NVIDIA 9800GTX+ Review Thread

[Sind]

Man I am loving all this competition and them trying to one up each other. Please keep it up! It's been sorrily missed.

 

[tuteja1986]

Originally posted by: Zaitsev

Originally posted by: Allio
That firingsquad review is weird, it shows the normal 9800GTX equal to or beating the 4850 in most tests which is nothing like what other reviews have shown. Compare Bioshock results from firingsquad and anandtech...

That is weird. The firingsquad setup is almost 13fps slower at the same resolution. I wonder if it has to do with anand using cat 8.5 and firingsquad 8.6 beta. You wouldn't think newer drivers would make games slower though...

Compare the call of duty result of techreport with other. These review are questionable. In less than 10hrs firingsquad benchmarked ten games and cliamed 9800GTX+ the winner.

None of them actually did benchmarking with 8xAA and some not even 4xAA. these are questionable since other site just received a 9800GTX+ just in less than 20hrs ago.

 

[BTRY B 529th FA BN]

I think one of these companies is distorting the view and making price adjustments to put themselves on the same playing field when they had two releases to the others one release.

 

[Wreckem]

Originally posted by: Zaitsev

Originally posted by: Allio
That firingsquad review is weird, it shows the normal 9800GTX equal to or beating the 4850 in most tests which is nothing like what other reviews have shown. Compare Bioshock results from firingsquad and anandtech...

That is weird. The firingsquad setup is almost 13fps slower at the same resolution. I wonder if it has to do with anand using cat 8.5 and firingsquad 8.6 beta. You wouldn't think newer drivers would make games slower though...

Well Anandtech got lower 9800GTX scores in BioShock compared to other reviews as well. Avg's around 10 lower than most other benchmarks. Its hard to compare reviews, because setups and settings can be different.

 

[Wreckem]

Originally posted by: tuteja1986

Originally posted by: Zaitsev

Originally posted by: Allio
That firingsquad review is weird, it shows the normal 9800GTX equal to or beating the 4850 in most tests which is nothing like what other reviews have shown. Compare Bioshock results from firingsquad and anandtech...

That is weird. The firingsquad setup is almost 13fps slower at the same resolution. I wonder if it has to do with anand using cat 8.5 and firingsquad 8.6 beta. You wouldn't think newer drivers would make games slower though...

Compare the call of duty result of techreport with other. These review are questionable. In less than 10hrs firingsquad benchmarked ten games and cliamed 9800GTX+ the winner.

None of them actually did benchmarking with 8xAA and some not even 4xAA. these are questionable since other site just received a 9800GTX+ just in less than 20hrs ago.

It not a review. Its a "quick" take.

 

[Wreckem]

Originally posted by: BTRY B 529th FA BN
I think one of these companies is distorting the view and making price adjustments to put themselves on the same playing field when they had two releases to the others one release.

Its a smart business move on Nvidia's part. They are now competitive with ATI when it comes to price/preformance.

Anyone that though Nvidia wouldnt adjust its prices across the board lacks all knowledge of business. And rumors of the 55nm 9800 have been floating around for awhile.

I dont know why people are f'ing bitching. Nvidia and ATI are engaged in an all out slugfest, and the consumer wins with better preformance and cheaper cards.

 

[AzN]

Originally posted by: Allio
That firingsquad review is weird, it shows the normal 9800GTX equal to or beating the 4850 in most tests which is nothing like what other reviews have shown. Compare Bioshock results from firingsquad and anandtech...

Firingsquad has always been a Nvidia sponsored site. Looks like legit review is too. They are the only ones who got to play 9800gtx+ first and nobody else has them. Some sites just aren't trust worthy.

 

[tuteja1986]

http://www.computerbase.de/art...v770/#abschnitt_prolog


Unreal Tournament 3 1600x1200 4xAA/16xAF:
GTX 280 : 145
HD4850 : 65
9800GTX : 81

Unreal Tournament 3 1600x1200 8xAA/16xAF:
GTX 280 : 72
HD4850 : 63
9800GTX : 34

Lost planet 1600x1200 4xAA/16xAF:
GTX 280 : 58
HD4850 : 31
9800GTX : 34

Lost planet 1600x1200 8xAA/16xAF:
GTX 280 : 45
HD4850 : 29
9800GTX : 24


Company of heroes 1600x1200 8xAA/16xAF
GTX 280 : 84
HD4850 : 46
9800GTX : 47

Company of heroes 1600x1200 8xAA/16xAF
GTX 280 : 66
HD4850 : 42
9800GTX : 35


No wonder Nvidia is happy that everyone benchmark games in 4xAA , look at huge drop for nvidia card soon you go to 8xAA. ATI only looses few FPS :!

