11 GTX480 vs. 5870 reviews in one chart (Updated with temps & OCs)

Tempered81

Diamond Member
Jan 29, 2007
6,374
1
81
105 benches, 11 reviewers, 4 cards, Avg FPS at 1920x1200 4xAA 16xAF:

GTX480: 83.66
HD5870: 73.23
GTX470: 68.70
HD5850: 62.25​

Thanks to MAJORD for making this:
gfxcomparerev2smallpng.jpg


xby34m.jpg


fullscreencapture331201.jpg

fullscreencapture331201.jpg

fullscreencapture331201.jpg

fullscreencapture331201.jpg

fullscreencapture331201k.jpg

fullscreencapture331201.jpg

fullscreencapture331201.jpg


15ez287.jpg
 
Last edited:

Lonyo

Lifer
Aug 10, 2002
21,938
6
81
Wow in some games it was 30-40% faster.

Given that it has 50% more transistors and a 50% wider memory bus, 50% higher power consumption, 30~40% faster in only some games is pretty bad. 30~40% across the board would have been impressive. 10~15% when they have gone all out isn't.

It's still not too bad price wise compared to the inflated ($420) current price of the HD5870, since it gives 10~15% more performance for only 20% more cost, which at the high end is pretty much a bargain, but considering what it is, and how late it is, it's underwhelming, just as was predicted by certain people who everyone likes to discredit.

The fact that it loses in some tests is more surprising than the fact that it wins by 40% in some, given the difference in design approach.

Also 2560x1600 numbers would be very interesting, considering these are $400+ cards and you would think at least some of the target market would be using 30" monitors at 2560x1600.
 

Genx87

Lifer
Apr 8, 2002
41,091
513
126
Given that it has 50% more transistors and a 50% wider memory bus, 50% higher power consumption, 30~40% faster in only some games is pretty bad. 30~40% across the board would have been impressive. 10~15% when they have gone all out isn't.

It's still not too bad price wise compared to the inflated ($420) current price of the HD5870, since it gives 10~15% more performance for only 20% more cost, which at the high end is pretty much a bargain, but considering what it is, and how late it is, it's underwhelming, just as was predicted by certain people who everyone likes to discredit.

The fact that it loses in some tests is more surprising than the fact that it wins by 40% in some, given the difference in design approach.

Also 2560x1600 numbers would be very interesting, considering these are $400+ cards and you would think at least some of the target market would be using 30" monitors at 2560x1600.

A lot of those transistors are geared for HPC, not gaming. So it isnt surprising it doesnt scale 100%.
 

ugaboga232

Member
Sep 23, 2009
144
0
0
See, I can do that too. :sneaky: :rolleyes:

I believe this is essentially what we were expecting, wasn't it? About 5-10% performance difference across most titles.

The problem becomes, while 5850's and 5870's can oc really high rather easily (to 1100 each on the core), these review 480's can oc about 15% (for comparison, 1100 for 5870 is 30%+). All said and done, with max OC's on both, the 5870 is = in performance. It is also cooler and uses less power even at such extreme oc's.

5850 is the best bet though. Nvidia didn't do too badly. Just wish they got the 30-40% performance level so we could get real competition.
 

ugaboga232

Member
Sep 23, 2009
144
0
0
A lot of those transistors are geared for HPC, not gaming. So it isnt surprising it doesnt scale 100%.

On the AT review, it barely gets over 2X performance of the 285 in PhysX. It isn't faster in HPC. It has a lot of geometry and tessellation power, but it hasn't really changed its core too much.
 

reallyscrued

Platinum Member
Jul 28, 2004
2,618
5
81
Some of these benchmarks actually surprise me. It looks like 1.5 GB is the new sweet spot for video memory.

So does this mean ATi street prices will come back down to MSRP?

The 5870 needs more VRAM and a 10% overclock, shouldn't be too difficult as they run pretty cool (At least cooler than my 4890.)
 

dug777

Lifer
Oct 13, 2004
24,778
4
0
A lot of those transistors are geared for HPC, not gaming. So it isnt surprising it doesnt scale 100%.

