290X OC Vs 7950 OC Clock For Clock Comparison

[tommybhoy]

290X Vs 7950 OC Clock For Clock Comparison@1080p

Performance comparison on everyday oc's that both can comfortably achieve:

Hitman AB-290X ~51% avg faster.

Sleeping Dogs-290X ~47% avg faster.

Tomb Raider-290X ~48% avg faster.

BioShock Infinite-290X ~29% avg faster.

 

[tommybhoy]

Updated op, forgot to state 1080p results.

You should also run this at 1440/1600

I can re do the tests after you send me the appropriate monitor.

Good work Tommy.

BTW very nice comparison

Thanks.

 

[tommybhoy]

To those querying the results, if ones low/high, the other is too as they were both done in the same system/driver.

H AB, SD and BI were on max settings, I can't remember if I changed TR settings from maxed out settings or used the TR BM settings(not maxed out) from the thread here.

Kind of makes CF 7950's pointless then doesn't it? I'm looking a 2nd to add to my rig but looking at these the 7950's seem well off the pace.

I don't think Matt for example found his CF7950's to be pointless.

Definitely not. 7950 crossfire was great bang for buck.

Spot on, I went from 7950 CrossFire to the 290(X) as I just got the new toy itch along with wanting to try out True Audio with Thief.

7950 CF@1100/1300MHz Score 12059
http://www.3dmark.com/3dm/1685603?

290X@1150/1450 Score 10744
http://www.3dmark.com/3dm/1693034?

12% slower on Firestrike on those clocks, but both could go faster with the 7950's pulling way ahead on higher clocks:

290X maxed out, SCORE-11136

7950 CrossFire maxed out-17038

 

[tommybhoy]

Good post tommy. Surprised the Ice q turbo's only hit 1100 core and 1450mem.

They didn't get 17038 points on 1100/1300Mhz, more like 1220/1600Mhz.

Can you go onto a 64 player paracel storm server and log your mins/mean fps when the storm kicks in. That map seems to give cpu's/gpu's a right arse kicking.

If I remember/have time.

I can tell you that my 290X@1100Mhz has lost about 25fps from my 7950's@1100Mhz on BF4 64 man DM in general (playing the same server that I use).