An exclusive gaming industry community targeted
to, and designed for Professionals, Businesses
and Students in the sectors and industries
of Gaming, New Media and the Web, all closely
related with it's Business and Industry.
A Rich content driven service including articles,
contributed discussion, news, reviews, networking, downloads,
and debate.
We strive to cater for cultural influencers,
technology decision makers, early adopters and business leaders in the gaming industry.
A medium to share your or contribute your ideas,
experiences, questions and point of view or network
with other colleagues here at iVirtua Community.
Spring IDF 2006 has officially kicked off, and it seemed Intel was eager to prove itself innocent of lying. Conveniently, an Intel system and an AMD system were on display at IDF 2006, and Anand Shimpi (and perhaps another Anandtech fellow) benchmarked the two systems.
Initial benchmarks were reported by Anand Shimpi here, but after criticisms by Rahul Sood in his blog, the Athlon FX-60 faced the Intel Conroe E6700 for a second round here, addressing Sood's criticisms. The Conroe tested here is not an Extreme Edition, but rather a docile, yet high-end desktop chip (estimated retail value, $529). Keeping in mind that this chip is operating not only on stock speeds but also at lower power than it's enthusiast-oriented rival (which retails for $1,061).
In all fairness, Conroe hasn't launched yet, so AMD has a few more months of dominance. All the same, I have to say I feel that Conroe will be an outstanding performer in July. AM2 benchmarks do not indicate that the K8 benefits very highly from the faster memory, and in many cases reduces overall speed, albeit slightly. This article, as Kooldrew points out, should be taken with a grain of salt as Tom's Hardware knowingly posts it with CPU-Z reading a memory speed of 100 MHz, and with the setup knowingly suffering from an apparent \"memory bug.\"
Still, AMD will have maintained the lead for slightly over a year by the time Conroe hits, so it is with the competitive spirit in mind that I close this entry with a nod of the head. Well played, AMD. Well played.
-Pikl
EDIT - 12:04 PM MST; 3/9/2006: Fixed system specifications (as they are now known much better), included links to the initial Anandtech article, Rahul Sood's blogpost, and the revised Anandtech article, deleted the first (unnecessary) part of the newspost. It was a little biased. I apologize. :D
* The initial tests seemed to be using timings of 5-5-5-15. This was fixed to the set value of 4-4-4-15 yeilding very little performance increase for Conroe.
** The initial Anandtech benchmark was caught by Rahul Sood as using the DFI BIOS version \"D49C-32,\" issued in October of 2005. Anandtech has re-done the benchmarks with the new BIOS version, \"D49C-33\" (dated December of 2005) up and running.
Last edited by A_Pickle on Thu Mar 09, 2006 1:06 pm; edited 1 time in total
In that article Tom was using an early engineering sample with a bugged memory controller. A look at the synthetic memory benchmarks confirms this as DDR2 667 in a synthetic bandwidth benchmark would beat the crap out of DDR/400. So don't expect these results to be representative of anything.
EDIT - Actually, look at the CPU-Z screenshot. It shows that the memory is running at PC2-3200 speeds. So it's a BW of PC2-3200 CL4 against PC-3200 CL2.
The processor we used for our tests was an Athlon 64 X2 4800+ for Socket AM2, which made its way to motherboard companies in mid-December 2005
In other words, it's old.
It has the same technical characteristics as its Socket 939 brother: 2 x 1 MB L2 cache, and 2.4 GHz clock speed. However, while we could select DDR2-800 speed in the BIOS of the engineering test motherboard, it obviously did not set the memory faster than 667. As already mentioned, the memory is suspected to suffer from a performance bug, which may be true or simple a matter of the early product stage.
In other words, we know our results are meaningless, but we'll use them anyway for page hits.
Really this article is so deeply flawed it is pathetic. This is why I never visit THG. Everybody should do the same or at least take what you read with a grain of salt.
Last edited by KoolDrew on Wed Mar 08, 2006 3:18 pm; edited 1 time in total
EDIT - Actually, look at the CPU-Z screenshot. It shows that the memory is running at PC2-3200 speeds. So it's a BW of PC2-3200 CL4 against PC-3200 CL2.
The current CPU-Z versions do not read the memory information correctly.
Quote:
Really this article is so deeply flawed it is pathetic. This is why I never visit THG. Everybody should do the same or at least take what you read with a grain of salt.
That particular article has it's flaws, most definitely, but I don't think AM2 is going to astound us by any stretch of the word. K8 is not a bandwidth starved architecture, and it hasn't been since the shift to Socket 939. Well, DDR2 isn't doing much beyond adding more bandwidth, but it should result in overall lower power consumption and less total system cost. Still, the \"memory bug\" Tom's made note of leads me to believe that the full scope isn't being seen.
I take the IDF Conroe benchmarks with more legitimacy that I do these benchmarks of Tom's Hardware. I think Conroe is looking mighty good right now.
-Pikl
EDIT - 11:36 PM MST; 3/9/2006: In response to criticisms made by Rahul Sood in his blog, Anandtech (with the permission of Intel) have been given the opportunity to address Sood's criticisms over at IDF. The systems have been more closely inspected, and they have been tested using third-party timedemos. I've updated the system specs on the initial post of this thread.
Last edited by A_Pickle on Thu Mar 09, 2006 12:36 pm; edited 1 time in total
As much of a AMD fan I am, I'm inpressed with this Conroe that Intel is coming out with. Now let's just hope this is really how they'll be when they come out.