Memory Latency: No Integrated Memory Controller Necessary

One thing we pointed out in our earlier preview coverage of Intel's Core 2 Extreme is that the new processors have extremely low latency memory access, despite relying on the same memory controller as the previous generation of Intel CPUs.

Without an on-die memory controller Intel's Core 2 processor must use the memory controller in its chipset, which currently means the 128-bit DDR2 memory controller in either Intel's 965 or 975X chipset. The confusing thing is that although the Core 2 processors use the same memory controller as the old NetBurst processors, memory latency has been improved tremendously:

Memory Latency Comparison

Intel's Core 2 processors now offer even quicker memory access than AMD's Athlon 64 X2, without resorting to an on-die memory controller. While Intel will eventually add one, the fact of the matter is that it's simply not necessary for competitive memory performance today thanks to Intel's revamped architecture. Update:As many astute readers have pointed out, Core 2's prefetchers are able to work their magic with ScienceMark 2.0, which results in the significant memory latency advantage over AMD's Athlon 64 FX-62. This advantage will not always exist; where it doesn't, AMD will continue to have lower latency memory access and where it does, Intel can gain performance advantages similar to what ScienceMark 2.0 shows.

 CPU Everest
CPU-Z 1.35 (8192KB, 128-byte stride) CPU-Z 1.35 (8192KB, 64-byte stride)
AMD Athlon 64 FX-62 (2.8GHz) 45.9 ns 43.2 ns 19.3 ns
Intel Core 2 Extreme X6800 (2.93GHz) 59.8 ns 52.8 ns 10.9 ns

 

Because of the Core 2's intelligent prefetchers doing too good of a job with ScienceMark memory latency test, we wanted to also showcase situations where Core 2 would feature slower memory accesses than the AMD platform with its integrated memory controller. Everest's results are more in line with what we'd expect to see, with the FX-62 offering over 23% faster memory accesses than the X6800. CPU-Z's latency tool also reported somewhat similar findings, with an 18% performance advantage due to AMD's integrated memory controller. CPU-Z also provided us with numbers that showcase how well Core 2 can perform if its prefetchers are able to "guess" correctly; at lower strides the Core 2 Extreme manages faster memory access than the FX-62. The 128-byte stride numbers are indicative of what will happen if the pre-fetchers are not able to get the Core 2 the data it needs, when it needs it, while the 64-byte numbers show you what can happen when things go well.

L2 Cache: 4MB or 2MB? FSB Bottlenecks: Is 1333MHz Necessary?
Comments Locked

202 Comments

View All Comments

  • finbarqs - Friday, July 14, 2006 - link

    just to show you how EASY it is to O/C the system! (and for the Futuremark junkies :-) )

    http://www.futuremark.com/forum/showthread.php?p=1...">http://www.futuremark.com/forum/showthread.php?p=1...

    You would want one :)

    if you want to know, 3DMark 2001SE posted a score of 45k, and 3DMark 2003 posted a 33k, FACTORY.
  • phaxmohdem - Friday, July 14, 2006 - link

    I was wondering if you could possibly do some "single core" performance testing in the CPU's by simply changing the HAL to "ACPI Uniprocessor PC" instead of "multiprocessor PC" This would tel windows to only use one of the Core2 Duo cores, and I think would give us a good indication of how it will perform when released. Especially on the 2MB cache models, since I'm guessing the single cores won't bust out the door with a full 4MB.

    Just some food for thought/consideration. I personally would love to see a few tests run this way and compared to some single/dual core A64's.
  • JarredWalton - Friday, July 14, 2006 - link

    Conroe-L is so far off that it's really an afterthought. The only reason it's not shipping is that Intel has a ton of Netburst stuff to offload, IMO. At $145, the PD 945 is still interesting in terms of certain computational tasks. (I miss the QMD Folding@Home cores....)
  • fishbits - Friday, July 14, 2006 - link

    I had hoped AMD could gain even more market-share before something like this happened, would rather see the two CPU makers on more even footing. Intel just hit this one out of the park however, if pricing holds and availability is decent. Was looking at upgrading my 3500 (939) to a $300 X2 4600 after the price drop, but now? It looks like I'll probably keep this system and build a new rig around a $300 Intel 6600, which wins over or smokes the 4600 depending on benchmark. I really don't know what AMD can do to keep me in the short term, because I don't know if they can make any money dropping prices as far as it looks like they'll need to go.
  • SpaceRanger - Friday, July 14, 2006 - link

    One thing to consider though, if you decide to go with the C2D 6600, then you will also have to spend $$ on a new Mobo + Memory (since your DDR RAM won't work on the platforms for C2D. Keep that in mind when you're upgrading.
  • fishbits - Saturday, July 15, 2006 - link

    When I said that if I went with Intel I'd keep my current rig and build a new one around the Conroe, didn't it dawn on you that I'd already "considered" the need to get a new mobo, memory, etc?
  • SpaceRanger - Sunday, July 16, 2006 - link

    Holy arrogance.. EXCUSE ME for pointing something out..
  • epsilonparadox - Friday, July 14, 2006 - link

    Well the loads of money they made when they had the better performing architecture can hopefully keep them doing very well also the Opteron isn't really being challenged and thats really their moneymaker right now. They shouldn't have any problems lowering prices on their desktop chips while keeping the status quo on the Opterons.
  • zsdersw - Saturday, July 15, 2006 - link

    They can't really keep the status quo on the Opterons either. Woodcrest excels (or, at the very least, is equal to the Opteron) in the 1P and 2P server space.. which is a huge chunk of the overall server market.
  • LoneWolf15 - Friday, July 14, 2006 - link

    While I'm sad to see AMD lose the performance crown, I'm not so obtuse as to deny it's happening.

    I'm excited by Intel's newest chip, but I think the results for me will be that I'll buy a faster Athlon 64 X2 when the prices drop (assuming Socket 939 ones become cheaper as well, I don't plan a move to Socket AM2 for some time to come). So Intel's newest chip should benefit even those of us sticking with an AMD system. :)

Log in

Don't have an account? Sign up now