Doom 3: CPU Battlegrounds

by Anand Lal Shimpi on 8/4/2004 2:14 AM EST
Comments Locked

59 Comments

Back to Article

  • kabob983 - Thursday, September 9, 2004 - link

  • kabob983 - Thursday, September 9, 2004 - link

    So where does the A64 3200+ fit on this scale anyways...above or below the P4 3.4EE?
  • T0918273645 - Monday, September 6, 2004 - link

    this statement from their review is partially wrong.


    "Doom 3 sees system memory as one big cache and drives performance up considerably. It is also the on-die memory controller that makes cache size less of an issue on the Athlon 64, while too small of a cache seems to make or break performance with the Pentium 4."

    notice the 3.2 E with its 1mb cache gets 72.8 while the 3.2 C with 512KB cache gets 68.7. the only difference between this chips is the cache, which results in a difference of 4.1 FPS.

    look at the FX53, with dual channel, and 1mb cache, it gets 103.4

    and the 3800+ with 2.4ghz, 512kb cache, and dual channel gets 99.8

    a difference of 3.6. So the larger cache does help. I think maybe anandtech might have mixed up the fact that they were comparing with amd the difference between dual channel, and cache. while with intel they were comparing only the difference between caches.

    i think looking at the AMD charts you can see how close amd's view of the tradeoff for dual channel with the loss of half the cache actually is, at least as far as doom3 is concerned.

    The 3400+ (no dual channel, 1mb cache, and 2.2ghz) performs at 80.4 FPS

    The 3500+ (dual channel, 512KB cache, 2.2ghz) turns in 79.5FPS. So the dual channel/ cache tradeoff is so close you'd never be able to tell the difference.

    So basically they mixed up the fact that the smaller cach size of the 3500+ compared to the 3400+ is masked by the dual channel ram which nearly completely makes up for the loss in cache.

    I'm surpised they overlooked that fact.

    Or maybe they thought that a difference of 4.1 is more of an issue than a difference of 3.6, but really that difference is nearly the same.
  • manson909 - Monday, August 9, 2004 - link

    i can't believe no one has thought of this: since Doom 3 is capped at 60FPS, why not compare the MINIMUM framerates the CPU's encounter when playing? a system that has an average of 60FPS but is locked in at this speed (no dips) is far superior than a system that has an average of 70FPS but with lows in the 20's & high's in the 100's imo... anyone?
  • HermosaBeach - Friday, August 6, 2004 - link

    Short version:

    I would like to see the final graph comparing CPUs using 1600x1200, Highest Quality, 4xAA (Anti-Aliasing) and 16xAF (Anisotropic Filtering)

    With this graph we may actually see a flattening of the frames per second where an increase in the CPU has no real impact on frames per second. This is where the GPU (GeForce 6800 Ultra) hits the wall and becomes the limiting factor.

    Dave
  • HermosaBeach - Friday, August 6, 2004 - link

    Dear 53, I disagree. Here's another way to look at the problem or question. Let's say I own an ATI X800 XT PE. I would like to know what computer (CPU/MB) to get. I want to play games at 1600x1200 with 6xAA and 16xAF. Starting from the slowest computer like the AMD XP 2000+. As you improve the CPU and climb the performance ladder we expect the frames per second to increase. But, there may come a time where an increase in CPU does not appreciably increase the frame rate. At this CPU level getting a faster CPU is not really going to help you. I was hoping to find this magical CPU point. Early 2003, with my ATI 9800 Pro, the sweet spot was the AMD XP 1800+. A faster CPU imporved frames per second, but not by much and was not worth the significantly increase in cost. For example, with FarCry and Doom3 the AMD 64 3000+ might be the sweet spot where getting the AMD 6400 3700+ really does not significantly increase your frame rate. If this is true, I would get the AMD 64 3000+. If the frame per second continue to significant improve as you increase the CPU then I would get the top end CPU. Sadly, at 1280x1024 without 4xAA and 16xAF you really can't tell.