That's very nice I'm sure, but 100% irrelevant to people who play video games, particularly as I understand they've majorly gimped all that for the 'gaming' cards ;)

As it is, it then appears to be deadweight, and damn power-hungry deadweight at that...if gaming is what you are after. If not, then all power to you :)

EDIT: Perhaps more importantly, do you think I could run a GTX 470 (or a 5870) on a reasonably decent 500W PSU? Q6600@3.4, 4GB DDR2, 1 HDD, 1 optical drive is the rest of the system, in an Antec Sonata III with the stock 500W PSU (Earthwatts 500W with 2x 17A 12v rails).
 
Last edited:

Genx87

Lifer
Apr 8, 2002
41,091
513
126
That's very nice I'm sure, but 100% irrelevant to people who play video games, particularly as I understand they've majorly gimped all that for the 'gaming' cards ;)

As it is, it then appears to be deadweight, and damn power-hungry deadweight at that...

Of course it is irrelevant to gamers. But it is the reason why it has more transistors but isnt blowing the doors off the 5870 in "games".
 

sciwizam

Golden Member
Oct 22, 2004
1,953
0
0
That's very nice I'm sure, but 100% irrelevant to people who play video games, particularly as I understand they've majorly gimped all that for the 'gaming' cards ;)

As it is, it then appears to be deadweight, and damn power-hungry deadweight at that...

EDIT: Perhaps more importantly, do you think I could run a GTX 470 on a reasonably decent 500W PSU? Q6600@3.4, 4GB DDR2, 1 HDD, 1 optical drive is the rest of the system, in an Antec Sonata III with the stock 500W PSU (Earthwatts 500W with 2x 17A 12v rails).

You don't have to understand French, but take a look at these temp. readings with that same case. They have an ambient temp. of 26C. I'm pretty sure it gets hotter than that in Australia. :D
 

ugaboga232

Member
Sep 23, 2009
144
0
0
You got me there. I didn't read that page. Looks Fermi does have some serious power. That is good. Maybe a Fermi refresh will be competitive with the 6XXX series.

Gaming-wise, I still think a 5850 is the best bet.
 

n7

Elite Member
Jan 4, 2004
21,281
4
81
Has anyone found a single review in which Borderlands is compared?

I know reviewers love to all use the same new games always :rolleyes:, but some older titles like that would be nice.

Anyway, i'm particularly curious how it does.

I know it's UE3, so i can infer results from other UE3 titles, but i'd still like to see actual numbers.
 

MadJackalIto

Junior Member
Mar 27, 2010
19
0
0
wow GTX480 is an amazing VGA Card... but in price/performance?
20% better performance in some games...awesome.
 

Qbah

Diamond Member
Oct 18, 2005
3,754
10
81
Has anyone found a single review in which Borderlands is compared?

I know reviewers love to all use the same new games always :rolleyes:, but some older titles like that would be nice.

Anyway, i'm particularly curious how it does.

I know it's UE3, so i can infer results from other UE3 titles, but i'd still like to see actual numbers.

http://pclab.pl/art41241-13.html

:) It's in Polish, but you can read the charts ;)
 

Tempered81

Diamond Member
Jan 29, 2007
6,374
1
81
Thanx!

GTX 480 is way better w/o AA, but slightly worse with. :\

I wish AA didn't tank fps so bad in that game.

Borderlands is great. I'm currently running at 1024x768 @ 70hz V-synced and maxed at 70fps on an overclocked 4890. N7, are you an avid Borderlands player? What level are you?
 

imaheadcase

Diamond Member
May 9, 2005
3,850
7
76
Most don't play with AA anyways, so really a moot point. I think it was firingsquad or maybe some other site that had a poll that asked what level of AA you use, and little more than half did not use it, and rest was 1x 2x. Like %5 said "What is aa?". lol