Zardon

I could go into a detailed description regarding photoshop and the multiple improvements across the coding specifically with threads, but basically ill just sum it up, anandtech are using an outdated photoshop (that being V7, 2 revisions beyond the latest release), hence the different results.

www.hardwareheaven.com/photoshop shows some of the differences and we have a sticky thread in here showing various performance differentials across the board between different versions/hardware.

VikingThunder

I wonder how the the Athlon64 X2 3800+ would fare... doesn't seem as promising.

rgreen83

This is somewhat to be expected, the PSbench test used here usually favors clockspeed on single cores also, right?

Just comparing the AMD cpus though, something does seem not right with the X2 cpu utilization, it never goes above 50%, which it wouldnt if it were only using one core for photoshop. The X2 could still be faster than the FX57, as windows would still be offloading the other system tasks and completely devoting one completely to photoshop, but photoshop is def not using both. This could explain why the most cpu intensive (longer) tasks are faster on the fx57 while the short ones are faster on the X2, the opposite correlation of the intel single vs. dual cores.

Seems odd also that the intel is using three times as much memory and running 12 more processes, not saying anything sneaky is going on, just that things like background processes and such should be more or less removed from the equation, it may help the intel to have less background processes running.

I dont doubt the validity of the scores and Im not a conspiracy theorist, I just think that these few points, especially the cpu usage or load balancing, should be investigated further in the interest of your readers to find out what the problem is.

*edit* spelling

Sonata

This Intel CPU is cheaper initially but if you factor in an expensive mobo and the extra power it will consume over the course of its life, its not worth a gold award in my eyes. Photoshop isn't the be all an end all when it comes to benchmarking a computer.
This might be a silly question but what did Intel give the reviewer to say such nice things about its CPU?

Zardon

You are right it is a silly question, any processor at that price that can put out the performance it does in our testing suite will get a gold award. We dont need more hardware as we are literally overflowing with it - or maybe you mean a cash bribe? obviously you dont know Intel at all!

I guess going on your theory all the previous AMD gold awards already made us rich or have all the prior awards with the exact SAME benchmarking suite suddenly been forgotten?,

http://www.hardwareheaven.com/reviews/fx57/conclusion.htm
http://www.hardwareheaven.com/reviews/...conclusion.htm
http://www.hardwareheaven.com/reviews/FX55/Conclusion.htm

just a little surprising that there are so many "experts" making snide remarks when not one has yet to prove the so called controversal photoshop CS2 benchmark script results wrong. I actually think very few people read it indepth just jumped on the nonsense posting that the inquirer made today about AMD "fanbois" getting all miffed. boy were they right.

rgreen83

I hope I wasnt being referred to in the "snide remarks" category, as that wasnt my intention at all. My intention was only to point out that it looks like photoshop wasnt using both X2 cores as well as the P-D. I do know some of the routines of PS are not fully multithreaded, but the ones that the P-D takes advantage of, it seems the X2 does not. The P-D is definately showing how a dual core cpu should perform in PS.

I really dont care which companies cpu is faster, I have some from each and like them both. I do think maybe there is a problem with the X2 or PS that warrants further investigation.

swimtech

What the heck, throw this into the fire too, then throw a little water on it. These last posts remind me of the old Apple vs. PC debate...

The fire...

For years, PC fans had problems with certain Adobe products running apparently much faster on Apples (PowerPC/G3/G4, now G5 dual processor). There were accusations within the PC community that Apple had the benchmarks optimized for their hardware, and worse. True or not? Eh...

Now apparently, Adobe products are running faster on Intel than AMD. Has Intel "rigged" the bench - or good ol' DriverHeaven? I really, really, really doubt it.

The water...

It's about the money, that's all...

For years now, I have worked on and supported Apple hardware as part of what I do for a living - and every single one of those computers was running Adobe software - great software - does the job. And, Apple users are some of the most loyal customers for both hardware and software on the planet.
I think ( don't really know for sure, but it sure looks like it - finally! ) that Adobe has put some time and effort to "optimizing" their code for Intel hardware! They no doubt did not do it to slight AMD, but to expand their market base - to sell more software to professionals who pay top dollar for the Adobe software (and stick with it for upgrades because it's their work...) that runs on Intel hardware.

My point is that this bench result means that millions of folks running Intel hardware, for not much money CPU wise, will be able to run the latest Adobe software much faster. I think that's a good thing (too). It's not about Intel vs. AMD, it's really more about Adobe on IBM CPUs vs. Intel CPUs, under the surface - maybe - likely - I think so...

And didn't Apple just announce they were moving to Intel hardware?

Hmmmm.....

__________________
It's not so much getting your way that matters or not - what matters is how you go about getting it.

pr0digal jenius

Very nice review guys...and I can vouch for the gold award on this one.

