5700 Ultra review

[Genx87]

Interesting review. Seems like the cards are pretty close to each other in performance. Some goes to ATI, others go to Nvidia.
One thing that is interesting is I think it was X2. Nvidia card beats up on the 9600 pretty good and this is a DX9 based game.

 

[nemesismk2]

oops my mistake!

 

[Sazar]

Originally posted by: Genx87
Interesting review. Seems like the cards are pretty close to each other in performance. Some goes to ATI, others go to Nvidia.
One thing that is interesting is I think it was X2. Nvidia card beats up on the 9600 pretty good and this is a DX9 based game.

not sure if I concur with the beating up thing

the tables and the results overall dictate that the 9600xt and the 5700ultra are a lot closer than they really should be...

however it seems there may be a bug with the drivers and the 5700ultra...

it has higher clocks and more bandwidth than the 9600xt... it should theoretically score better @ high res and filtering situations... but this is not the case... not quite sure what is going on there...

 

[McArra]

It seems NVidia is reacting at last. The only problem is that driver cheatings are still very close and I don't know if i should trust the results or take em with a bit of salt

Anyway it seems both companies are working quite hard and results are being quite nice for us, consumers.

(looking forward to see Volari in action and see how good it is )

 

[nemesismk2]

Maybe my Ti4200 can finally be retired when I upgrade to a 5700 Ultra instead, I never did like those fx 5600's!

 

[Genx87]

Interesting to note ATI goes to Bilinear AF in Toms review which Nvidia stays at Trilinear.

That is a big improvement imo.

AF comparison


Also interesting to note the IQ is pretty much identical between 52.16 and 3.8.
I guess we can attribute those issues with 51.75 as bugs and not cheats.

But this is rather interesting........

"ATi, on the other hand, repeatedly claimed that their drivers implemented no illegitimate optimizations. The accusations leveled against ATi at NVIDIA's Editors' Day two days ago thus become that much more serious. Epic's Mark Rein confirmed that in some cases, high-res detail textures were not displayed in some areas by ATi´s drivers and that standard, lower-res textures are used instead. Randy Pitchford of the Halo development team also mentioned that there were optimizations present in ATi's drivers which are detrimental to Halo's image quality. However, Randy didn't want to go into more detail here. Finally, Massive's new DX9 benchmark, AquaMark 3, also displayed some irregularities of ATi drivers in the overdraw test.
"

 

A quote from Anands final words: "After testing the GeForce FX 5700 Ultra, we have been very pleasantly surprised by NVIDIA. We mentioned last week that the 5700's new architecture might help to close the gap. In fact, NVIDIA has flipped the tables on ATI in the midrange segment and takes the performance crown with a late round TKO. It was a hard fought battle with many ties, but in the games where the NV36 based card took the performance lead, it lead with the style of a higher end card."

Not too shabby NV. Getting back to where they once were.
The 5700U even hung with a 9700pro on more than one occasion. Now thats impressive considering the power the 9700pro has.
199.00 after rebates I think would be well spent on this card. NV still needs to do some serious work on there AA performance although they have already
made up some decent ground there.

 

[Ferocious]

Neither card can touch the 9700 Pro for price/performance right now.

Wow, the card I bought a year ago is still the best today?!?!?

Who would've ever thought.

 

[jorojr]

The 9700 Pro looks like it has enough staying power to last through most of next year. At the very least, it should be more than enough until PCI-Express makes its debut. Quite impressive.

 

[BoberFett]

I really would have liked to see the 5900 included in the graphs, as that's the next step up in the Nvidia line.

At $199, I'm not sure how this product will do. The 5900 isn't much more expensive, and should give quite a bit better performance.

 

[Sazar]

just a thought about the review...

after all the publicity webwide concerning the quality of the last couple of 'reviews' done here... would it not be prudent to change some benchmarking policies to eliminate doubt and have SOME accountability...

when weighing the benchmarks and considering all factors... there is nothing like the TKO that the reviewer notes in the conclusion...

as one of the 'big three' perhaps anandtech should look @ getting better reviewers... or better review guidelines...

I welcome your comments on this evan...

ta...

 

[kylebisme]

i am guessing they are happy with doing their reviews the way they are now, and i imagine that nvidia is rather happy about it as well.

 

[bunnyfubbles]

Originally posted by: Sazar

Originally posted by: Genx87
Interesting review. Seems like the cards are pretty close to each other in performance. Some goes to ATI, others go to Nvidia.
One thing that is interesting is I think it was X2. Nvidia card beats up on the 9600 pretty good and this is a DX9 based game.

not sure if I concur with the beating up thing

the tables and the results overall dictate that the 9600xt and the 5700ultra are a lot closer than they really should be...

however it seems there may be a bug with the drivers and the 5700ultra...

it has higher clocks and more bandwidth than the 9600xt... it should theoretically score better @ high res and filtering situations... but this is not the case... not quite sure what is going on there...

could very well be bandwidth efficiency, it's why the 4600's superior memory bandwidth means little compared to the inferior 9500 Pro's bandwidth yet the 9500 Pro whipes the floor with the 4600 in memory bandwidth intensive situations

 

Can we please see your review Sazar?
I'm dying to see what you came up with..

Thanks in advance,

GM

 

[VIAN]

The review says 199 with rebate. Sounds pretty expensive, but still a nice gain.

I'm still bothered that ATI doesn't show all of the image and takes out some parts because they are barely noticeable. That could help a lot in performance. Thats their bandwidth saving techniques. Nvidia renders the entire image.

