• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

5870+9600GT to Gigabyte 480 SOC results

straxusii

straxusii

straxusii

Soldato
Joined
18 May 2003
Posts
4,884
Had my 5870 since launch and ran it stable at 900/1250, great card. I also had a 9600GT which I used for Physx

Here are some of the results of my move to a 480, running at 860/2060, all tests at 1920x1200 apart from Heaven 2.5

Mafia 2 - all settings max, AA, high physx
5870 + 9600GT – 32.2fps
480 – 33.5fps
480 + 9600GT – 35.8fps

I note in this demo that the 9600GT peaked at about 60% usage with the 480, 480 wasnt maxed either at about 80% average usage.


Metro 2033 - DX11, all settings maxed except detail on high, not very high, and DOF off
Didnt run this on the 5870
480 - 41.96
480 + 9600GT - 43.39

The 9600GT was hardly used, got up to about 25% usage.


Crysis warhead - DX10, enthusiast, 0xAA, frost
5870 – 38.38
480 – 42.54


Heaven 2.5, defaults apart from 4xAA, 1680x1050
5870 - 34.4 (867)
480 - 45.8 (1153)


3d mark 11
5870 - P4446
480 – P5518

In graphics test 4 I note the 480 is around 70% faster


Overall I'm going to keep the card over my 5870 – just. It is quieter than my 5870, I won't be keeping the 9600GT as a dedicated physx card though, it increases the temps of the 480 too much.

As an upgrade from anything less than a 5850 in power this is a great option at £240.




Edit: Just cause 2 1920x1200, dark tower, all settings max/on 4xAA 16xAF, bokeh & gpu water off
Found an old forum post with my 5870 result
5870 : 52
480 : 70.59
 
Last edited:
Canned benchmarks only tell half the story, going from a 1000 core 5870 to an 850 core 480 gives a massive leap in min FPS and average FPS in most games, gameplay is much smoother.
 
Last edited:
Freddie1980 said:
So going on his gaming results he gained 11% in Crysis. *UPGRADE WIN*

The only thing that's "FAIL" is your arithmetic. :o
Click to expand...

but really, would you notice 3-4FPS?

i guess it will make everything smoother so is a worthwhile upgrade, but to someone looking at those benches, the gains dont look all that.

although i do agree with you that an 11% increase is well worth the money, 3-4FPS does not sound so great, which i think is the point he was making.
 
Off the top of my head those 480 results look low, not much higher than what I was getting with a stock GTX470 IIRC (tho that was 1920x1080 not 1920x1200).
 
Freddie1980 said:
So going on his gaming results he gained 11% in Crysis. *UPGRADE WIN*

The only thing that's "FAIL" is your arithmetic. :o
Click to expand...

Congratulations on learning how to work out a percentage!

* Double fail for calling an incorrect mistake!!! mafia was 3fps difference and cysis was 4fps ! Thats why i stated he had gained 3-4fps LOSER*
 
Based on the crysis and mafia results in isolation I'd agree it's an upgrade fail.

3dmark and heaven don't count as clearly they don't represent most games.

I wished I'd benched Metro on the 5870 as almost certainly that would have demonstrated a large improvement.

Those 2 were just the 2 games I had to hand, there's plenty of other games out there that if I had to hand would show a good 20-30% boost.


However the proof of the pudding is the game I play most is TF2 and with the 480 I can now enable 8xAA and get no slowdown in mental firefights.


If you look at the TPUP review you see at 1920x1200 the summary shows a gain of 28% stock 5870 to stock 480 SOC http://www.techpowerup.com/reviews/Gigabyte/GeForce_GTX_480_SOC/27.html

This doesnt tell the whole story though as it contains some cpu bound bechmarks and for some reason they are not benching dirt 2 with DX11.


The game results alone give:

AVP +35%
BFBC2 +11%
Battleforge +55%
COD4 +23%
COJ2 +42%
Crysis +10%
DOW2 +17%
Dirt2 (dx9) +8% (for reference anand DX11 dirt2 shows a stock 480 at +23% over 5870)
FC2 +62%
HAWX +58%
Metro +50%
Riddick +16%
StalkerCS +29%
SupCom2 +11%
UT3 +33%
WOW +32%
 
Last edited:
paradisiac said:
but really, would you notice 3-4FPS?

i guess it will make everything smoother so is a worthwhile upgrade, but to someone looking at those benches, the gains dont look all that.

although i do agree with you that an 11% increase is well worth the money, 3-4FPS does not sound so great, which i think is the point he was making.
Click to expand...

If it were 3-4 FPS gain when your already 50+ FPS then no your not going to see but when it comes to Crysis I'll take anything especailly at that range.

jej_knowles said:
Congratulations on learning how to work out a percentage!

* Double fail for calling an incorrect mistake!!! mafia was 3fps difference and cysis was 4fps ! Thats why i stated he had gained 3-4fps LOSER*
Click to expand...

Oh dear, one doesn't take criticism very well do they.

And yes I can see the performance difference in the two games by deducting the two numbers in isolation but you failed to see what that meant in reality as you looked at the absolute numbers and dismissed them completely.
 
Last edited:
hmm......im lost now. the guy has upgraded and benched games he plays, and is happy with the results, yet now people are saying they dont play those games so the gains arent important/relevant :confused: but is the OP not stating that his new purchase is good value for money, not telling everyone to sell their cards and go out and buy one of these.
 
It's not all about the max fps. I went from a 5850 clocked at 1000/1250 and gameplay was getting some stuttering so i got the Asus GTX 480 when it was on special and even at stock the gameplay is smoother and much more enjoyable. When i benched them the minimum fps was way higher with the 480 than the 5850. Take Stalker Clear Sky for example:-

Min/Max/Avg fps. All setting to max @1920x1200.

Day
5850@1000/1250 28.7/123.1/58.1
GTX 480 @stock 45.6/233.2/156.0

Night
5850@1000/1250 38.9/144.0/81.6
GTX 480 @stock 63.1/244.8/149.4

Rain
5850@1000/1250 44.6/156.3/95.5
GTX 480 @stock 65.5/248.5/161.8

Sun
5850@1000/1250 25.5/91.7/50.7
GTX 480 @stock 40.0/154.3/96.4

Some huge gains there. Forget about the max and look at the min and avg fps. The 480 is a cracking card and is so much smoother than the 5850.
 
You must log in or register to reply here.
Back
Top Bottom