PDA

View Full Version : Is this proof that Intel is better than AMD A64 @ min fps?


acrh2
08-26-04, 04:47 PM
http://www.xbitlabs.com/articles/cpu/display/doom3-cpu.html

myshkinbob
08-26-04, 05:18 PM
It proves x-bit have the best p4 motherboard available for testing, but only mediocre a64 ones, no nforce3 250 was used in those tests. Infer what you will. :)

acrh2
08-26-04, 05:47 PM
Not true. The differences between Intel and AMD in other sites' tests with nforce motherboards are the same as xbit's for timedemo1 test. It is only when they used their custom timedemo, did the things go very differently for intel cpu's. Also [H] had an article with fraps results of similar kind.
And yes, thank you very much. I'm trying to infer the truth.

BuzzLightyear
08-26-04, 05:50 PM
It's a tricky call, regardless of hardware, to extrapolate much from that, other than all of those setups seemed, in terms of how much difference the scores were, they were all pretty darned close.

If the demo is faster on AMD, but realtime play is faster on P4, it suggests that the chips are simply better than each other at different things. When AI and hard maths is involved, the P4 seems to squeak ahead by a hair, but when calculating and moving chunks of data around AMD takes the lead.

Overall it paints a picture of a very close match between the two - until you map the averages onto the respective prices.

sabrewulf165
08-26-04, 06:46 PM
HardOCP's benches don't look anything like Xbit's benches... HardOCP shows AMD delivering much more consistent framerates in situations where they aren't crippling the graphics card by trying to attain "the highest playable settings".

I am all for more realistic benching, but HardOCP's new testing methods leave me with no idea of what to expect at the setttings that I play at... I don't really care who is better at 16x12 4xAA, my monitor doesn't even go that high! :rolleyes:

Anyway, I won't try to say who's chip is better or worse, but I will say that I personally have always found Xbit's numbers in many of their reviews to be very out there and far off from what I've personally experienced, so I tend to disregard most of what they say about performance. Every other site in the 'net has said AMD owns Intel at Doom, so I'm inclined to disregard the Xbit anomaly. And please don't say "but they used real testing" because they aren't the only ones that have done that, and they're still the only ones showing Intel as being that competitive.

Anyway not trying to step on anyone's toes, just stating my opinion. Even if Xbit is right and everyone else is wrong, AMD still has the better price/performance and is better in most every game on the market, so it's all kind of a moot point.

On a personal note, I have both the rig in my sig as well as a P4 3.2 HT, and I personally feel (and have tested) that the P4 has a lot more framerate dips in various games than the A64. Again, that's just me talking.

myshkinbob
08-26-04, 07:29 PM
acrh2, i really wasn't trying to flame you or anything, no offense meant at all.

burningrave101
08-26-04, 09:57 PM
The majority of games are primarily GPU intensive and when your playing at high resolutions with AA + AF enabled your not going to notice much of a difference if any between a high-end P4 and a high-end A64.

Very few games are truly CPU intensive and if you will look at benchmarks that compare a P4 to an A64 at 1600x1200 you'll see them scoring almost exactly the same fps and sometimes the P4 is even a little faster because of the motherboard or other system configurations used.

Personally i'm probably going to be buying a high-end A64 and an nForce 4 motherboard next spring/summer but gaming is NOT a CPU intensive sport unless your a Morrowind fanatic like me. :)

Multitasking, rendering, encodeing, compression, productivity software, office applications, and so on, are CPU intensive tasks and better represent the true performance of a CPU. The A64's gaming performance is coming from having an on-die memory controller which drastically cuts the latency between the CPU and the RAM thus increasing gaming performance which is all about latency. If the A64 didn't have the memory controller built into the chip it would NOT be faster then a Pentium 4 in gaming.