AMD Motherboards: Processor Support
by Wesley Fink on June 28, 2005 1:00 PM EST- Posted in
- Motherboards
The Processor Factor
Another concern has raised its head in the last few months that should be a consideration for anyone shopping for an Athlon 64 motherboard - the processor factor. When we tested the nForce4 SLI boards we used the 4000+ clawhammer chip as our standard CPU. We also had done some testing with the early Winchester chips which were based on the 90nm production process instead of the 130nm process used for clawhammer. Overall these early 90nm chips were mainly a die-shrink, and performance - and compatability - were much the same whether clawhammer, winchester, or newcastle.Recently, however, we have some new choices from AMD in Revision E chips and dual-core. These new Revision E parts support SSE3, are based on the 90nm process, and they do not always behave as earlier chips did in the same motherboard. We saw this for ourselves in our Gold Editor's Choice MSI K8N Neo4/SLI. While we experienced outstanding overclocking with a clawhammer chip, users with Venice chips were experiencing perfectly miserable overclocking results. It has taken MSI some time to find a solution to this problem, but we are happy to report that a new BIOS has just been released that claims to fix the Venice issues on the MSI. The point of this is that the newest Athlon64 Revision E chips, code-named Venice and San Diego, do behave differently than earlier Athlon 64 chips in some boards. This is likely a temporary concern as the market adjusts to the newest CPU architectures, but it is a factor that should be considered.
This becomes an even larger issue with the new Toledo dual-core processors. The Athlon 64 X2 joins two Venice or San Diego cores on a single CPU. These Revision E X2 dual-core CPUs - the 4200+, 4400+. 4600+, and 4800+ - theoretically will work in any Socket 939 board. However, you will certainly need at least a BIOS upgrade. Most of the major manufacturers have quickly brought the needed BIOS upgrades to market, but if you plan to run a dual-core chip you need to check before you buy.
It should also be mentioned that AMD has implemented a hidden feature in Revision E processors, namely additional memory ratios that can be implemented in BIOS. We saw a Revision E chip mounted on the Abit Fatal1ty AN8 motherboard. Adding the new processor gave the additional options of 433, 466, and 500 to the available memory ratios. This has to be coded in the BIOS to be available, but the new asynchronous ratios are a feature of the Revision E Memory Controller.
The good news, unlike the Intel dual-core, is that just about any Socket 939 motherboard can run dual-core. But you will definitely need a BIOS update and you will definitely need to check to make sure the board you are looking at has an available BIOS supporting dual-core.
Facebook
Twitter
RSS
57 Comments
View All Comments
PrinceGaz - Wednesday, June 29, 2005 - link
#32- those TWINX memory products are matched pairs of 1GB modules, which gives a total of 2GB. They are not individual 2GB modules. AFAIK all of the 2GB modules available are registered ECC such as Crucial's 2GB PC3200 module CT25672Y40B. That makes them unsuitable for A64's, but fine for Opterons (which is what you would expect as modules of that size are only likely to be used in servers).I'm almost certain that A64 Rev.E DOES support four DIMMs at 1T command rate, but ONLY if single-sided. Being able to use 1T command-rate is dependent mainly on the load placed on the memory bus, ie the number of memory chips on the channel. Two double-sided modules per channel is 32 chips which is never going to run at 1T.
#27- all 1GB modules are double-sided, it's highly unlikely that your freind with four 1GB modules is running at 1T, or if he is that they are running error-free. I suggest he lets Memtest86 loop through all its tests overnight as it will probably report errors (test 5 in particular).
smn198 - Wednesday, June 29, 2005 - link
#28 Wesley - I also would love to see the performance of dual core with the new async. memory speeds. I imagine that dual core would benefit dual core more than the FX57 but I'm not sure how bandwidth starved a dual channel X2 is."AMD has also added additional "hidden" features in the AMD on-processor memory controller. Additional asynchronous ratios are available at 433, 466, and 500 memory speed on boards that implement the necessary code to access these memory controller features. These options should be available with any Revision E chip if the manufacturer implements the controller option read in BIOS."
sprockkets - Wednesday, June 29, 2005 - link
anyone want to make a wiki for processor support here for AMD stuff?justly - Wednesday, June 29, 2005 - link
"If we overlooked boards and/or BIOS revisions that belong in our Supported list please let us know."The Asrock K8Upgrade and Combo-Z also have X2 BIOS support.
Slaimus - Wednesday, June 29, 2005 - link
What about single sided sticks? I have two single sided 512MB sticks in my current system specifically to allow two more to be added.jiulemoigt - Tuesday, June 28, 2005 - link
Your not actully saying the same thing 4 gigs ram != 4 sticks of ram, there are 2gb sticks just not many:TWINX2048-3200PRO 2048MB 3-3-3-8 2x184
TWINX2048-3200C2 2048MB 2-3-3-6* 2x184
TWINX2048-3200C2PT 2048MB 2-3-3-6* 2x184
TWINX2048-3200 2048MB 3-3-3-8 2x184
TWINX2048-3200PT 2048MB 3-3-3-8 2x184
gettting 1T commands out of any of those sticks on the other hand may simply not be possible, the problem was not the size of the memory but shared latency, with four sticks you have four over lapping fields that have to line up really nice. With 2gb sticks you only have two fields :)
cryptonomicon - Tuesday, June 28, 2005 - link
"Unfortunately the long-rumored 1T Command Rate with Rev. E AMD processors appears to have been just a rumor. We could not run 4 matched dimms at 1T in a motherboard with Rev. E/x2 support. Four dimms still required a 2T Command Rate."Doesn't surprise me... thats the price you pay for the amd platform. However its pretty much the only bad part.
Wesley Fink - Tuesday, June 28, 2005 - link
#27 - People who know memory also tell me no 2T with 4 dimms, so I am skeptical of your friend's claim. Sandra appears to report 1T when it is not really reporting Command Rate which confuses many. I know Oskar's BIOS' for every memory - I believe last count was more than 55 BIOS revisions for the DFI nF4. That's not the issue. The 4 dimms 1T would be a function of the mem controller on the new Rev E and not have a lot to do with the board. The new Rev E mem controller is indeed a better overclocker than the earlier 0n-CPU controllers.KeithDust2000 - Tuesday, June 28, 2005 - link
Wesley, thanks a lot for the informative and prompt reply, I can´t wait to see the results!:)
Wesley Fink - Tuesday, June 28, 2005 - link
KeithDust2000 - Wish I had an answer to that. I'm still waiting for a FX57 and/or 4800+ from AMD for our motherboard reviews. When supplies ease a bit and I get the CPUs I'll include some asynch benchmarks in a future review. The difference will not be dramatic, but with DDR500 available at pretty fast timings these days, you will get a bit of a boost from a 400 CPU running at stock and driving memory at DDR500 or 533.Only a few boards implement the new memory options correctly, and that may be why a lot of reviewers overlooked them. AMD didn't advertise the new ratios in their press kit and most weren't looking for them.