 

[Killrose]

I'm surprised at how poorly the nVidia cards do with 4xAA/8xAF enabled. Too me, those are the real numbers to pay attention to. If you compare the 5600U against the 5700U, you are seeing huge improvements across the board.

I really can't see a buy recommendation on the 5700U over the 9600XT based on the numbers shown by Anand.

 

[kylebisme]

Originally posted by: VIAN
I'm still bothered that ATI doesn't show all of the image and takes out some parts because they are barely noticeable. That could help a lot in performance. Thats their bandwidth saving techniques. Nvidia renders the entire image.

what evedence have you seen to back that argument?

 

[Sazar]

Originally posted by: gorillaman
Can we please see your review Sazar?
I'm dying to see what you came up with..

Thanks in advance,

GM

I know what you are trying to say but look @ it this way....

I do not run a site that releases reviews gorillaman... I don't get hits/ad revenue based on traffic nor do I have the same obligation to consumers looking for information to base their purchases on as AT does with the reviews...

it is the responsibility of whoever does the review to do it as thorough and balanced as possible...

heck I only started coming to AT in the first place due to the quality of their reviews... back in the day... its just something I expect...

when you start seeing the big three routinely stating things that look completely out of order... even v/s their OWN tabulated data... it does look odd... its almost as though the conclusion was written first and nothing was correlated @ the end of the day...

the conclusion was a little odd considering the tables as well as a few other quirks...

sort of right up there with tom's sitting describing the nv35/38 architecture as being 16 pipes

take it more as constructive criticism than slander...

 

[Sazar]

Originally posted by: VIAN
The review says 199 with rebate. Sounds pretty expensive, but still a nice gain.

I'm still bothered that ATI doesn't show all of the image and takes out some parts because they are barely noticeable. That could help a lot in performance. Thats their bandwidth saving techniques. Nvidia renders the entire image.

I was under the impression it was $199 retail... dunno bout the rebate...

thats a good price point and the card is shaping up quite nicely... a nice reworking of things by nvidia...

 

[Sazar]

Originally posted by: TheSnowman

Originally posted by: VIAN
I'm still bothered that ATI doesn't show all of the image and takes out some parts because they are barely noticeable. That could help a lot in performance. Thats their bandwidth saving techniques. Nvidia renders the entire image.

what evedence have you seen to back that argument?

tom's has pointed out some situations in his review...

lars makes it a point to note that some of ati's numbers should be taken with a grain of salt.. bear in mind there were no such disclaimer's when nvidia was pulling off their 3dmark03 and other optimizations so take that how you will...

as it stands... there may be a coupla things to check out.. AFAIK dave from b3d is doing some work checking out halo...

 

[kylebisme]

tom's article assumes that nvidia is rendering aquamark3 correctly in order to back their allegation that ati is cheating, and considering what we have seen from 3dmark and other applications that is hardly a valid assumption to make.

 

[reever]

Uhh go read Hardocp's review if you want to argue about filteirng issues. It is and has been a known issue that Nvidia's, not Ati's bri/tri filtering method is the one causing lower filtering and noticeable mip-map transitions. Now if only the geniuses at anandtech would have included fullscreen uncompressed images in their last IQ comparison, all of you would have seen that too. But I guess they felt it was necessary to shield your eyes from the real issue and instead cram BS down your throat and expect you to believe what they say with 0 proof whatsoever

 

Originally posted by: reever
Uhh go read Hardocp's review if you want to argue about filteirng issues. It is and has been a known issue that Nvidia's, not Ati's bri/tri filtering method is the one causing lower filtering and noticeable mip-map transitions. Now if only the geniuses at anandtech would have included fullscreen uncompressed images in their last IQ comparison, all of you would have seen that too. But I guess they felt it was necessary to shield your eyes from the real issue and instead cram BS down your throat and expect you to believe what they say with 0 proof whatsoever

Umm dude? Anand stated in the 5700 review that they are doing a separate IQ comparison and they are working on it right now. Will be posted up on the site when it is finished. It was also said that they want to do the best job possible with all of the facts.

Here is the quote from the article:
"We are currently working on an entire article devoted to the image quality of current generation graphics cards. Where there are important or overt visual anomalies, we will note them here. Other than that, our IQ judgments will be compiled into our coming article. We attempt to be as thorough as possible and delve into as many aspects of image quality as we can. Stay tuned, as its stacking up to be very interesting."

So I guess you'll just have to give em a chance. And then you might get the anwers that you really want to hear. You know,
like (ATI has much better image quality than nvidia because....) or (Nvidia has caught up to ATI on the image quality because...).

I dont think there is any reason to think there is a conspiracy of "eye shielding" going on. LOL.

GM

 

[reever]

I was talking about the IQ comparison done in the high-end shootout review, hence why i said "last IQ comparison"

Talk to me when anand re-edits that review with fullscreen uncompressed jpg's that actually include the ground portion of the games

 

[ronnn]

Originally posted by: Sazar
just a thought about the review...

after all the publicity webwide concerning the quality of the last couple of 'reviews' done here... would it not be prudent to change some benchmarking policies to eliminate doubt and have SOME accountability...

when weighing the benchmarks and considering all factors... there is nothing like the TKO that the reviewer notes in the conclusion...

as one of the 'big three' perhaps anandtech should look @ getting better reviewers... or better review guidelines...

I welcome your comments on this evan...

ta...

I agree, if I was in the market for a mid range card - this review would not really help me. Mind you the reviewer really reccomends not buying a midrange card and having a 9700pro, I agree with his reccomendation there.