Engineering Sample of A64 Benchmarks

Athlon 64 is basically an Opteron so we already know what to expect performance wise - its gonna rock especially in gaming benchmarks.

<font color=purple>Ladies and Gentlemen, its...Hammer Time !</font color=purple>
 
But do you really think that? Imagine a single Opteron with less cache (right?) used for the purposes that 64 bit probably won't help with. That's really going to be an excellent chip?

I'm just your average habitual smiler =D
 
An opteron @ 2.2ghz has been reviewed with the same cache as A64 on an early nforce 3 board. It already beats the 3.2 P4, and once the boards/memory and process has matured, things will only get better. I think that it will at least compete with the new pentium. For the A64 to really shine though, it will need cas 2 memory and a few tweaks here and there.

<A HREF="http://service.futuremark.com/compare?2k1=6868334" target="_new"> MY RIG </A>
<A HREF="http://service.futuremark.com/compare?2k3=1150155" target="_new"> 3D-03 </A>
<font color=red> 120% overclocker </font color=red> (cheapskate)
 
The 940pin A64 will still have 1MB cache...the low end 754pin will have 512KB L2 cache and single channel mem controller.

The 754pin version will probably be equivalent to a low/mid range Athlon XP.

<font color=purple>Ladies and Gentlemen, its...Hammer Time !</font color=purple>
 
The only thing that bothers me is that the A64 sucess really lies in the acceptance and the fast migration to 64 bits. It's good to see amd trying to go a new path but they depends on many other compagny too...To bad AMD doesnt have the Intel budget to really accelerate the adoption...

So at first it will depend on A64's 32 bit performance...
I'd like to see AMD succeed. Last amd success, the early athlon series as done nothing but good to the CPu market. Put more pressure on Intel and its always good.
 
Exactly. In my opinion A64 will really shine in Q1 2004, when the 939pin version will be released. That CPU will be able to accept CAS 2 DDR400 memory, and possibly when used with a newer chipset (Nvidia's Crush maybe?), then we will see even more performance.
 
I think AMD64 adoption is going pretty well already. Opteron has only been out for 4 months and they already have the following support:

1) SUSE Linux x86-64
2) Red Hat (still under development)
3) Windows XP/Server 64 (under development)
4) Oracle beta 64bit for Opteron (avaliable to download now)
5) 64bit Java (due 2H04)
6) Apache Web Sever 64 (under development)
7) 64 bit version of RSA Security Encyption

The snowball has already started rolling down the hill...

AMD64 will really take off when Microsoft release 64bit XP.

<font color=purple>Ladies and Gentlemen, its...Hammer Time !</font color=purple>
 
Check out <A HREF="http://www.amdzone.com/articleview.cfm?ArticleID=1329" target="_new">this</A> review of Opteron @ 2.2GHz. I was personally impressed by its performance. The first 940pin A64s will be basically Opterons and we already know that no CPUs under 2GHz will be released. So this review is exactly what we can expect from the dual-channel version of A64. Later on, in 2004 when A64 will be able to also accept non-ECC memory (so this means CAS 2), then we will see even more performance. Of course many will say that this is a review from a site biased towards AMD and all that but even if these benchmarks are not 110% accurate, they are still very good indication about what we will get in about 1 1/2 month.

So to answer your question, I am not sure whether the A64 will be able to beat Prescott, but on the other hand I don't believe Prescott will dominate either. What I am sure though is that A64 will not be a flop.
 
"AMD64 will really take off when Microsoft release 64bit XP."

that is the single most important thing in the whole pc market right now probably.

"keyboard error or no keyboard present, press F1 to continue or DEL to enter setup" spot the deliberate mistake
 
Indeed...I predict on the Athlon 64 launch date we will see a beta version of Windows XP 64 with the final release sometime in Q104.

My only worry is 64 bit driver support from hardware manufacturers.

<font color=purple>Ladies and Gentlemen, its...Hammer Time !</font color=purple>
 
why?