 

[Wreckem]

Originally posted by: tuteja1986
http://www.computerbase.de/art...v770/#abschnitt_prolog


Unreal Tournament 3 1600x1200 4xAA/16xAF:
GTX 280 : 145
HD4850 : 65
9800GTX : 81

Unreal Tournament 3 1600x1200 8xAA/16xAF:
GTX 280 : 72
HD4850 : 63
9800GTX : 34

Lost planet 1600x1200 4xAA/16xAF:
GTX 280 : 58
HD4850 : 31
9800GTX : 34

Lost planet 1600x1200 8xAA/16xAF:
GTX 280 : 45
HD4850 : 29
9800GTX : 24


Company of heroes 1600x1200 8xAA/16xAF
GTX 280 : 84
HD4850 : 46
9800GTX : 47

Company of heroes 1600x1200 8xAA/16xAF
GTX 280 : 66
HD4850 : 42
9800GTX : 35


No wonder Nvidia is happy that everyone benchmark games in 4xAA , look at huge drop for nvidia card soon you go to 8xAA. ATI only looses few FPS :!

Just looking at all the 9800GTX(non +) benchmarks and all the 4850 benchmarks.

Avg it out, and the 9800GTX beats out the 4850 in about half of the benchmarks when AA isnt enabled. They are neck and neck at 4xAA, with the slight edge going to the 4850, and the 9800GTX loses horribly at 8xAA.

Its pretty clear that ATI is going to win the battle when AA is involved, and its likely because they have likely spent more time making more efficent AA, a side effect of being DX10.1 compatible.

 

[Cookie Monster]

Um i dont think most nVIDIA users will be using the traditional 8xAA since theres another alternative ala 16xCSAA which is much better.

edit - or 16xQ CSAA for that matter.

 

[AzN]

Originally posted by: Wreckem

Originally posted by: tuteja1986
http://www.computerbase.de/art...v770/#abschnitt_prolog


Unreal Tournament 3 1600x1200 4xAA/16xAF:
GTX 280 : 145
HD4850 : 65
9800GTX : 81

Unreal Tournament 3 1600x1200 8xAA/16xAF:
GTX 280 : 72
HD4850 : 63
9800GTX : 34

Lost planet 1600x1200 4xAA/16xAF:
GTX 280 : 58
HD4850 : 31
9800GTX : 34

Lost planet 1600x1200 8xAA/16xAF:
GTX 280 : 45
HD4850 : 29
9800GTX : 24


Company of heroes 1600x1200 8xAA/16xAF
GTX 280 : 84
HD4850 : 46
9800GTX : 47

Company of heroes 1600x1200 8xAA/16xAF
GTX 280 : 66
HD4850 : 42
9800GTX : 35


No wonder Nvidia is happy that everyone benchmark games in 4xAA , look at huge drop for nvidia card soon you go to 8xAA. ATI only looses few FPS :!

Just looking at all the 9800GTX(non +) benchmarks and all the 4850 benchmarks.

Avg it out, and the 9800GTX beats out the 4850 in about half of the benchmarks when AA isnt enabled. They are neck and neck at 4xAA, with the slight edge going to the 4850, and the 9800GTX loses horribly at 8xAA.

Not really. In raw performance they are 1 or 2% apart even in lower resolution. With AA 4850 has the clear advantage especially @ 1600x1200 and up resolution or 8xAA.

Performance average by resolution and filters
http://www.computerbase.de/art...rmancerating_qualitaet

 

[Mwing]

9800gtx+ has a better cooler, but is it longer than 4850?

 

[AzN]

Originally posted by: Cookie Monster
Um i dont think most nVIDIA users will be using the traditional 8xAA since theres another alternative ala 16xCSAA which is much better.

edit - or 16xQ CSAA for that matter.

