Hardware AMD Opteron 2435 CPU Review

The lead in statement on the front page of bit-tech (i.e. not the front page of the article) is not bias against Intel because it comparing the specs of Intel and AMD's six-core CPUs. In this case AMD's six-core CPUs do have far more memory bandwidth and are more power efficient.

However, given that Intel's six-core CPUs (Dunnington) are now pretty much end-of-life, and a server, not a workstation product, in the review itself, as you can see we've benchmarked the Opteron 2435 against Intel's flagship CPU, the Xeon W5580. And clearly the article itself isn't bias against Intel either - the Xeon W5580 comes out better in every benchmark and gets recommended, while the Opteron 2435 doesn't.

 

I was just wondering if in future some sort of fluid dynamics simulator might be a worthwhile benchmark for these heavily multicore systems?

 

The Euler3D benchmark on page 6 uses CFD (computational fluid dynamics) calculations - it's very well multithreaded so loves the extra cores these sorts of systems have.

 

Sorry, I realised that yesterday when I read the article but this morning that popped into my head. Guess I had a brain fart.

 

Your first point is valid. With fresh eyes on it today, I see that you were comparing the AMD 6-core to Intel's 6-core dunnington.

However, you did not even address my second issue, and the misleading information exists in the article. Let me refresh your mind:

The AMD Operton 2435 (like all 24xx series) has a TDP of *115 watts*, NOT 75W. The value of 75 watts is AMD's "ACP" value -- their so-called average power consumption -- NOT thermal design power. So it is NOT fair to compare AMD's ACP versus Intel's TDP.

Additionally, I think it is disingenuous not to mention that the 2nd fastest Xeon 55xx chip running at 2.93Ghz (and 3.2Ghz in turbo mode) has a 95W TDP. So In fact, except for the W5580, Every Xeon 55xx chip has a lower TDP than the Opteron 2435. (95W v 115W)

 

Sorry for not addressing both points earlier, it's *beeping* annoying how Intel and AMD quote power consumption numbers in such different ways. However, please rest assured that because of this (and the figures could be PR dogs droppings anyway), the analysis in the review is based on our own measured power consumption figures. These are taken at the wall (i.e. the whole PC) because we don't believe its useful to know how much power a CPU consumes - as that doesn't take into account other components in the rest of the system.

 

Thanks for replying.

I did see that you have a power draw test which is good. However, the information at the beginning of the article is STILL MISLEADING. I don't understand why you won't correct the error. Here is what you wrote:

"Despite having two more cores than earlier Opterons, the new 2435 has the same low TDP of 75W - nearly half the 130W of Intel's fastest workstation/server CPU, the Xeon W5580. "

Again, regardless of the actual specifics in how AMD or Intel measures power consumption, AMD's *published* TDP for the Opteron 2435 is 115W, NOT 75 watt. Specifying it as 75W is incorrect. Why do you willingly continue to publish false information?

 

I'm more than willing to update the article but as yet AMD have failed to confirm the TDP to us, and as far as I can see the TDP is not published on any of the spec sheets/product pages. Not that I'm doubting you, but what is your source for the 115W TDP figure?