The a64 will be able to process 32 bit app's with ease...i feel that there will be stability issues (like we saw in ME) if manufacturers scramble to write 64 bit drivers...32 bit drivers will be fine IMO...eventually the manufacturers will switch over but i don't think it is a high priority...

If i put my k6 in a Ferrari it would be faster than your your pentium 4 or Athlon XP :tongue:
 
i dont really know anything about this kind of stuff, but isnt it quite easy just to recompile 32bit drivers/applications into 64bit?

"keyboard error or no keyboard present, press F1 to continue or DEL to enter setup" spot the deliberate mistake
 
i really don't know...i beleive you could...but the program would run inefficiently as 64 bit because it would not take full advantage of all the extra registers?

I dunno i am over my head...but if it were that easy...then windows me would have been a stable os!

If i put my k6 in a Ferrari it would be faster than your your pentium 4 or Athlon XP :tongue:
 
i dont really know anything about this kind of stuff, but isnt it quite easy just to recompile 32bit drivers/applications into 64bit?
Not all that easy.

First the driver code has to be 64-bit clean. That basically means no stupid assumptions about integer or pointer sizes. You'd be surprised how often minor programmer mistakes slip by and blow 64-bit cleanliness out of the water.

Then the driver code has to account for different PCI memory mapping. PCI devices and drivers have the luxury of full access to the 32-bit address space for bus-mastering et al (although some, like the SBLive!, impose their own artificial limits on PCI DMA addressing). When the addressable space expands to 64-bit quantities, that may become a problem. PCI devices may find themselves having to use a "DMA hole" under the 4GB range, like some ISA devices require under 16MB.

Third is possibly different overall conventions for things like SMP et al. Remember, NUMA is something that isn't yet supported in MSWindows. When it is, device drivers may have to adjust the way they work to get maximum mileage out of NUMA topology. Of course, this won't matter so much on uniprocessor systems.

Using the expanded register set is a rather distant fourth. It's not so important, except for bumping speed a bit. Also, updated compilers can theoretically take care of that without any fuss.

That being said, driver support isn't all that far behind. nVidia's already posting AMD64 drivers, at least for Linux. I imagine they'll have corresponding Win64 drivers ready to roll.

<i>I can love my fellow man...but I'm damned if I'll love yours.</i>
 
yeah, those amd64 drivers have been there for a while havent they.

do you think applications will be easier than drivers to make 64bit? is that just a question of recompiling, or does it require re-writing from scratch in 64bit?

also, i am almost cerain the answer to this is yes, but, if (or when) intel eventually releases a 64bit Pentium, will it be 100% compatible with 64bit drivers/applications/OS' etc. that have been written based on the Athlon 64?

"keyboard error or no keyboard present, press F1 to continue or DEL to enter setup" spot the deliberate mistake
 
I still am suspicious of those benchmarks. AMDZone is pro-AMD, god knows what they could have fitted the P4 with behind those already revealed specs.

Furthermore they used a lot of pro-AMD benchs, like the Science ones where Athlons always did well.

--
<A HREF="http://www.lochel.com/THGC/album.html" target="_new"><font color=blue><b>The sexiest website ever, guaranteed XXX!!!</font color=blue></b></A> :wink:
PS: New sexy users' sites now added! :smile:
 
What's wrong with PC66 CL3?

----------------
<b><A HREF="http://geocities.com/spitfire_x86" target="_new"> My Website</A></b>

<b><A HREF="http://geocities.com/spitfire_x86/myrig.html" target="_new"> My Rig</A></b>
 
Should work in async mode? (100/66, 133/66, 200/66)

----------------
<b><A HREF="http://geocities.com/spitfire_x86" target="_new"> My Website</A></b>

<b><A HREF="http://geocities.com/spitfire_x86/myrig.html" target="_new"> My Rig</A></b>
 
One may argue about how importantly the amount of memory affects those benchmarks, but 2G on opteron vs. 256M on P4? Shame on AMDZONE!