Actually 8x CSAA is lesser than Q. CSAA is performance advantages but traditional 8xMSAA is called Q which has smoother edges.

 

[Cookie Monster]

Um, i meant 16xCSAA not 8xCSAA (i have no idea why one could consider 8xCSAA when 16xCSAA does it better and both have similar performance drop). And can you show me how the 8xQ is superior than 16xCSAA? most users point to the latter being superior in both IQ AND performance.

 

[AzN]

Regular CSAA takes a less of a nose dive than regular MSAA Q but it really doesn't improve image either. It's only about 10% difference in performance from 4xAA. Regular CSAA is bandwidth saving using less samples.

 

[AnandThenMan]

Firingsquad= Nvidia shill. Don't trust any results from them they make sure Nvidia outshines ATI regardless.

 

[Cookie Monster]

Yes and in the past they were regarded as one of the top reviewing sites.

I think people should just look at the overall picture instead of making this accusations. Different sites use different methods of benchmarking with different software environment and hardware setups i.e will result in different outcomes. (Well variances are a given but outliers can be suspicious). These reviews all give a wider perspective for the consumer to then make a decision. I guess products are just really competitive nowadays for a consumer to even remotely see a black/white difference.

On a personal note, I never thought FS would be deemed "nVIDIA shill". Quite disturbing how times have changed.

 

[tuteja1986]

Originally posted by: Rilo
Apparently firingsquad.com got a hold of one of these boards to post some numbers:

http://www.firingsquad.com/har...850_geforce_9800_gtx+/

well , firingsquad review is questionable. Also looks the review was rushed out. They also some how managed to review another card less than a day. They failed to do 8xAA benchmark which they use to do 2 years ago and were very respected site. This review benchmark does not match the general community finding. Disappointing review that suffers from being rushing it out. Also the 9800GTX+ is not even out and will be out untill next month and the way this review shows 8900GTX+ beating up a 4850 would make this card cut into nvidia GT260 sales which nvidia would not want to happen. Why would they want to kill GT260 sales by releasing a $230 card.

http://www.computerbase.de/art...tt_unreal_tournament_3

Unreal Tournament 3 1600x1200 4xAA/16xAF:
GTX 280 : 145
HD4850 : 65
9800GTX : 81

Unreal Tournament 3 1600x1200 8xAA/16xAF:
GTX 280 : 72
HD4850 : 63
9800GTX : 34

---------------------------------------------------------------
http://www.computerbase.de/art..._clive_barkers_jericho

Clive Barker's Jericho 1600x12004xAA/16xAF:
GTX 280 : 53
HD4850 : 29
9800GTX : 25

Clive Barker's Jericho 1600x1200 8xAA/16xAF:
GTX 280 : 26
HD4850 : 28
9800GTX : 7


---------------------------------------------------------------
http://www.computerbase.de/art...v770/8/#abschnitt_fear

FEAR 1600x1200 4xAA/16xAF:
GTX 280 : 147
HD4850 : 92
9800GTX : 83

FEAR 1600x1200 8xAA/16xAF:
GTX 280 : 93
HD4850 : 60
9800GTX : 49


---------------------------------------------------------------
http://www.computerbase.de/art...schnitt_call_of_juarez

Call of Juarez 1600x1200 8xAA/16xAF
GTX 280 : 30
HD4850 : 21
9800GTX : 20

Call of Juarez 1600x1200 8xAA/16xAF
GTX 280 : 24
HD4850 : 17
9800GTX : 6


---------------------------------------------------------------
http://www.computerbase.de/art...nitt_company_of_heroes

Company of heroes 1600x1200 4xAA/16xAF
GTX 280 : 84
HD4850 : 46
9800GTX : 47

Company of heroes 1600x1200 8xAA/16xAF
GTX 280 : 66
HD4850 : 42
9800GTX : 35

---------------------------------------------------------------
http://www.computerbase.de/art...#abschnitt_lost_planet

Lost planet 1600x1200 4xAA/16xAF:
GTX 280 : 58
HD4850 : 31
9800GTX : 34

Lost planet 1600x1200 8xAA/16xAF:
GTX 280 : 45
HD4850 : 29
9800GTX : 24

Now i am looking for more review that has done 8xAA to confirm these sort of result.

