What do you think of this?

Summary: The P4 3.2 EE Wins 32 Times, The Athlon 64 FX-51 15 Times - An Uncertain 64-bit Future For AMD

After four weeks of strenuous uninterrupted testing, sleepless nights, unending phone calls, innumerable updates, tiresome conferences and lots of back-and-forth with the manufacturers, the results were clear: Thanks to its ideal configuration and use of the best components, the P4 3.2 in the Extreme Edition (actually a Xeon labeled 'P4') wins the performance crown. Its former gaming weakness against the Athlon 64 has been ironed out by the 2 MB L3 cache. The AMD Athlon 64 FX-51 is only marginally slower. Especially with 3D games, both the 1 MB L2 cache as well as the fast memory access (HyperTransport) help it decisively beat competitor Intel P4 3.2 (standard edition).

It's not yet possible to judge the performance capability of the Athlon 64 in pure 64-bit operation, since the alpha version of "Windows XP 64" we used and the dearth of available programs don't allow a clear evaluation. And the 64-bit Linux isn't really practical for daily desktop tasks. AMD delivers its GZIP Packer optimized for 64 bits with the standard package, which is a whole lot speedier compared with 32-bit operation. But a closer look at this fact reveals a marketing trap: if the code is optimized for the P4, too - as it was in our test - then the Intel CPU turns out to be faster under conventional 32-bit operation. So there's no compelling argument for 64 bits there. AMD makes its case using the chicken/egg scenario: Only once a sufficient number of Athlon 64 CPUs are available on the market will the software industry respond accordingly. Besides for solid performance in the 32-bit realm, users should consider x86-64 compatibility as something of a bonus option, even if there's no software for it at the moment.

Anyone who can't handle the whole, awesome magnitude of this mega-test can download our new video number 10 right now.

The new Video 10 can be downloaded here:

* Server 1
* Server 2
* Server 3

It contains a summary of the essential highlights of our battery of tests. The video is also suitable as an introduction or overview of this topic.

Intel doesn't have to decide yet whether it wants in the medium term to build its desktop CPUs on the complex IA-64 architecture or to go with x86-64 like AMD. But if the market should unexpectedly shift towards 64 bit, the manufacturer still has its secret Yamhill project up its sleeve.

Since Intel already has an inkling of what the outcome of the eternal duel between Athlon 64 and P4 will be, the manufacturer hastily introduced the "P4 Extreme" a few days ago at the IDF (Intel Developer Forum 2003) in San Jose. We were there: the processor is nothing more than an Intel Xeon with a P4 label tacked onto it, complete with a 2 MB L3 cache, now offered with FSB800 (200 MHz real FSB speed) and 3.2 GHz. To get the faster clock speed under control, the ECC checking in the CPU was unceremoniously deactivated. A few hours before posting this article the Athlon 64 was ahead of the Pentium 4 Standard Edition. But with the P4 Extreme Intel managed to considerably spoil AMD's launch. Now the latest Intel CPU wins in most of the benchmark tests. So was it a fair move for Intel to make such cosmetic changes prior to the actual launch of the Athlon 64? We see it as the infantile reaction of a monopolist who's naturally inclined to act like a general at a sand table exercise.

Click to expand...

Avg said:

That article seems to be pretty old since it is comparing the 3.2EE and the fx51 and both Intel and AMD have new High end chips, but even if the EE is a little faster then the fx51 which used registered ram (the new fx53 uses regular ram) the fx51 was still much cheaper then the EE.

Click to expand...

j3lly said:

athlon64 would be much more "future proof" than any 32bit P4
PCI-E would be nice,that is until its bugs and kinks are worked out.
DDR2 is not such a big deal tho,it could be skipped over soon.
have you heard about FB-DIMM? the new RAM standard:

http://www.theinquirer.net/?article=18358

Click to expand...

hUMANbEATbOX said:

agp will antiquate your systems faster than 32bits ever could. by the time 64bit apps are actually the standard, amd will probably have gone through 5 more sockets types, none of which will be compatible with each other.

also to call fb-dimm the "new standard" is VERY premature, ddr2 will be around way before this, and will most likely be more prevalent. when i think fb-dimm, i think rdram, but time will tell. (and rdram still kicks ***).

Click to expand...

j3lly said:

LOL,actually intel "will probably have gone through 5 more sockets
types,none of which will be compatible with each other."

any 64bit OS or 64bit apps don't care whichever socket xyz is used,
AGP or PCI-E. as long as the CPU support 64bit,it'll be compatible.

btw,what does DDR2's advantage over DDR1 again?

Click to expand...

hUMANbEATbOX said:

umm, any kind of change to the system ram with a a64 will entail a socket switch. single channel>dual channel = 754>939. ddr1>ddr2 = 939>900. i belive this is due to the on-board memory controller.

ddr2, while not extremely useful right NOW, it will be, when fsb speeds pick up. ddr667 IS going to be a mainstay, at least on p4's, where memory bandwidth has always been a prority.

what gives you the impression that p4's will require socket changes? right now, you can get lga775 board with agp support, pci-e support, ddr1 support, ddr2 support, all with the SAME socket. what would give you the impression that intel will change sockets? they haven't said anything to support this, yet amd has already announced s900.

Click to expand...