear they will continue to make driver changes to prevent operation of any "non-standard" (8X/8X) operation of their SLI driver. This also throws into question whether the VIA "dual graphics" mode on the 894 Pro chipset will ever work with nVidia graphics cards. If you are interested in the current UT Ultra-D we suggest you buy one now if you can find it. Future versions of the UT Ultra-D will not have the same capabilities as a result of these actions."
ImJacksAmygdala - Friday, February 4, 2005 - link
I guess beauty is hard to quantify. Like Wesley said he has saved the Editor's Choice Award for the SLI board review that will provide more hard data to support his opinion on the DFI board.justinmann - Friday, February 4, 2005 - link
Dear Gerbil, I am not wasting my time. I read the article, and like any good scientist I looked at the data. The data do not support the conclusions, therefore I felt it necessary to ask why he did not perform an experiment that backed up his proclamation. It's really a pretty simple concept.Justin
Happy Buddha - Friday, February 4, 2005 - link
Hey Wesley,You did use a IDE HDD for benchmarking so the question about SATA and overclocking pop right to my mind: Is Overclocking as stable with SATA drives?!?
I know that the NF3 250GB was, but sometime engineer can make silly mistake. ;)
An can you or DFI confirm that the low end (Utra-D) will have as much success with OC? Thanks...