 

[BFG10K]

Regular CSAA takes a less of a nose dive than regular MSAA Q but it really doesn't improve image either.

That just isn't true with respect to image quality.

One of the places that have the most noticeable aliasing is first-class polygon edges and there CSAA has almost the same IQ as the equivalent MSAA level.

8xQ is more robust (and overall better) than 16xCSAA but the effects of CSAA should not be downplayed.

 

[AzN]

Originally posted by: BFG10K

Regular CSAA takes a less of a nose dive than regular MSAA Q but it really doesn't improve image either.

That just isn't true with respect to image quality.

One of the places that have the most noticeable aliasing is first-class polygon edges and there CSAA has almost the same IQ as the equivalent MSAA level.

8xQ is more robust (and overall better) than 16xCSAA but the effects of CSAA should not be downplayed.

It's just coming from my experience when I play a game. It does ever slightly improve image but nearly as much as 8xQ.

 

[BFG10K]

It's just coming from my experience when I play a game. It does ever slightly improve image but nearly as much as 8xQ.

Fair enough; I personally find 16xCSAA is a great in-between mode when I have power to burn @ 4xAA but not enough juice to do 8xQ.

Now 8x, that's generally useless as every test I've tried 16xCSAA is literally only 1-2 FPS slower. But that's just a testament to how good 16xCSAA is rather than a failure on nVidia?s part.

 

[AzN]

Don't get me wrong I use it all the time when my card isn't fast enough to use 8xQ. It's better than 4xAA and very low performance drop.

 

[Keysplayr]

Originally posted by: munky

Originally posted by: Wreckage

Originally posted by: Spartan Niner


You do realize that AMD has made their cards on 55nm since the 38xx series, right? From the AnandTech preview article:

You do realize that the 4850 runs 25% hotter than a standard GTX right?

So, I'm not sure what your point is.

I also realize the 9800gtx runs 10dB louder, which puts the whole cooler/quieter thing in a different perspective.

If you ran the 4850's fan where it's supposed to be to adequately cool it, i guarantee that would change. According to the article you linked to, how high would you need to run the 4850 fan to get it from 68C to 48.7C. A 20C drop. Check the noise then.

They should have just made the reference cards with Dual slot coolers. Charge 15 bux more. Big deal. That would have been preferable to a single slot cooler that can barely do the job. And look how small that fan is. That annoys me. I felt the same way about the 8800GT's with that single slot cooler and super tiny fan.

 

[bryanW1995]

Originally posted by: DaveSimmons
No power usage or noise testing :| .

For all we know they just overclocked a regulat 9800GTX to GTX+ speeds.

Anand talked about the 9800gtx+ in his 4850 review. He explicitly stated that it was 55 nm. It IS clocked at exactly the same speed as one of the evga 9800gtx OC models, however...

 

[bryanW1995]

Originally posted by: keysplayr2003

Originally posted by: munky

Originally posted by: Wreckage

Originally posted by: Spartan Niner


You do realize that AMD has made their cards on 55nm since the 38xx series, right? From the AnandTech preview article:

You do realize that the 4850 runs 25% hotter than a standard GTX right?

So, I'm not sure what your point is.

I also realize the 9800gtx runs 10dB louder, which puts the whole cooler/quieter thing in a different perspective.

If you ran the 4850's fan where it's supposed to be to adequately cool it, i guarantee that would change. According to the article you linked to, how high would you need to run the 4850 fan to get it from 68C to 48.7C. A 20C drop. Check the noise then.

They should have just made the reference cards with Dual slot coolers. Charge 15 bux more. Big deal. That would have been preferable to a single slot cooler that can barely do the job. And look how small that fan is. That annoys me. I felt the same way about the 8800GT's with that single slot cooler and super tiny fan.

When I fold on my 3870 I have to bump my fan up from 37% (80c +) to 42 % (69c). you should soon (if not now) be able to manually set your 4850 fan in rivatuner to an acceptable level with something significantly less than a 10db penalty.

having said that, we definitely agree on our feelings about that horrible fan on both 4850 and 8800gt. :thumbsdown: to both teams on that call :|