Helping to make this the longest thread in history.

What's the record right now?

Maybe because some deleted swearing from Nail/Chryx?

Hmm...thread count is still off by 4. Anyone care to add 2 more posts and take us to the 300 mark? :D

Damage, unless you completely missed it, the Athlon XP they compared latency with runs at the same Mhz!

So i would say it is fairly accurate, at least between the two AMD cpus.
If i recalculate the latency figure for the P4 due to speed difference, it should land at about 90-95% of the latency of the XP 1.6Ghz, and i dont think thats too far off either.

Also, remember that those are theoretical benchmarks, and probably gives the "best case".

THG has posted their Opteron review and I have to say I'm fairly impressed for the most part. It's completely dominating the server benchmarks, and I think that's the market that AMD is truly aiming to penetrate. Its workstation performance is a little lackluster in some cases and rather competitive in others, but overall, it doesn't perform nearly as good in the workstation benchmarks as it does in the server benchmarks.

I'm thinking that once the clockspeed increases on these babies, their workstation performance will get a lot better. Moreover, if developers start optimizing for the Opteron like they have for the P4 and Xeon then that can only increase performance further. Overall, it's an excellent showing for the Opteron and I'm curious to see which OEMs are going to announce their support in the next 12 hours or so. Now, if September would just hurry up and get here :)

I think we will be seeing more reviews on the Athlon 64 on Tuesday because that is when the NDA will be lifted.

it also says "283 Comments(s). 6 Pages(s)"

Comments(s) & Pages(s) lol

I'm the guy who sent you that link, Damage :) As far as the A64 preview is concerned, I must admit that I was a little taken aback by the performance. Somehow I just expected it to blow the XP 2800+ and the P4 2.8GHz out of the water. Unfortunately, this didn't turn out the be case. Yes, this is clearly an engineering sample, but at the same, it would've been great if it actually did blow those other processors out of the water. That being said, looking back at the article now, there are some pluses and minuses to the performance that we've witnessed.

First off, the A64 looks incredibly strong in gaming and compression, particularly for a sample that's only working at 1.6GHz and doesn't even have a proper platform to work on. However, it is clear that the 1.6GHz clock speed is holding back the A64 big time, as it is effectively a XP 1.6GHz with an insane memory controller. And that's why I'm worried. Unless the performance of this chip increases exponentially with every increase in frequency, I have to wonder how the A64 is going to compete against a Hyperthreading Prescott on a 800MHz FSB that's simply going to get faster and faster as the year progresses.

Sure, the A64 will probably dominate the gaming benchmarks and the compression benchmarks, but in my opinion, I think the only way for the A64 to be truly successful is to dominate all the other processors out there. Winning those two fields are simply not going to be enough to dethrone the P4. AMD cannot settle on simple performance parity, especially not with a new architecture. If this thing doesn't crush everything else, then it can almost be considered a failure. I'm just hoping that things improve a lot come official launch time. And I personally think they will. After all, AMD, more than any of us, knows how much they need this chip to succeed.

For God's sake... STOP POSTING YOU BASTARDS. Don't you realize that you are making yourselves look like idiots :roll: Saying that AMD sucks denotes how stupid a person can be. The same goes for the ones saying Intel sucks. DIDN'T YOU SEE THE LAST REVIEW ABOUT THE AXP 3000+ AND INTEL P4 3.06? One processor didn't wiped the floor with the other. Amd ruled in UT2k3, Intel kicked ass in SSE2 optimized code. Please do a favor to the Tech-Report community... SHUT THE HELL UP

Everybody's taking this article a little too seriously considering how far we are from seeing retail versions of the CPU and chipset that were tested. The good performances show some possible potential, but the poor performances may just show where they make the most improvements by production time. I hope nobody's making buying decisions based on this test.

#233, Let's get something strait. People will not invest money in the business that producing products with a very low potential for success. That is how stock market works.

LOL. I guarentee this person hasn't made a dollar in any american stock exchange. This is the most oversimplied foolish perspective on stock markets i've ever seen.

Ok, I just had to launch a couple of tactical nukes and impose sanctions on a pair of rogue posters. Ugh. Grow up, guys. The bans will be lifted in a week, if everybody involved can grow up.

For any of those who are curious, and I'm sure most if not all of you are, Ace's Hardware should be dropping their Opteron review at midnight :D

good god last time i looked at this thread there were like 62 replies.

this beast needs to be tamed.

What will be really interesting to see is how the 2 cpu and 4 cpu Opterons will be able to perform. Most 3D programs have multithread capabilities, and looking at the specs it seems the Opteron platform will be able to scale well.

It's ashame AMD isn't really living up to all the hype! perhaps they should rebrand that chip to 2500+ instead of 2800+ as it seemed to perform on par (slightly) with the p4 2.5. none the less this is still a beta chip and shouldn't be judged just yet. I could care less who makes the fastest chip out there all i want is the fastest i can buy for the money i have with the best quality.

Is Athlon 64 1.6G faster or Pentium-M "Banias" 1.6G faster ??

Good lord, when I left earlier tonight this was like a 70 something thread after like 2 days, when the heck did it blossom an extra 130 threads in the space of a few hours :P!

Yes, I have to agree. This thread has gotten a little out of hand.

Jeez, you'd think this was Slashdot. Or Usenet. WTF?!??