FYI, this chip is currently being pimped as the be all end all chip for media center PC's, and is being shipped in the HP 7160n as well as the Gateway 835GM, both of which run XP Media Center and are designed to be an 'all in one' sort of device, not gaming rigs.


And as for the photoshop stuff....that's freakin crazy how well this thing did. Just curious allan, have you tried using SMP Seesaw with one of these dual core chips? I know it is sorta flaky on HT chips, but I wonder if it would kick a little harder in a dual core?

Sonata

[QUOTE=Zardon]

It's interesting that a different review site has given the Intel D 820 a 3 and half rating.
Please see the below site.

[color=white]http://www.hardwarezone.com/articles/view.php?id=1666&cid=2&pg=13[/color]

That's a long way from a gold award.

Anyway no hard feelings I can take what ever you throw at me

By the way I don't trust large corporations to act in an ethical manner, especially the likes of Intel.

Zardon

well if all they do is play games then yes 3.5 out of 5 seems fair - actually thats very good for a 2.8ghz chip (I know they didnt but its just a comment - I mean does anyone class 3dmark03 or 05 as a valid cpu test ??). then again Driverheaven doesnt solely base reviews on games, I think our readership do more than just play First person shooters (especially those contemplating this processor), and at the price the intel dual core is very competitively priced.

mike2h

sonata, do you trust the 'likes' of amd? if so, why? if not, where does that leave you?

__________________

The_Neon_Cowboy

For what you get, for what you pay, The 820 is and AWESOME DEAL

[QUOTE=Sonata]And look at thier tests looks like they looked at mostly single threaded applications
and compaired them with single core multi threaded cpu's with 2 mb cache Vs the
1MB cache on each core for the 820 (so 1MB on an non multi threaded application)

Were else can you get 5.6 GHZ of processing power for under $250 retail box
keep in mind the cheapest dual core amds are how much? $520 oem

You don't see its true power with out a multi threadeed application!!!

THERE IS NO WAY you can compaire a SINGLE CORE multithreaded CPU
to a DUAL CORE cpu as it much faster and better in multi tasking

__________________

bug77

I could fit into that category.
But hey, it's the first time I have seen a product given a top award based on a single application performance (if the AMD X2 claims are real, then a patch might blow this award to pieces). I just wanted some more detail to understand that decision. You see, many people don't use Photoshop, so for them this isn't really a 'gold award' product.

PS: Why wasn't it benched with games? While it may not be targeted at the gaming community, it would certaily add to the completeness of the review, imho.

PPS: I still like DH and I do not think those results were influenced in any way.

Zardon

Well i mentioned earlier in the review it wasnt benched with games due to time limits on stuart and the fact it had to be passed onto another reviewer. However as ive also said, we have two key points to this comment. the first being we did mention in the review that the processor isnt for die hard gamers, I mean I had expected the people reading this review to be of a higher education than those reading for example a gamespot review and know already that a 2.8ghz P4 wouldnt exactly be excelling in a pure gaming environment, I guess we need to spell that out in extreme detail next time if we dont have time to run a barrage of gaming tests at various resolutions. Very few games use HT never mind dual core so we felt it was more worthwhile in the limited time we had to concentrate more on multi threaded applications and real world testing as we feel this is the market place for the cpu. Most people contemplating a dual core 2.8ghz P4 would understand the benefits and certainly wouldnt be purchasing for a 1600x1200 or 1920x1200 hardcore gaming situation.

Really this debate could go on for months, with everyone having a differing viewpoint on it, but I really dont know what else to say

bug77

I wasn't thinking of a barrage, I was thinking of a few standard (as in run for each review - how about Riddick, Doom3, HL2 and RTW/LotR:BfME at one or two resolutions?) tests that would give a more clear picture. If time is a problem, you can always leave some tests out and run them at a later date. Or something.

Anyway, point taken about not reading thoroughly. Though I try...

Edit: Man, I'm starting to tell reviewers how to do their jobreview stuff. I should have my head checked.

The_Neon_Cowboy

It pretty pointless to do that unless you can test in in multi threaded gameing situations.
(If you a hard core gamer you'd go for nothing less then 840 anyways)

Or doing things that people rarely do like running a game on one core and running some encodeing on another (and about wich virtaully no one will care...wasted effort)

You'd need to use a mix of multi threaded game to be fair and untill we see things
like the unreal 3 engine in use. More future games will surely ue it but many havn't
botherd becouse multi threading or dual CPU support wasn't trageted for gameing.
But for intence applcations....

Photoshop is used by more people then you'd think too...
I'd hate to fathom a number but many times and many
Millions more use photoshop then even the most popular
game ever!

Most people that buy intel are look for the superior multi tasking support anyways.

__________________

bug77

You kinda missed the point: imho a review should be complete, in order to allow a fair comparison to other products. For CPU that would include both single and multithreaded applications. That's all I was trying to say. 'Complete' being the keyword.