I hate to be the one to do this, but:
There’s an Anandtech memory review article out there which says among other things that 4 DIMMs are better than two.
This is based on some unbuffered Sandra memory benchmarking, though the improvement doesn’t show using other benchmarks.
The article says the following:
“So why do we believe Sandra UNBuffered Memory benchmarks?
“The answer, surprisingly, is in Intel 875P Chipset Memory Configuration Guide White Paper and the Intel 865P Chipset Memory Configuration Guide White Paper. The tables below are taken from Page 13 of the 875P White Paper, but the tables are identical in both the 865 and 875 documents (our emphasis).”
The proper link to the 875P datasheet is here, and here’s what it is:.
So far so good, those are the results shown in the article.
Unfortunately for the author, here’s what the 865 datasheet says:
As you can see, it shows the opposite first-second place finish, so they’re obviously not identical.
Update 8/1/03: This item has been corrected. However, upon looking further upon the article, although an 865 and 875 board were claimed to be used, test results only come from an 875 board.
One needs to be careful about these things.
Making Lemonade From Lemons
Some rather inclined to memory fetishism have been saying recently that unbuffered Sandra scores are the real deal rather than the standard buffered Sandra memory scores.
Here is an ideal chance to test this theory out and redeem oneself.
If Anandtech is known for anything, it is known for running tons of benchmarks based on real-life applications.
Why not do that here? Run some benchmarks with these setups and see which does better. You need not run many benchmarks, actually you could just run Quake in a pinch, but the more, the merrier.
Actually, you could even kill two birds with one stone. Some gamers wonder what would be better for their system, 512Mb or 1Gb. One could test a few games at both 512MB and 1Gb and see if performance improves in particular games as a result.
That would make a very useful follow-up article.