Seems AnandTech in their Nehalem preview (http://www.anandtech.com/cpuchipse [...] spx?i=3326) has deliberately modified the scores of Penryn from a cople of monts ago. Anand is claiming that the reduction of scores is due to a Vista update. (is that true?)
Scientia (http://scientiasblog.blogspot.com/ [...] hokes.html) has an article describing the diferences between scores of Penryn from january and the scores it used to compare Penryn with Nehalem. And the finding is that Penryn scores were higher back then by 980 points. (Efectively reducing Nehalem's advantage to 10% in a multithreaded scenario)
I don't know why Anand hasn't provided an explanation for reduced Penryn scores in the Nehalem comparison and if there really is a valid reason to those reductions (a Vista update? that has reduced the score of 3D Max 9 by 21%? I personally doubt it, and I would like to see a confirmation that this indeed happened)
Anyway I hope to see more professional testing soon.
Message edited by Cryogenic on 06-06-2008 at 10:50:05 AM
We'll see... Pretty soon many other sites will post their results and if your claim is true Anandtech will be discredited...
The Nehalem thing is getting interesting, on TH there's that video of "insane overclocking" on air without disclosed speed, on AnandTech they only benchmark an 2.66 GHz Nehalem (which is lower than Penryn) and also Nehalem has a higher power usage than Penryn in their tests ... then there's the "ajusted" Penryn scores ... Frankly I believed all along that Nehalem will deliver about 30% over Penryn clock for clock and higher frequencies but now I'm beginning to have some doubts about that.
if anand did modify the numbers, why? they don't seem to have anything to gain by doing so. did you take into account the fact that they underclocked the penryn to match the nehelem's speed for a clock-by-clock comparison?
I would hate to think that shenanigans are afoot over at Anand. But, if memory serves me correct (and maybe not after smoking one too many blunts) Anand posted some shady early benchmarks regarding Conroe. So, with that said, it's plausible that some numbers "modification" took place.
I have no doubt that Nehalem will be a performer, but I'm gonna reserve judgement until more 3rd party reviews are available.
Shame that Nehalem will initially be priced outside of my budget.
Message edited by chunkymonster on 06-06-2008 at 01:50:12 PM
---------------
You got your cozy little corner
All night you're jammin' on your feet
Hangin' out just like a street sign
And put a twenty dollar trick
if anand did modify the numbers, why? they don't seem to have anything to gain by doing so. did you take into account the fact that they underclocked the penryn to match the nehelem's speed for a clock-by-clock comparison?
They actualy updated their Penryn Scores in the article
Quote :
"(Updated: The original single-threaded Penryn Cinebench numbers were incorrect, we've included the correct ones):"
"Cinebench shows us only a 2% increase in core-to-core performance from Penryn to Nehalem at the same clock speed. For applications that don't go out to main memory much and can stay confined to a single core, Nehalem behaves very much like Penryn. Remember that outside of the memory architecture and HT tweaks to the core, Nehalem's list of improvements are very specific (e.g. faster unaligned cache accesses)."
Message edited by Cryogenic on 06-06-2008 at 02:07:10 PM
I've noticed before that AT does not really establish a 'coherent' and consistent base of benchies (like Tom's CPU Charts). They run through a bunch of different tests and systems and it gets really difficult to make comparisons.
He did the same thing with Skulltrail. He ran both SLI and Crossfire on Skulltrail and compared the results side by side ....
The only problem was in comparable tests (games and resolutions) several months previous on the 790fx chipset the Crossfire results were substantially higher. Kinda gave a false impression ....
Vista could take some performance out but recently it has been performing much like XP. I am not sure what they did to make the socres what they were.
I think they said the Penryn setup was a bit handicapped as was the Nehalem setup since it wasnt at its true peak performance.
I do have no doubt that Nehalem will probably lay the smaketh down on Penryn when we have nice mobos from Asus and such and the silicon is in its final stages.
I also think the power consumption will go down. Its possible that the mobo was not regulating the voltage sufficiently so it could have possibly caused it to use more power than it truly will.
For someone so uptight about proper testing methodologies and calling into question all the details, it would be helpful if he actually looked the details of the tests before spouting off. When Sci points out that Anand's old review showed the Q9450 get a score of 3297 in the single threaded Cinebench run, which is obviously higher than the one reported in the Nehalem preview score of Q9450 = 2931, it should be noted that was because the old test was using the 64-bit Vista OS while the new one is using 32-Bit Vista.
Looking at this Vista comparison benchmark by Extremetech, the Vista 64-bit gives a 10% advantage over 32-bit, which easily explains a the difference.
