Are these benchies on par (HIS 6950 2GB)

Merlin5 · 13 Feb 2011 at 03:12

I just did two benchmarks on the Dirt 2 demo. With 4 x AA, everything on high (except post processing which only goes to medium with my card), I got 52.9fps average and 43.1fps minimum.

Then, with everything set to ultra and post processing still on medium, I only dropped 5fps with 47.6fps avg and 37.9fps minimum. This is with my sig rig. I could not discern any difference whatsoever in smoothness or graphical quality from high to ultra, the demo looked great both times. So frankly, keeping everything to high and not ultra would be fine if I was actually interested in playing Dirt 2.

Anyway, I spoke to phil2008. I asked him to do another bench with the same ultra settings, and to take his post processing down to medium to match mine.

His came out with 61fps average, 44 fps minimum. His screenshot above shows 59.6fps avg with ultra settings and high post processing. So he only gained less than 2fps average by reducing post processing from high to medium.

Considering I have 4gb ram, exactly half his amount, a 3.1ghz dual core cpu, and only a 4870 1gb, shouldn't his card pull away from mine even more with higher framerates?

metalmackey · 13 Feb 2011 at 12:43

Merlin5: You cant compare your card to Phil's at Ultra settings as his will be in DX11 mode and yours isn't.

Merlin5 · 13 Feb 2011 at 13:06

That's a good point metalmackey.

metalmackey · 13 Feb 2011 at 13:24

When I first got my 5970, I had it in a QX6850 system and was a bit dissapointed with the perfomance, and that was with a quad core. I think some games like the extra mem bandwidth of DDR3 and the other improvements in newer CPU's. As soon as I got the I7, the 5970 showed what a good card it was (and still is LOL).

Merlin5 · 13 Feb 2011 at 14:10

metalmackey said:
When I first got my 5970, I had it in a QX6850 system and was a bit dissapointed with the perfomance, and that was with a quad core. I think some games like the extra mem bandwidth of DDR3 and the other improvements in newer CPU's. As soon as I got the I7, the 5970 showed what a good card it was (and still is LOL).

So if phil had the same i7 920 cpu as in the vortez test setup, he would achieve, give or take, around that 79.6fps average? The only reason he gets 59.6 now is due to his dual core processor?

metalmackey · 13 Feb 2011 at 14:26

Yeah I'm guessing he'd get near that with an I7 or even an I5. More and more games are taking advantage of quads now as they become the norm.

tommybhoy · 13 Feb 2011 at 14:43

[email protected] 8g ram, 6950>70
Same settings as yours(max everything with with 4xaa @ 1920x1080):

This however is with power control settings at +20%.

With power control settings at 0, I was getting:

Big difference in minimums! :eek:

You could try and see if the power control settings at +20% makes a difference for you.

metalmackey · 13 Feb 2011 at 14:51

Remember the benchmark is different every time so will get slightly different results each time.

Phil2008 · 13 Feb 2011 at 14:58

Thanks guys, its like when I first bought my 3870, it was performing pretty badly in games and I didnt really like the 3870 at all. But once I built my wolfy system, coming from a amd opteron 185 system, the 3870 came alive and Ive been happy with the 3870 ever since

Im more then happy with the 6950s performance with my wolfy system as theres no slow/chuggyness in games at all, and thats the reason I upgraded. All this benchmarking lark will stop now as I know my 6950 is performing as it should be, and when I upgrade cpu,mem and board in a few years, it will release the extra fps from my 6950 and max it out.

Something else Ive notice but I dont know why it should make a diff, but dirt2 levels are loading a much much quicker with the 6950 :confused:

Phil2008 · 13 Feb 2011 at 15:09

tommybhoy said:
[email protected] 8g ram, 6950>70
Same settings as yours(max everything with with 4xaa @ 1920x1080):

This however is with power control settings at +20%.

With power control settings at 0, I was getting:

Big difference in minimums!

You could try and see if the power control settings at +20% makes a difference for you.

Thanks. so its deffo the cpu thats letting the side down in dirt 2. To be honest Im not that fussed, like my saying goes, just as long as games run nice and smoothly, who cares about the fps

tommybhoy · 13 Feb 2011 at 16:59

Phil2008 said:
To be honest Im not that fussed, like my saying goes, just as long as games run nice and smoothly, who cares about the fps

That's the main thing at the end of the day.

TheRealDeal · 13 Feb 2011 at 17:13

tommybhoy said:
[email protected] 8g ram, 6950>70
Same settings as yours(max everything with with 4xaa @ 1920x1080):

This however is with power control settings at +20%.

With power control settings at 0, I was getting:

Big difference in minimums!

You could try and see if the power control settings at +20% makes a difference for you.

Your q6600 is holding back your gpu as my 6870 gets better results at 1920x1200 with my rig. I usually get 73-74 average and around the same minimum as you.

tommybhoy · 13 Feb 2011 at 19:22

TheRealDeal said:
Your q6600 is holding back your gpu as my 6870 gets better results at 1920x1200 with my rig. I usually get 73-74 average and around the same minimum as you.

I know, I'm just waiting to see what bulldozer brings to the table before deciding whether or not to take the plunge.
Whatever gives me best bang for buck would be getting my cash.
On the other hand, I don't know if £400 is worth ~10fps when you get above 60 fps.

Phil2008 · 13 Feb 2011 at 20:07

Its like when my cpu,memory and board needs changing in 1-2yrs time, I will notice a huge diff in games and general use and my money will be well spent. But if I upgrade now I wont notice much diff at all.

Thats just how I look at it, I run all my hardware into the ground before I upgrade.

Rroff · 13 Feb 2011 at 20:14

metalmackey said:
When I first got my 5970, I had it in a QX6850 system and was a bit dissapointed with the perfomance, and that was with a quad core. I think some games like the extra mem bandwidth of DDR3 and the other improvements in newer CPU's. As soon as I got the I7, the 5970 showed what a good card it was (and still is LOL).

Was that with the CPU at stock? I get similiar fps with my overclocked Q9550 w/ GTX470 SLI as people with overclocked i7 setups with 470 SLI except in a very small number of games that can effectively use the HT units. Its certainly not holding me back.

metalmackey · 13 Feb 2011 at 20:26

Rroff said:
Was that with the CPU at stock? I get similiar fps with my overclocked Q9550 w/ GTX470 SLI as people with overclocked i7 setups with 470 SLI except in a very small number of games that can effectively use the HT units. Its certainly not holding me back.

No it wasn't overclocked but I think it was more to do with the crap chipset I had it on than the CPU speed itself. The O/C helps of course

chaparral · 13 Feb 2011 at 20:33

Rroff said:
Was that with the CPU at stock? I get similiar fps with my overclocked Q9550 w/ GTX470 SLI as people with overclocked i7 setups with 470 SLI except in a very small number of games that can effectively use the HT units. Its certainly not holding me back.

How many FPS do you get in dirt 2 at 1920x1200 4xAA max details ??

Rroff · 13 Feb 2011 at 20:44

Don't have dirt 2.

chaparral · 13 Feb 2011 at 20:46

Rroff said:
Don't have dirt 2.

That no good...

setter · 13 Feb 2011 at 20:53

With an i7 @4ghz and stock sli 470's i get the following in the dirt 2 bench from the demo

Average, 128 fps
Minium,103 fps

Res of 1920x1200, 8 x qcsaa all detail settings on highest, currently running the cards at stock as ive messed up the airflow in my case due to testing new case intake fans.