Searching for the Memory Holy Grail: Part 1
by Wesley Fink on July 27, 2003 11:13 PM EST- Posted in
- Memory
Mixed Configurations
In general, Intel’s White Papers assume that you will be using two or four matched DIMMs. The reality for many users, however, is that while they may have a pair of matched DIMMs, they do not likely have four matched DIMMs at their disposal to use in their new 865/875 boards. Intel’s White Papers address mixed memory configurations only to say that they will work, but they will default to the slowest speed and SPD timings of the mixed DIMMs.
To get a better idea of what happens to performance when mixing DIMMs at default performance at DDR400/800FSB, refer to the statistics below.
| DIMM Configuration | Best Memory Timing | UNBuffered Sandra 2003 Memory Test (MB/Second) |
% Changed from Matched DIMM Performance |
| 2x256MB DS + 2x512MB DS | 2-7-3-3 | 2094 INT 2148 FLT |
-25% |
| 2x256MB DS + 2x256MB SS | 2.5-7-3-3 | 2064 INT 2132 FLT |
-22% |
| 2x512MB DS + 2x256MB SS | 2.5-7-3-3 | 2097 INT 2150 FLT |
-23% |
| 4x256MB Matched DS | 2-7-3-3 | 2861 INT 2848 FLT |
---- |
The performance declines significantly from mixing different pairs of memory, even with two pairs of matched dual-channel. A 22% to 25% drop in memory performance compared to four matched double-sided DIMMs is certainly nothing to sneeze at.
Undoubtedly, the memory modules themselves are having an impact on this drop in performance. However, the four matched DIMMs run at 2-7-2-2, which is certainly not the best score I have seen at DDR400. In fact, the 2x512MB DS pair used in the mixed tests run as a dual-channel pair at DDR400 with 2-5-2-2 timings. The SS pair of 256MB DIMMs do require CAS 2.5 for best performance, and run fine as a pair at 2.5-6-3-2 at DDR400. Yet, these slower SS DIMMs perform better mixed with the DS GOLD modules than the faster DIMMs.
There will be great variation in mixed DIMM performance, with some close DIMMs running much better than these results. However, you can clearly see from the results that the very best performance at 1:1 requires four matched DIMMs. The performance loss from running two different matched pairs of DIMMs can be enormous.
Facebook
Twitter
RSS
42 Comments
View All Comments
jsalpha2 - Tuesday, July 29, 2003 - link
Pardon me, cause I'm tired. Did the article say if 4x(256) is better or worse than 2x(512). Assuming identicle brand and speed of RAM.I think I heard somewhere to go with just two sticks for better performance. Plus then you have open slots for later.
Question #2 Would 2x(512) of cheaper DDR333 be better than 2x(256) of DDR400?
thanks
Anonymous User - Tuesday, July 29, 2003 - link
Great article, it's just missing latency benchmarks.Anonymous User - Tuesday, July 29, 2003 - link
Ok - pardon the newbie question, but - I'm building a P4c with Asus P4P800 board. I want 1 gig of DDR400 ram - what brand/model number do I buy - ?Thanks for your help.
Anonymous User - Tuesday, July 29, 2003 - link
This is all nice and good, but what does it all mean in the real world, run some benchmarks in these various modes and show us whether we should care about it :) bottom line to me is what it does for the games, if i'm losing/gaining 4 FPS i'm more likely to care about the price differences then memtest.Wesley Fink - Tuesday, July 29, 2003 - link
These are quoted form Intel's White Paper, p.13 "NOTES: Ranks per Dimm (1 Rank is a single-sided DIMM, 2 Ranks is a double-sided DIMM)". The common practice of using higher-density Dimms every other Dimm on both sides (4 chips per side) is FUNCTIONALLY a Single Bank or Single-Sided Dimm.As for confirming that 4 dimms was faster, only the tests on the 3.0 were CPU-limited. We also determined maximum overclock on a 2.4C which was not CPU-Limited. Please check Page 7.
Philippine Mango - Thursday, January 25, 2007 - link
Wrong, you didn't use a 2.4C, you used a 2.6 processor which from what I know doesn't overclock as well as the 2.8C or 2.4C..Anonymous User - Tuesday, July 29, 2003 - link
"...we confirmed that the added memory bandwidth more than makes up for the slightly lower overclock with four double-sided DIMMs"To say you 'confirmed it' is quite a leap indeed... as you notably stated, you were CPU limited in going any higher for 1 and 2 sticks, whereas you clearly reached a blockade with the 4 sticks of memory. It could be that 4 sticks of memory causes a blockade in the chipset performance at some GHz, but with a better CPU you might have gone much higher with the opposing configurations.
-Robert
Anonymous User - Tuesday, July 29, 2003 - link
"If you plan to run DDR400 as your base memory speed with an 800FSB processor, your best memory performance will clearly be with four matched double-sided DIMMs"Can somebody help me to understand this?
I have only heard about 2 matched DIMMs...
Four matched DIMMs is 2 X 2 matched DIMMS?
Thank you very much!
Anonymous User - Tuesday, July 29, 2003 - link
While the article was interesting in that it at least confirmed Intel's white paper, I would be interested in your also testing ECC. I have a machine which does double duty as a backup server (plug the disks in the SCSI port and away it goes!). I am just curious as to the performamce hit when ECC is being used.Anonymous User - Tuesday, July 29, 2003 - link
The writer does not distinguish between DS and double bank module ;)