    Dave
  • skiboysteve - Friday, August 6, 2004 - link

    Dave, your retarded. Try reading the review, esp the first page.
  • HermosaBeach - Thursday, August 5, 2004 - link

    This article should have used 1280x1024 and 1600x1200. The 800x600 resolution was a waste of time and reading. Who's going to purchase any AMD 64 or any ATI X800 series or NVidia 6800 series card and play at such pathetic resolutions. I found the article a waste of time and it certainly did NOT answer the question - what CPU to get. I own a ATI 9800 Pro and it certainly did not help with what CPU to get when I purchase my next graphics card - sigh.

    Dave
  • xxxfubar187xxx - Thursday, August 5, 2004 - link

    Thanks for the benchmarks guys! You do great work over there. Why wasn't the Athlon64 3700+ included in these benchmarks? Being the top of the line Socket 754 processor I figured it would be included in the tests.
  • gimpyd00 - Thursday, August 5, 2004 - link

  • Anemone - Thursday, August 5, 2004 - link

    I suggest you look at the low and average framerates for the EE @ 1600x1200. From memory I recall on HardOCP was that the AMD chips had a higher top end and a lower low end, making the average, whereas the P4's had high and low ends not so far away from their average.

    AMD64's fly, but at higher res's 1280 up to 1600, the difference is a few fps, and P4's seem to not fly as fast, but also not bottom out as bad in the super heavy scenes.

    While I'm torn on the issue of which platform to go to, that is something I noticed.
  • DigitalDivine - Thursday, August 5, 2004 - link

    Shinei, hardocp did not use apples to apples comparisons.
  • dmxlite - Thursday, August 5, 2004 - link

    "When tested at 1600 and real gameplay, Athlon 64 falls like a brick, as shown in the "Official Doom 3 hardware guide" at HardOCP."

    Yes, that must be the reason why in the conclusion they say "AMD came out ahead in DOOM 3 performance with the strongest CPU in our tests, the Athlon 64 FX-53 processor," and why they chose the stock FX to be in their [H]ard|OCP Ultimate DOOM 3 System and an overclocked A64 3000+ to be in their [H]ard|OCP Ultimate Enthusiast DOOM 3 System.
  • nlr_2000 - Wednesday, August 4, 2004 - link

    It's called an on die memory controller ;)
  • Staples - Wednesday, August 4, 2004 - link

    Interesting how P4s were always top dog in Q3 but now with D3, it is the opposite.
  • Shinei - Wednesday, August 4, 2004 - link

    I don't know about you, T8000, but I can't complain about my (underclocked) 64MB Ti4200 and 3200+ XP combination. It drags in some areas (9fps as a low, 60 as a high, average gameplay flow is around 20; settings are 10x7 medium quality), but it's definitely playable in singleplayer, with very little mouse lag.
    So, for my Athlon XP to be performing fairly well with 2-year-old, underclocked hardware, I'm willing to say that HOCP is probably making things retarded for AMD by changing settings or something. I find it very hard to believe that the vastly superior Athlon 64s (compared to my Athlon XP) would falter so heavily in gameplay when my own processor does not.
  • T8000 - Wednesday, August 4, 2004 - link

    I think you cannot say a CPU is better because it produces more frames in situations where it is hardly stressed, like lower resolutions and timedemos. Furthermore, I think most people want to play at an acceptable framerate with as much detail as possible, so testing a GF6800U only at 1280 and without playing the game, is not realistic.

    When tested at 1600 and real gameplay, Athlon 64 falls like a brick, as shown in the "Official Doom 3 hardware guide" at HardOCP. Only the overclocked Athlon FX could do slightly more then the P4EE at stock settings.
  • flexy - Wednesday, August 4, 2004 - link

    >>>
    Whether the top-performing FX53 is worth the $811 price is up to you,
    >>>

    The answer is of course: NO

    Neither is the PIV EE worth its price :)
  • flexy - Wednesday, August 4, 2004 - link

    the most important info for me is that it's actually nonsense to wait for socket 939...or useless to spend more money for 1mb cache.