So far I've noticed three things that have changed on the Athlon64 that could and should affect scalability. Packaging, Process, and Pipeline (HEHE 3 P's!!). I think it would be naive to believe AMD went through the process of adjusting all three of these aspects of the Athlon core in order to get a measely 18% clockspeed improvement over the Athlon XP. I'm also more convinced that we might see quite a nice clock disparity between the Athlon XP and Athlon 64 given the information that the Opteron die is a lot smaller than previously expected (to the tune of 140mm^2 vs. expected 180~200mm^2). With that known it might be a case where the Paris (256KB L2 Athlon64) will be smaller than the current Barton processor (can you imagine how low power consumption might be?? 45watts @ 2Ghz??) and will be able to pull on the numerous improvements made to scale quite considerably above it (25-35%?). Granted that's just my own silly beliefs, but it's certainly plausible :).

I'm looking forward to April 22nd. Ace's Hardware is going to drop their Opteron review on that day. And I'm pretty sure a lot of other sites are planning on doing the same thing. Going to be an interesting read and an interesting day, that's for sure :)

Found this posted anonymously to Slashdot:

Being that I am an AMD employee, I know for a fact that B0 (the version of clawhammer that they are benchmarking) is a very early rev, and should not even be considered when thinking about final benchmarks. Geez.

Well, one thing is for sure, xbit labs just blew their chances of ever getting their hands on another upcoming bleeding edge technology again.

Of course since it was posted anonymously, we can't be sure if it's for real. But it's interesting if it is. :)

Well, I'm still liking the idea of a dual 2+Ghz Opteron system :)

I agree with Xylker. Anything less than 2 gig at launch and this puppy will lose.

This time next year...maybe a die shrink and I'd bet the 64 will start stretching and showing its legs.

This is really fun, but I am a little concerned about their memory latency numbers. They haven't specified what units those numbers are in, but latency numbers come out of programs like cachemem in CPU cycles. Obviously, processors with higher clock speeds will see more clock cycles pass per second than processors with lower clock speeds. One must convert those numbers into comparable units, such as nanoseconds, in order to compare CPUs at different clock speeds. I do expect the Athlon 64 to have low memory access latencies because of its integrated memory controller, but I don't think the gap will be so great as the X-bit numbers would seem to indicate.

Don't they give scores for a similarly clocked Athlon XP system on an nForce2 motherboard? Why don't you just use the percentage of improvement from that to guessimate leveled performance.

I just wanna know why Damage had to edit/update the front page to post his .02, instead posting here like us. :)
Show off

Sorry for repeating, this is the last time...
You need to release at 2.2GHz (37.5% faster) to achieve superiority.

Athlon64, P4, XP2800, Athlon64
2.2GHz, 2.8GHz, 2.2GHz, 1.6GHz

7234 , 6943 , 6411 , 5261 , PCMark2000 CPU Score
49 , 31.4 , 35.5 , 35.5 , Business Winstone 2002
329 , 300 , 246 , 239 , SysMark 2002
263 , 220 , 197 , 191 , Sysmark 2002 Office Productivity
239 , 232 , 263 , 329 , MP3 Encoding, LAME 3.92
81 , 129 , 142 , 111 , Data Compression WinRAR 3.0
43 , 38.82 , 32.75 , 31.24 , MPEG-4 Encoding Flask0.78.39/DiVX 5.02
198 , 197 , 237 , 272 , Windows Media Encoder 9
839 , 689 , 626 , 610 , 3DMark2002 Default CPU Score
22425 , 16385 , 15517 , 16309 , 3DMark2002 SE CPU Score
278 , 217 , 186.7 , 202.3 , RTCW (Checkpoint)
97 , 61.59 , 67.96 , 70.62 , Unreal Tournament dm
289 , 210.16 , 197.15 , 210.16 , Unreal Tournament flyby
84 , 88.07 , 87.91 , 115.07 , ScienceMark Molecular DYnamics
400 , 474.34 , 449.41 , 550.13 , ScienceMark Primordia
13 , 18.28 , 15.14 , 17.81 , ScienceMark Cipher
281 , 250 , 313 , 386 , 3DMax Final Rendering Underwater
87 , 102.5 , 140.3 , 119.9 , LightWave 7.5 RayTrace
122 , 128.5 , 134.1 , 167.5 , LightWave 7.5 Sunset
314 , 338 , 262 , 228 , Cinema 4D Cinebench 2003 Raytracing

"preview" indeed. Why do people bother with this crap? It's done when it's done. Geez, society really has you all in this MUST HAVE NOW mentality, methinks it ain't so healthy.

They did? Didn't notice it, but I was skimming... and now the server's hosed :(
Yes obviously what speed they're able to milk out of it will make a big difference. But this is 6 months in the future, right? Maybe we'll know more in a few days with the opteron launch. I don't see why they would be launching it at this speed. It'd be shooting themselves in the foot. As far as I know the changes should be POSITIVE for going to higher clockspeeds than the Athlon XP (longer pipeline, SOI). But yeah I suppose it works out well for them that the Opteron is coming out first. In the server arena, the clockspeed probably won't be as important as the 64 bit capability. At least there, there's a reason to buy it. At 1.6Ghz for the desktop, what's the point? I can't see them releasing it in september at much less than 2Ghz and still being competitive. They might want to rethink the PR ratings on these too. The XP 2800+ trounces it in most cases. Something a bit more conservative might have more credibility. Let's not go back to the days of Cyrix.