New P4 vs Athlon benchmarks

[chiwawa626]

http://www.watch.impress.co.jp/pc/docs/article/20001121/hotrev85.htm

http://www.watch.impress.co.jp/pc/docs/article/20001124/hotrev86.htm

they are in japaneese, but u can understand who is wining in the bechmarks from the graphs

 

[KarsinTheHutt]

Tom's Hardware Benchmarks, Mark 3

Tom's also had a new set of benchmarks up... based on what I've read Intel has some MAD software engineers!

 

[Mr.IncrediblyBored]

Sorry, Chiwawa, but Karsin's link was much better.

 

[Adul]

I think a wait and see attitude is more appropiate for the p4.

 

[Remnant2]

Intel's re-compiled scores are very impressive.

What really impresses me, is how much the Athlon and P3 gain from the new compiler : Performing almost twice as fast as they do with the previous compile! From what I've heard/seen so far, it looks like Intel has probably the best vectorizing compiler out there, able to take more or less standard code and extract useful parallalism out of it. I don't know of anything else on the market that can do this.

It would be so interesting to see what the Athlon could do with optimizations actually made for it, but since AMD doesn't have the kind of budget required to write a compiler, I doubt we'll ever see this happen.

Among other things, I've heard that implementing prefetch commands in that kind of streaming-heavy app will give at least 20% improvement; the default recompile does NOT include the improvements from prefetch.


This has actually revised my opinion of the P4 upwards quite a bit. What has always made 3dnow, SSE, et al have very little impact, is that they required special coding. If you can get this kind of performance gain for no extra work on the programmer's side, Intel is going to be back on top in short order...

 

[Valhalla2000]

yup, intel really have very good software engineers to do theses things. Just imagine if they were in AMD.

 

[Soccerman]

heh, I'm not that surprised that applications optimized for the P4 do much better (that's the way it was made).

if AMD wants to compete, they're gonna have to do some thing to make software optimizations easier for developers.

 

[urbantechie]

LOL. I see all these post, "P4 Is suck!!!!" and now the scores smacks the Athlon in the face and everyone is complementing the P4.

 

[Remnant2]

P4 still is suck for current-day code. However, if the IntelC++ compiler can outperform MSVC++ on both Intel & AMD machines, its probably pretty likely that it will get some use.

We'll see. The real kicker is when Clawhammer comes out, with support for SSE2. When that happens, all the time and effort Intel has been spending developing a base for the P4 will instantly benefit AMD.

This maybe explains why some people lately have said that AMD seems to be shifting its focus away from K7 improvements, and going full throttle for the K8.

 

[IBMer]

We'll see. The real kicker is when Clawhammer comes out, with support for SSE2. When that happens, all the time and effort Intel has been spending developing a base for the P4 will instantly benefit AMD.

Of course just like MMX, AMD will have to pay for licensing of the technology. If it is super popular, the more money they will have to pay.

 

[BFG10K]

I see all these post, "P4 Is suck!!!!" and now the scores smacks the Athlon in the face and everyone is complementing the P4.

I knew that was going to happen. And it's only the beginning...

 

[LocutusX]

I said before that I see the P4 today as the equivalent of the PPro when it first came out. WHEN IT FIRST CAME OUT. i.e. late 1995/1996. That was *BEFORE* WINDOWS NT 4.0 came out! So obviously, when all the mainstream idiots ran benches on Win95 and Win311 they were complaining about it's lacklustre performance (since Linux was a bit immature still and no-one used the various weird PC Unices available at the time). Later in 1996, QUAKE first came out, and so did Windows NT 4.0... THEN all the idiots retracted their comments as they saw a potential for this bohemoth of a CPU (Quake 1 on a P6-150 beat a PMMX233!). I kinda knew that all along, because what intrigued me initially about the PPro was the RISCness of the CPU, so it was fairly obvious it was useless for all the 16-bit software of the time. I bought one in early 1997 and of course I used purely Windows NT 4 and I played a lot of Quake 1 (running in DOS). Likewise, we see here that the P4 only has an advantage in Quake 3 (!) and MPEG-4 encoding. So give it a year and it'll also kick ass, just as the PPro did.

Don't flame me though; I understand you all are VERY disapointed that it is useless for TODAY'S software, which is obviously the most important thing to most people. But this is to be expected of any CPU which brings a whole new architecture to the table.

 

[xtreme2k]

people saying P4 is suck before any software optimisation is PURELY short-sighted

 

[BoberFett]

Actually, for most people it is a stinker. Intel is selling this thing on the shelves at Best Buy. There was never a Pentium Pro system for sale at any large retail chain AFAIK. The fact that they're marketing this toward mainstream users where it will give them the least benefit is what stinks. If they were limiting it towards a small, select segment of the market I'd be fine with it, however they're trying to sell one to everybody. That's the difference.

When the PPro came out, it was a fine chip. Great in 32 bit performance and for some certain tasks. I would never have purchased on for myself running DOS and Win 3.1 though, or told any friends or relatives that they need one of these new "wunderchips." That would have been a flat out lie, bit Intel needs to sell chips so they need to convince everyone that it's the latest, greatest thing in every situation.

 

[RagingGuardian]

Until the price falls and it starts using DDR then it still sux. Right now a P4 upgrade is way too much for the minimal performance gains it gives.

 

[TopQ]

For TODAY the P4 SUCKS period. When software becomes optimized for SSE2 then we SHOULD see a big improvement in performance. IMHO the P4 should had come with a stronger FPU unit to run current software at decent (not even fast) speeds but also be ready for new SSE2 software. Yes from a technical point the P4 has lots of potential but from a practical point is it crap today. By the time we get widespred utilization of SSE2 code we should have P4 running at 2 Gig and higher anyway.
All this discussion kind of reminds me of the voodoo vs Gforce and T&L
issue with the difference that in that argument the Gforce was still very fast without any T&L software (which by the way is not even comon nowdays)
I am tired of P4 supporters trying excuse the P4 for its poor performace today. Like I said earlier you dont give the crown to future prospects but to current winners.

TopQ