Discussion in 'Reviews' started by W1zzard, Jan 4, 2012.
Once again I am amazed at the idle power consumption.
Nice review Wizzard.... As always! Any news on TriXX support for the Sapphire version?
I'd say something like:
"In 2D, there are no additional fans running compared to a single card, due to ZeroCore power. So any difference in noise is made by the change in airflow path alone, and probably won't be noticeable.
In 3D, obviously you've got too fans rather than one, and, more significantly, stifled airflow so the top card's fan will have to run faster. Whereas a single 7970 is about as noisy as an X, two 7970s are noticeably noisier- about as noisy as a Y."
You have to fill in the X and Y, I've never worked with these cards or enough other cards to compare them with
Some reviewers also go for DB readings, though these can be misleading as certain types of noise are more noticeable even when they are no noisier (e.g. high pitched).
Don't take any of this as criticism, though. I made my original post just to point that out, not to have a go. I think you consistently make reviews that are amongst the best on the internet.
Excellent idea... would be very interested in reading it when/if you do.
So you're saying 2 speakers in stereo isn't louder than 1 speaker in mono? I agree adding a second speaker does not double the level of sound, but 2 speakers from a distance can produce a louder volume at a distance. Example: Take a single speaker or fan and sit it 15 feet away and then take 2 speakers or 2 fans and sit them the exact same feet away and they will be louder than the single speaker or fan. Another example turn on a 40mm fan in a server room.. then turn on a 1u server with say 6 40mm fans and it will sound like a huey taking off.. hehehe Just my 2 cents..
@W1zzard : Great review as always!
Awesome power right there
No. You can tell that I'm not saying that, because I didn't say it
But that's what I actually said. Excellent.
100% scaling in some games made my pants sticky........I know, I have low expectations of life.
Err that was me being sarcastic.. I should have added the, "I get it" at the end.. hehehe I was going for the student asking the teacher.. hehehe bit..
And I was agreeing?
Then I was trying to help you explain it in a non tech way.. You're welcome!
I can't wait for the 7950's to come out! I might replace my 5850's with them.. I really want that second card to cut off using only 1w when not gaming....
So....according to the summary... adding another 7970 for crossfire only gives it a 24% boost over the single card??
that makes it real hard to justify getting 2 cards.
AMD still haven't improved much the software side of things for that. Crossfire doing worse than a single card in Skyrim, for example, contributed to lowering that overall percentage.
yeah, with the amount of issues ive recently had with AMD drivers im considering selling both my 6970s and grabbing one 7970 for my needs
Is there an error on the Skyrim benchmarks page in regards to the Geforce 590? Is there really negative performance scaling for the 590, or are the 590 scores switched with the Geforce 580 scores?
There are a few benchmarks that don't do well with crossfire, but mostly it's just crossfire only doubles your GPU computational performance - it doesn't double anything else to do with the GPU (bandwidth, for example) or CPU speed. As you get to higher resolutions, scaling will appear better and better because the bottleneck shifts towards GPU computational performance (right until you run out of VRAM ).
sli not working, but eating cpu time + lower clocks = lower score than 580
W1zz, I need your personal professional opinon ...would you pefer a single 7970 3GB or a GTX 560 Ti 2Win ..which is Dual GPU however only 1GB GDDR5 per GPU...yes the GTX 560 Ti 2 is faster in most cases, but more and more games are going to be requiring more than 1GB GDDR5..
I am still waiting for the 7950 to come out. I think that crossfired will be the best bang for your buck
Apparently Nvidia the maker of all good software has just decided not to bless its users with that option either for some unpopular games like skyrim.
AMD is de devil and makes bad software we can both agree though.
Any way mounting dimensions could be posted W1zz?
You know, I told myself I was going to skip the 7-series a few months ago.
Today, I want three.
Although I see some scaling issues, I see the smae problems on my 6-series cards more often than not, so for an initial showing, I think the results are pretty good.
That said, the numbers are one thing. the actual usability is another.
I am very interested to see if ZeroCore is working With Sandybridge's IGP, with Lucid software switching between devices, as this may provide the ultimate in power savings, yet raw horsepower...IGP gets used @ the desktop, and the 7970 is idle @ 1W...then under load the 7970 kicks in, and you get great gaming performance, too!
Of course, that brings up the question of whether such tech would work with Crossfire as well. For me, that would be ideal, and I'd not need multiple rigs, one for work, with more CPU than GPU, and one for play, with more GPU than CPU. Potentially you can get the best of both worlds here...
Isn't Ivybridge supposed to supprot multiple monitor outputs, too? IvyBridge IGP for 2D, and 7970 Crossfire for Eyefinity, on the same rig? Will that be possible?
if i had something that would serve me for basic gaming i wouldnt buy anything until at least summer
It all depends what you're going to be playing. The games where scaling does work, it's near 100%.
The thing is that these cards are still very new. Scaling will get better by the next set of rivers for sure.... and better and better each time drivers are released. My GTX 580's didn't scale very well in everything either when they first came out but 2 to 3 months into it they sure did. There is still a reason for getting two cards..... and that reason will get even better as time goes on.
nice review w1zz!!! here's hoping that this "zero core" thing actually works as hopped and doesnt just mess up crossfire for these cards. I know with earlier cards amd has had issues getting the power states right at a driver level causing users to have to hack the driver or create certain profiles for gaming.
Oops, sorry, forgot to respond..
No, I was referring to my i7-920 @ 4.0Ghz versus my i5-2500k @ 4.8Ghz.
That's what gave me the 15% difference.
Also, I can't claim anything about the HD 7970, but PCIe 2.0 x16 has no noticeable performance difference compared to PCIe 2.0 x8 with my GTX 480s.
W1z has done numerous tests showing what kind of difference the amount of PCIe lanes makes, and if I recall, x8 to x16 on 2.0 does make a difference.
it probably doesn't, but trackr would complain about it anyway
Separate names with a comma.