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

  • mlittl3 - Tuesday, May 23, 2006 - link

    Until K8L (which will have microarchitectural improvements), there are a few things that could allow AMD to look good against Core 2.

    1) Price drops so that comparable K8 and Core 2 processors are the same price giving the same performance/$ ratio (this metric is important for the budget constrained)

    2) Nov. '06 release of 65 nm AM2 processors so that K8 and Core 2 processors will have the same performance/watt ratio (65 nm could give K8 a 20% drop in power and with Core 2 being 20% faster, they will have the same performance/watt ratio which is popular now). Also, everytime AMD transitions to a new die process they add some minor bug fixes and minor memory/microarchitectural enhancements which could also boost performance by a few percent.

    3) Continual improvements to DDR2 latency might yield a 2-2-2 DDR2 800 memory module which will probably benefit K8 more (maybe ~5% improvement) than Core 2 but this is a wild guess here and I don't know if it is even possible. However, DDR400 latency started around 4-4-4 and dropped to 2-2-2 so it could happen.

    With the same performance/price and performance/watt as Core 2, K8 could stay competitive and OEMS and users decide on which company (if not both) they would like to do business with. This is all speculation and of course everyone is more than welcome to rip my reasoning to death.
    Reply
  • JarredWalton - Tuesday, May 23, 2006 - link

    I seriously doubt we will ever see DDR2-800 running at 2-2-2 timings. (Feel free to quote me on this in the future and make fun of me if I'm proven wrong. :-)) Just think how long we had DDR memory around, and no one ever managed to create 1-1-1 DDR-400 memory. I do think we will see 3-3-3 DDR2-800, and possibly even higher bandwidth with those timings. In fact, we almost have that already judging from my experiences so far with socket AM2. (I can post and run benchmarks, but I wouldn't call the system 100% stable.) Reply
  • mlittl3 - Tuesday, May 23, 2006 - link

    LOL! I will! Reply
  • MacGuffin - Tuesday, May 23, 2006 - link

    From Patriot's PDC22G8000+XBLK Rev. 2 review on PCSTATS.
    Rated for DDR2-667 @ 3-3-3-9 (Maintains those timings through DDR2-940!)
    Rated for DDR2-1000 @ 4-4-4-12 (Goes Up to DDR2-1020!)
    Completely stable on the Intel platform they used. It's extremely expensive (saw it for $400+ at NewEgg). But yes, it is possible to run 2GB at these timings already. Its just extremely expensive.
    Reply
  • EdisonStarfire - Tuesday, May 23, 2006 - link

    any opinions on AMD offering a Clearspeed solution as stop-gap in the high end desktop arena ? Reply
  • Griswold - Tuesday, May 23, 2006 - link

    The bottom line is, we now know what we knew last fall, or rather (rightfully) assumed.

    quote:

    AMD does have one last trick up its sleeve before the end of the year, and you will hear about it in June. It's not K8L and it's not going to affect the majority of people, but it is an interesting stop gap solution for the high end in 2006...


    Now you made me curious. Could that be the "noise in june" which Henri Richards mentioned in a Register interview earlier this month?

    Reply
  • smn198 - Wednesday, May 24, 2006 - link

    It is called quad-core. Reply
  • temp2 - Tuesday, May 23, 2006 - link


    The extremetech.com article has a similar teaser at the end, but it is slightly more specific:

    "And given recent discussions with AMD, we can safely say that the company hasn't launched its last FX series CPU for the year quite yet."
    Reply
  • mino - Wednesday, May 24, 2006 - link

    This provided, 3.2 or even 3.4 FX's on 65nm are on the way... Reply
  • Scrogneugneu - Tuesday, May 23, 2006 - link

    Beware the mighty Sempron FX 32 ! Reply

Log in

Don't have an account? Sign up now