Does AM2 Reduce the Impact of L2 Cache Size?

We've already seen that Socket-AM2 doesn't really impact performance except for in games, but does the higher bandwidth memory controller reduce the impact of AMD's 1MB L2 cache parts compared to its 512KB L2 cache offerings? 

 Benchmark - Athlon 64 X2 2.0GHz Socket-939 (1MB vs. 512KB Advantage)  Socket-AM2 (1MB vs. 512KB Advantage)
Cinebench 9.5 Multi-Core Rendering Test 0.2% 0%
3dsmax 7 0.3% 0.6%
Adobe Photoshop CS2 2.7% 2.5%
DivX 6.1 0% 0%
WME9 0% 1%
Quicktime 7.0.4 (H.264) 0.9% 1.3%
iTunes 6.0.1.4 (MP3) 0% 0%
Quake 4 - 10x7 (SMP) 4.8% 3.5%
Oblivion - 10x7 7.5% 3.3%
F.E.A.R. - 10x7 8.6% 6.2%

 

In the application benchmarks there isn't really a difference in how performance scales with cache size between the two platforms, but looking at the games there is indication of a pattern that is developing.

In Quake 4, Oblivion and F.E.A.R. the 1MB L2 cache seems to make slightly more of a difference on the Socket-939 platform than on the Socket-AM2 platform.  While the 1MB cache offers a 4.8%, 7.5% and 8.6% performance advantage in those three games on the Socket-939 platform, on AM2 the advantage is cut down to 3.5%, 3.3% and 6.2% respectively.  The explanation being that with a lower latency memory controller and more available memory bandwidth, the benefits of a larger cache are reduced on Socket-AM2. 

However the differences in performance scaling that we're seeing here are small enough that once you take into account the amount of variation you can see between runs, it's not really worth concluding anything concrete based on this data.  What we do see here is a trend of the 1MB L2 cache parts doing less on Socket-AM2 than on Socket-939 (another way of looking at it is that the 512KB are doing better on AM2 than they did on 939), but the margins are small enough that we can't really say for sure what is causing the trend.

Once again, the trend only seems to impact games, as the other application tests we've run appear to be basically unaffected. 

The Question on Everyone's Mind: Is AM2 Faster? How Does the New 4000+ Stack Up?
POST A COMMENT

83 Comments

View All Comments

  • darkdemyze - Tuesday, May 23, 2006 - link

    z-ram isn't due for AMD procs for quite some time, I doubt this is their plan for June.. Reply
  • mlittl3 - Tuesday, May 23, 2006 - link

    Basically this is what I said above for my guess of the "trick" AMD will use. Anand said it will only affect some high-end users, read FX series so it can't be price cuts as some have suggested (that would effect everyone). Adding L3 cache is the only performance improvement I can think of that doesn't require changing the microarchitecture of the cores (well at least not a big change).

    However, TDP is still an issue here as someone above suggested. I don't know how much more power it takes to run L3 cache. Last time AMD did it was on K6 and power wasn't really measured back then.

    By the way, please ignore Questar's comment below about z-ram being pig slow. I really don't think he knows what he is talking about. /shields eyes from incoming Questar flame
    Reply
  • johnsonx - Wednesday, May 24, 2006 - link

    K6-III did not have L3 cache. It had L2 cache, making the cache that all socket-7 boards had then an L3 cache.

    So, let's stop saying things like 'AMD hasn't done L3 cache since K6-III', etc.
    Reply
  • mino - Wednesday, May 24, 2006 - link

    Well, IMHO the point is AMD has used exclusive 3-level cache structure in the past so they have som experience with thi arrangement. Reply
  • Questar - Tuesday, May 23, 2006 - link

    No flame here, look it up for yourself.

    Z-RAM has high capacitive loading, which results in slow speed.

    At 4MB it'll run half the speed of SRAM.
    Reply
  • Questar - Tuesday, May 23, 2006 - link

    Large amounts of Z-RAM are pig slow. Reply
  • Ecmaster76 - Tuesday, May 23, 2006 - link

    Seriously, the one area an Athlon X2 would be bandwidth starved and does it get tested in the preview? NO
    In the review? NO

    How long ago did we know that the K8 was not bandwidth limited in single application usage? YEARS


    So yeah, DDR2 din't increase the 3dMark, big surprise
    Reply
  • mlittl3 - Tuesday, May 23, 2006 - link

    I think 3dMark06 is multithreaded now so all available cores and bandwidth should be used within the limits of the program. I could be wrong about this however. Reply
  • Ecmaster76 - Tuesday, May 23, 2006 - link

    3Dmark06 is almost completely GPU limited. The 3Dmark CPU score did increase a bit, but I really was referring to graphics benchmarks in general. Reply
  • cscpianoman - Tuesday, May 23, 2006 - link

    I was just noticing the performance differences between the FX and the EE. In some cases the FX tromps the EE by "gasp" 30%! In other cases the EE makes it's mark. This is part of the reason I am skeptic on Conroe. Yeah it's good. But I always take what Intel, or AMD for that matter, with a grain of salt. Just today we saw the 30% advantage translate down to about 15%. This seems just like any other generation change where 15% is to be expected. The current hype for the Conroe is a product of Intel's excellent marketing dept. Reply

Log in

Don't have an account? Sign up now