Enhanced Performance Profiles, or EPP, allows consumers to easily expose new, advanced performance memory settings built into high performance memory DIMMs for even higher levels of overall PC system performance. If you are using DDR2 memory or plan to on AMD's upcoming AM2 platform then read on!
Since we are talking about technology on a platform that has yet to be released I find it hard to give my full thoughts on EPP, but it is interesting to say the least. Many reviewers in the industry see SLI-Memory with EPP as a way to "hand-hold" consumers, but I don't really see it like that as you can disable the feature if you don't want it. Right now I think NVIDIA and the memory companies are headed in the right direction. They are taking unused lines of space in the SPD and putting them to good use...
JEDEC Chair Says Enhanced Performance Profiles are News to Him!
This week an industry announcement was made for a “New Open Standard
Memory Specification” called “Enhanced Performance Profiles” or EPP. This
proposal suggests that it can “allow memory manufacturers to integrate
additional module performance data” by redefining the manufacturer’s data
region, bytes 99-127.
As chairman of the JEDEC SPD Task Group, I feel it is my responsibility to
inform the committee that this EPP specification was not developed in my
task group. In fact, we have not seen a first showing on the topic.
EPP is not a JEDEC endorsed specification and should not be misconstrued
in the industry as such. I do encourage the sponsors of the concept, or
any other companies wishing to properly document higher DDR2 speed grades,
to develop a truly open specification by submitting a proposal to the
JEDEC committee and to the SPD Task Group and following the well
established guidelines by which open standards are developed.
from an email from the Chairman of the JEDEC SPD Task Group
I'm very surprised that NVIDIA or Corsair did not even bring the proposal to JEDEC yet! That just seems like a necessary step in developing a standard in the memory industry and something that should be completed (or at least proposed) prior to the public launch of the technology.
Why would they need submit this to JEDEC? EPP is only specific to Nvidia chipsets.....What NV done was ask memory makers to add a bit more code into their SPDs so they can both shift more products....To me this isn't a new memory standard. The SPD is still the same, non-nvidia chipsets will still see thier required bits......I always tought that JADEC's specs were not specific to 1 platform/vendor
While i understand JEDEC's position, you also should concider that its like congress. Nvidia and Corsair would have to wait months if not years for something like this to become standard through JEDEC. By going to market with it them selves it could become an adopted standard just like PC4000. The best thing to do with this is a wait and see approch. Wait to see if it actualy comes out, then see if it works. Then see if any other manufactuers like OCZ, Mushkin, GSkill, PDP.... start using EPP. Who knows how this will turn out, but so long as it is not incompatable with JEDEC SPD then im all for it. Im no ram expert and it would be nice if my memory had advanced setting profiles on it because that would reduce the effort on my end to get computers set up.
"Don't open that! It's an alien planet! Is there air? You don't know!"
While I'm not a JEDEC expert, but I think that if JEDEC adopts it.... then it becomes an 'open standard.' So after thinking on this a bit more... I'm thinking corporate greed has a role in the way NVIDIA decided to handle this new technology. If/when this gets approved by JEDEC, then I'm wondering if NVIDIA will be forced to allow all chipset makers to read the EPP bits in the SPD. But until then, like Ken mentioned, this will be NVIDIA chipset specific and Intel, SiS, and VIA will not be allowed to use this new technology and be forced to wait for JEDEC approval.
IF THIS IS THE CASE, then it makes perfect sense why NVIDIA chose to release the technology first. So all the consumers will choose to go with NVIDIA chipsets between now and JEDEC approval of EPP since NVIDIA based mobos will have slightly better performance. And like Yves pointed out, that will be a while!
i doubt they will ever submit this to JEDEC and, my money is on JEDEC never adopting it.
I do however think this could be come used in the market place. Also since its nothing but a serial bit patern, there is nothing stoping other companies from reverse enginering it, so while intel will most likly never use it, Via, SIS and ATI could.
"Don't open that! It's an alien planet! Is there air? You don't know!"
There is no need for ATI/SIS/VIA/Intel to reverse engineer the specification. NVIDIA published the specification on our Web site for anyone to use, as did Corsair. The spec is open for public consumption.
is there any chance this eprom is going to be programable? Because i can see it now... send your sticks to some uber OCer who tests the hell out of your memory and writes back a custome EPP, or hell even users might like it. I know because of some veriation between chips most ram manufactuers are going to use slightly relaxed profiles, ones that are more aggressive than JEDEC but not as aggressive as the chips may beable to handle in case they get a lower quality chip or two.
The more i read about this the more i like it. Yes EPP will come with a cost, my guess is the extended testing and programing will add about $10-30 to a pair of modules. But thats not too bad when you concider the added performance. What im really interested in, is the posibility that existing memory could be updated using this. If the EPROM is large enuogh on existing memory i dont see why we could not take it, program a new EPP and plug it into a new EPP enabled chipset.
"Don't open that! It's an alien planet! Is there air? You don't know!"
NVIDIABDR wrote:There is no need for ATI/SIS/VIA/Intel to reverse engineer the specification. NVIDIA published the specification on our Web site for anyone to use, as did Corsair. The spec is open for public consumption.
NVIDIABDR thanks for dropping in! Are you saying that EPP is free for anyone in the industry to use? Therefore this is not an Nvidia specific thing?
Apoptosis wrote:
infinitevalence wrote:my guess is the extended testing and programing will add about $10-30 to a pair of modules.
A couple memory companies I have talked to said that they will indeed be increasing the pricing on their EPP based modules.
I can't see how it should cost more! To my knowledge the hardware used to program an SPD can already see these bits used for EPP......the memory makers just don't use it.........AFAIK adding a very small bit of extra code would be ~10mins work...and would only have to be done once for each SKU.
Sounds like the memory makers are just using it as an excuse to raise prices......The testing/validation for each new profile added would cost them money.....but more people would buy their ram, thus lowering costs and increasing profits!
