Graphics Card Hierarchy Chart

The GTX560Ti gets nothing. Sad, really.

5770 "Great 1920x1200 performance in most games"
6850 "Great 1920x1200 performance in most games"
that makes no sense, that basically says the cards are on the same level, the 5770 does NOT give great performance at 1920, the 6850 only just cuts it at that res. Please re-structure your comments as to how well these GPU's play games, all the lower end ones are way off too, a 5570 cant play recent games well at 1680x1050, not even at 1440x900. Everything seems pretty much way off.

For $220 a GTX 560Ti should be a good bargain.

I own a Sapphire Radeon HD 6950 2GB, pretty good video card. It's giving me good times with crysis, even with that useless tessellation workload they put on background, i caped that out, and i really like my vid card.

toms, did you know i can thumbs up my own message? it happens when im not logged in, then press a thumbs button, then it asks me to sign up or log in, and i log in. I can then thumbs up my own message.

I was wondering if we can include the HD6870X2. It has a good bang for the buck against GTX580....

Hey where is the 6970 ??? It has great performance and good price for a single card.

Should list Memory Bandwidth if you're going to list memory bus width and memory speed.

People looking at an article like this are really looking for memory bandwidth when looking at those two parameters.

Regarding 5670/6670:
the fastest card (at least the reference version we have from AMD) that doesn't require an auxiliary PCIe power cable. That is to say its power requirements are entirely satisfied by a second-generation PCI Express x16 slot.

If the motherboard only has PCI Express x16 1.0 (not even 1.1) will that be a problem?

i wish there was a free category

the famous card of the year 2011 is gtx 460

I stubbled across a GTX590 in a retail store the other day and snagged it. It is an interesting beast. The performance is epic, and so is the heat. GPU #2 runs 8-10 degrees hotter than the already hot GPU #1. This did not change even with an open air case. I really wonder how much faster this card could get with a more dedicated cooling solution. Aren't the GPU's dialed down like 40% from thier single-card siblings?

So the majority of GPUs on the market are in this list? How about being more strict? GTX 460 and 6850 can't be tied if they have the same performance and price but one uses 25% less power than the other.

Wow, SLI and Crossfire demoted. I knew this should be the case after reading your article about micro-stuttering, but I've thought this is not going to happen. Respect Tom's Hardware, I wish all journalist were so honest.

Maybe you could come up with some idea how to compare single and multi-card rigs? In my opinion baseline of those zig-zags we saw on charts in Micro-Stuttering And GPU Scaling In CrossFire And SLI article is good determinant of perceived frame rate, because every other frame rendered fast is just useless.
This should be discussed among journalists. I know hardware manufacturers wouldn't like it, but we would know the real performance, not just guess how much of our cards performance (and frame rate) is eaten by micro-stuttering. We would also led to minimizing micro-stuttering not just maximizing this illusional frame rate we measure now.

Id take a gtx 560ti hands down over a 6950, most of the benches toms uses favor amd and beyond that I own a gtx 560ti ASUS DC2 model and it holds a 1ghz core oc, at that speeds it kills a 6950 and is on par with a 6970, for me the ti is on the same tier as the 570, 295, 6950, etc, it is ridiculous to say that you can actually see a difference between a ti and a 6950 at stock in terms of real world performance, Toms seems to be more and more biased as time goes on, and also that raw performance alone does not tell the whole story of performance in some instances where the gtx will offer better minimums, heat, and noise...

It saddens me that SLI/CF has been downgraded, but I understand why you did it and I applaud you for it. IMO, minimum frame rate is just as important a factor as average frame rate for the actual play experience. A steady 30 FPS is better than an average 60 with occasional drops to 10.

Many HD 6950s reach 1GHz Core and they overclock a lot more on the Memory, so you're wrong.

Also, they're right about there being a noticeable performance difference between both stock:

http://tpucdn.com/reviews/ASUS/Mar [...] l_1920.gif

51/47: 8%. I think 8% is enough to call it a win overall. Oh, and the 6950 2GB is actually 1% slower than the 1GB card at 1920x1200.

AMD 6950 2gb in crossfire is definitely still my pick. Good price/performance value.... I got two of these about 6 months ago and really don't regret it at all....

I am still glad I got the XFX 1GB Radeon HD 6950 at tax time. It has played all the games I like to play admirably, and I was able to unlock the extra shaders with very little problem. I may be looking for a 2nd one, if I dont upgrade to a 7xxx series, depending on price and what the new version have to offer. Either way, Im sure Ill be happy.