• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Slow? Radeon7970

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
What do the 7970 reviews at this site state for Vantage? Again noting that CPU does play a role so you should be in the ballpark of those scores.

I went to 3DMark.com Results and searched 7970 non sli/crossfire but couldn't specifically find a i7 930..The results listed..and there is an endless number are mostly about extreme i7's I guess and similar.. Much faster than I'm using, but the numbers were in the 50K area for the 3DMark Vantage performance scores.

They have a tremendous variety of tests and results. Anyone interested could look around better than I could relate all here. Hopefully I'm reading their scores correctly, but new to their performance reviews. Far from knowledgeable.

http://3dmark.com/search
 
50k is on extreme cooling.

Look at the reviews HERE for the 7970's. We tested on 2600k with the GPU at stock speeds.
 
id say stock 7970 on your cpu .. anywhere from 20-30k should be normal.
 
50k is on extreme cooling.

Look at the reviews HERE for the 7970's. We tested on 2600k with the GPU at stock speeds.

I wasn't suggesting that I have such an expectation but was rather reporting the results..thats all. As I mentioned I couldn't find reports at 3dMark for slower machines, but my P 8,000 score seems poor. And I hope to do better or learn why.
 
Im telling ya, 5870s still a great card, if you had trifire 5870s it will eat that card alive, even 2 5870s would but Im sure ill get a lot of heat from what i said, look at my score
Captue2.PNG

heres the res i run on my monitor
Unigine score1.PNG
 
Last edited:
Im telling ya, 5870s still a great card, if you had trifire 5870s it will eat that card alive, even 2 5870s would but Im sure ill get a lot of heat from what i said, look at my score
View attachment 109482

That's been my experience for sure. I had two of the 2gb units and had to return one under warranty because the fan (bushing? bearing?) whatever got wasted and the fan would scream when it really ramped up and they wouldn't fix it. They sent a me a 6950 that works less well and tests less well too haha. I complained and they just cited their original tests---the one they do that sets their benchmark for all 6950's--not the one they sent me:)

I just tested my 5870 and it scores significantly below the results I've been getting for my new 7970, "but" talk about real world apps...when i play the fast pace shooters...the 5870 "rocks". and the 7970.."not". haha.. what can I say..my 7970 scores P7991 3DMARK11 and my 5870 2gb scores P4669 3DMARK11. But I'd rather take the 5870 into MW3 for sure!!
 
That just makes no sense!!! Its like twice as fast!

Ya it doesn't sound right, but real world/real person testing is not what they do when they run these 3DMARK11 type scores..I would suppose.

I'm no expert on graphic scores but I know how to use graphics "hands on". And I guess the 3DMARK11 numbers are maybe kinda simplistic in that perhaps they're measuring things that some of us actually aren't using.

For instance in fast pace shooters one isn't standing around gazing at the physics of the dust cloud etc. Players are concerned with "seeing" the enemy better and quicker.

I suppose when gpu's are designed they must make trade offs and try to balance all. Just guessing. So I suppose some gpu's are better suited to some applications than others.

That's my speculation.
 
I just got this VisionTek Radeon7970 and its slow compared to my 5870. Not sure about the problem. And how do I mean slow? Its not like I'm running FRAPS and unfortunately I don't have an up to date software package from Futuremark--though I may buy one tomorrow. But when I play shooter games like MW3--I'm suddenly going 0 and 6(and I quit) and similar when that's not normal for me. I was expecting "quicker visual responses" not slower. And I'm running MSI afterburner and can see the card is only working at 60% capacity in MW3 1920x1080 but just slow. It looks good alright, but suddenly everyone shoots me and I can't seem to return the favor. Wasn't like that yesterday or for years prior. Somethings wrong?!!

Unless something else has suddenly gone bad with my machine, possible, this thing is disappointing for me. Maybe its my mobo bios or something has gotten damaged on my machine? Could be. Tomorrow I plan to reinstall the 5870 and see how it performs. If I suddenly go back to scoring normal..then I can only think somethings wrong with the new card. I'm experienced at all the drive cleaner/current not so current AMD drivers etc. Its not that area. Its something else.

If i can't figure it out, i plan to return it for refund or exchange. I just got it today from Amazon. Anyone have a similar experience?

Are you using bios #2? I flipped my bios switch to #2 position about 30 minutes ago, I then installed the latest version of Sapphire Trixx.....I increased my core clock to 1200 and memory to 1600 , Board LI to 20%, underneath Board LI is an option to increase the voltage which I did to 1.200......So far its reacting better, no more weird freezing I was getting.

I just dl'd that 3dmark program. I will run it late tonight or tomorrow as I am watching a MLB game in another window.
 
Ya it doesn't sound right, but real world/real person testing is not what they do when they run these 3DMARK11 type scores..I would suppose.

I'm no expert on graphic scores but I know how to use graphics "hands on". And I guess the 3DMARK11 numbers are maybe kinda simplistic in that perhaps they're measuring things that some of us actually aren't using.

For instance in fast pace shooters one isn't standing around gazing at the physics of the dust cloud etc. Players are concerned with "seeing" the enemy better and quicker.

I suppose when gpu's are designed they must make trade offs and try to balance all. Just guessing. So I suppose some gpu's are better suited to some applications than others.

That's my speculation.

Edit: whoops wrong card comparison.

The HD7970 will destroy the HD5870 in just about any benchmark. Both cards should handle MW3 easily, over 60FPS. Especially the HD7970. Check if vsync is on though - this can bring about mouse lag which feels very awkward and to me makes it much harder to aim. Measure the actual FPS using FRAPS and you will see. The HD7970 will be quicker than a HD5870 CF setup too, seeing as scaling on the 5XXX series wasn't as spectacular, some games didn't scale very well for my HD5970. The higher the FPS, the higher the chance of seeing the enemy quicker (although this won't occur above the monitor's refresh rate).

3dMark11 is more representative of modern games - MW3 is still based on DX9 and is old, just a console port. Try a modern DX11 game and you will easily see a difference between the two. Games like BF3 are a good comparison - the HD7970 will pretty much handle it on ultra with FPS drops below 50 rare. The HD5870 would only yield about 30FPS on ultra by comparison.
 
Last edited:
Thanks I'll try that. I've been running all at my long time overclock. I'll switch to the OEM profile and give that a try. And yes I'm using AB 2.1.0..didn't know there was something newer. I'll get that for sure.

I'm pretty sure its something I'm doing/or not. As an example I've run a couple tests this morning of the 7970 at 3DMark11 and I'm getting scores like P7,991 haha..Can't be right. .

http://3dmark.com/3dm11/3346770 (P8,668)

Ran it at stock cpu (2.8) but it reports 2.9?

Ran gpu at 1200/1600 but it reports 150/300?

Had my CCC 3d app set to Morph AA, highest quality settings, vsynch, the same way I play pc games....using cat 12.4's

I think if I oc my cpu later I can hit 10k or better....
 
Last edited:
I wouldn't worry too much about 3dmark. If it can't report + detect accurate clock speeds on the cpu OR the gpu the scoring is also flawed. Haven't used it since it was MAD ONION, myself, till today.

I just ran it to see if the OP was scoring low.....
 
Scoring is not flawed. The displayed freq. in 3dmark are sometimes the idle freq.
 
its not flawed, but real world test is not the same as just benchmarking
 
Back