    So...socket 754 is even better/faster and cheaper. And the 512kb cache are plenty too for the A64.

    This is very helpful info in regards to upgrading my system soon :)
  • Pumpkinierre - Wednesday, August 4, 2004 - link

    Nice article Anand- strange set of results. I'd like to see the benchmark for a 2.4c at 3200MHz with same memory latency settings. If it is memory latency as the a64 results suggest then the Prescott should be lower or even beaten by N'wood. If it is L1 cache size then the EE result should be lower.
  • PrinceGaz - Wednesday, August 4, 2004 - link

    The amount of System memory (above 512MB) is unlikely to have any impact on framerate in the timedemo as I doubt it would need to swap anything out after the first run (which is discarded anyway).

    I found my 128MB graphics-card (a Ti4200) gave an almost identical framerate at Low, Medium, and High quality settings in the timedemo even when gfx-card limited, provided Aniso was disabled in the driver for High quality mode (which would otherwise use 8x Aniso and impact on performance in other ways). So increasing the videocard memory from 128MB to 256MB will have no effect whatsoever on the timedemo, except maybe at Ultra quality which I didn't bother testing.
  • Steve Guilliot - Wednesday, August 4, 2004 - link

    #27
    That's the OS load balancing between the two procs. Two D3 threads aren't running at once. That's why sum utilization of both procs won't go over 100%.
  • Succorso - Wednesday, August 4, 2004 - link

    Is this review using XP or XP64 beta with the amd64 ? Are the benefits the same using a 32bitXP as opposed to the 64 bit XP ?

    Succorso
  • SignalPST - Wednesday, August 4, 2004 - link

    its interesting how DOOM3 runs best in the Nvidia/AMD combo along with the amazing price/performance that they offer over their competitors

    the Athlon64 3000+ is on par with Intel's 3.4GHz EE, while the price difference is $840

    the GeForce 6800 GT is faster than ATI's X800XT PE, the price difference being $160

    so in this scenario, the Nvidia+AMD combo can save you $1000 and still outperform the ATI+Intel combo

    bottom line, for DOOM3 and future DOOM3 engine games, ATI+Intel=losers
  • cKGunslinger - Wednesday, August 4, 2004 - link

    Yes, I would also like to see some numbers benchmarking 256/364/512/768/1024/etc MB memory configurations. When does the average system have *enough* ram to run WinXP and play a game?
  • xtf - Wednesday, August 4, 2004 - link

    Would it be possible to add the cache (and other) specs of the K7s to certain charts?
    Because sometimes the 2700 and 2800 are slower than then the 2500 and it'd be interesting to know why.
  • tdent1138 - Wednesday, August 4, 2004 - link

    Great article AT! I'm happy to know my 2.53Ghz @ 2.717Ghz P4 and 9800pro will happily run D3 at 8x6 in medium quality. I can now wait until HL2 at least to upgrade to whatever makes sense at the time (A64 something I imagine). Thanks again!
  • tdent1138 - Wednesday, August 4, 2004 - link

  • Philbill - Wednesday, August 4, 2004 - link

    Great article, Do you plan to give an update with the high end ATI cards?
    Phil
  • dangereuxjeux - Wednesday, August 4, 2004 - link

    Somehow, I feel ashamed that the Sempron 3100+ crushes my ol' P4 2.4C.... please please please stop publishing articles like this that encourage me to spend any more of my money upgrading to a new AMD platform to go along with my 6800.
  • michael2k - Wednesday, August 4, 2004 - link

    Heh, what about frames/$ graphs?
  • Da3dalus - Wednesday, August 4, 2004 - link

    Lovely graphs, *pats my Athlon 64 3200+ while waiting for Doom 3*

    Another week for Doom 3 to hit our european shelves, damnit I hate waiting for something you americans already have :-/
  • elfy6x - Wednesday, August 4, 2004 - link

    I have a dual Opteron 246 setup, with 1Gb of ram, and a Radeon 9700 Pro. I'm not really a gamer, but I gave Doom 3 a shot, and it utilizes both of my CPU's when I play. I have nothing else running when I play the game. It doesn't tax both CPU's to 100% but one CPU fluctuates between 40-50% while the other one bounces around 10-20%. So something is processing two threads. Just my observation. :)
  • PotatoMAN - Wednesday, August 4, 2004 - link

    I second the idea of benching memory and video memory at AT for doom week. I have a 3200+ A64 and I am wondering if I am starving it more with a 9800 pro (128) or with my 512mb of RAM. Thanks AT for being awesome!
  • Gooberslot - Wednesday, August 4, 2004 - link

    I bet a 1-1.2ghz P3 or Athlon would cut it for a minimum. Just go back and look at the old benchmarks of the P4 1.5ghz on here. Only in Quake3 was the P4 on top. In UT the P4 actually tied with the P3 1ghz. Pathetic. Stating minimum requirements based on those old Williamete P4's is very misleading. Perhaps the real minimum cpu requirements should have been 1.2ghz P3 or Athlon, 1.5ghz P4, or 2.4ghz Celeron. :)
  • matman326 - Wednesday, August 4, 2004 - link

    Man I always knew that an athlon 64 system was powerfull but a 180 dollar proc. beating Intels Extremly Expensive 1000 dollar proc is just mind blowing. So much for the Netburst design kicking butt.
  • AtaStrumf - Wednesday, August 4, 2004 - link

    If you could find some time, I'd like to see a comparison between 128 and 256 MB R9800Pro, and 512 RAM and 1024 MB RAM. Basicly how much of a difference do video and system RAM make.
  • at80eighty - Wednesday, August 4, 2004 - link

    dammit! looks like im gonna celebrate Christmas with a loan..grrrr... : )
  • WooDaddy - Wednesday, August 4, 2004 - link

    Two words:

    AMD ROX!!!


    I've been an AMD user since the 386 days and even had a NexGen processor (pre-pentium, K5). Never went Intel, never will (maybe)...

    Thatsright,
    Naw.. The charts don't lie. But to be fair, let's wait for the Intel-funded people (aka Tom's (blow)hardware) to put up their benchmarks. 10 bucks says they'll make all kinds of excuses to why Intel procs aren't fast enough.
  • Regs - Wednesday, August 4, 2004 - link

    Wow. Good info. Dual channel offers nothing. And the socket 939 2.2 Ghz CPU offers nothing over my 3000 A64 which costs 200 dollars less. Given that if you play on higher resolutions.
  • thatsright - Wednesday, August 4, 2004 - link

    Great article, but just one quibble. Is it just me, or does it sound like AMD paid AnandTech for the last sentence of the article: "In the end, the winner of the final battle is clear: the AMD Athlon 64 is the processor for Doom 3"

    It sounds like a perfect quote from the AMD marketing dept! LOL
  • DAPUNISHER - Wednesday, August 4, 2004 - link

    Good stuff! That was the most entertaining set of benchies I've seen in awhile. D3 is evidently turning out to be a better benching tool than game =)
  • kherman - Wednesday, August 4, 2004 - link

    Dear anandtech,

    Please do a similar benchmarking wtih system memmorry.

    Sincerely,
    me
  • ViRGE - Wednesday, August 4, 2004 - link

    #9, Carmack has stated that Doom 3 does not have SMP support.
  • CrimsonDeath - Wednesday, August 4, 2004 - link

    How about pumped up XP-M's? I got one running at 2700Mhz, that would make it better than most cpu's out there.
    Overclock till it burns, then overclock more!
  • punko - Wednesday, August 4, 2004 - link

    Guess my XP 1800+ ain't gonna cut it . . .

  • kherman - Wednesday, August 4, 2004 - link

    Well, my 2800+ is staying in my box!
  • kherman - Wednesday, August 4, 2004 - link

    In the days of GPU shootouts, this article is an EXCELENT idea! Good work! Still have to read it though.

    It's funny to, because I'm contemplating a move to 1 gig of memmory.
  • Wesley Fink - Wednesday, August 4, 2004 - link

    Just to clarify before the comments start, there are TWO 3400+ Athlon 64 as you will see in checking at New Egg for instance. The NEW 3400+ is Socket 754, runs at 2.4GHz and has 512k cache. The original 3400+ is still available and is Socket 754, runs at 2.2GHz, and has 1MB cache.
  • Wesley Fink - Wednesday, August 4, 2004 - link

    Please keep in mind that the small Athlon 64 increases in Dual-Channel and Cache-size are cumulative. While it's true that Dual-Channel only adds about 3% to A64 performance and 1MB cache only adds about 5%, the combination adds 8 to 10% improved performance. The FX53 is the ONLY chip that combines BOTH 1MB cache and Dual-Channel memory and is 8 to 10% faster than a comparable chip using Single-Channel (Socket 754) and 512MB cache.

    The Dual-Channel 1MB FX53 should be about 8% to 10% faster than the latest Newcastle SC 512K 3400+ even though both run run at 2.4Ghz. Another interesting point is that since Cache mattered more in Doom3 than Dual-Channel, a 3700+ (SC, 1MB cache) might actually be faster in Doom3 than a 3800+ (DC, 512MB). This was certainly true in Anand's tests of the 3400+ (SC, 1MB0, which was faster than the 3500+ (DC, 512MB).

    Whether the top-performing FX53 is worth the $811 price is up to you, but it is still a bargain in Doom3 compared to the 3.4EE that is 18% slower and still sells for $1000.
  • Thorne - Wednesday, August 4, 2004 - link

    SMP Anyone? Doom III should support smp... has anyone found it?
  • Thorne - Wednesday, August 4, 2004 - link

  • jrphoenix - Wednesday, August 4, 2004 - link

    Intel and gaming don't mix. Where is Peteroy when you need him? LOL. My 3800+ is looking better and better each day :)
  • magratton - Wednesday, August 4, 2004 - link

    Is there any possibility that you could produce some numbers for the cpu overclockers who placed their bets on the Athlon XP2500M instead of the A64? Great article, thank you!
  • ceefka - Wednesday, August 4, 2004 - link

    Oh my, I guess the Pentium EE is indeed the Extra Expensive Emergency Edition.

    I'm not much of a gamer though. It's a shame I wrecked my car a few weeks ago. I wanted to build a 939 since they also perform so well on sound and video. Have to wait now. Well there's one upside to that, prices will drop a bit, or so I hope.

    Say AT, how about soundbenchies, especially Cubase SX, or similar, with a fixed arrangement, sound & midi with VST instruments and then add stuff like plug-in effects until the system throws up. I have read that Dual Opteron systems can handle 100 audio channels with about 80 realtime plug-ins. How far does say the top 20 of these CPU's carry you. I suppose Hyper Threading would help out Intel here. So I'm very curious.
  • msva123 - Wednesday, August 4, 2004 - link

    If only you could download hardware...
  • Sonic587 - Wednesday, August 4, 2004 - link

    I find it remarkable how the A64 3000+ manages to get within 2FPS of the 3.4GHz EE. We have a lowly $160 processor almost besting a $1000 EE. Wow.
  • Adul - Wednesday, August 4, 2004 - link

    drool :D cant wait to upgrade
  • mkruer - Wednesday, August 4, 2004 - link

    LOL im still not buying and AMD 64 chip til the end of the year, but its good to know that the 3500+ on socket 939 still beats*ntels best

Log in

Don't have an account? Sign up now