A Great Article!

I bought HD5850 and I am glad that I can overclock it to perform as good as HD5870, which costs as much as 40% more than HD5850 in my country.
Alas, I cant find a reason why to OC it anyway...

Does the overclocked frequency influence the idle frequency, or is the higher power draw caused by increased voltage?

I'd also like to see a review of HD5770 on xbitlabs, as the early ones on other sites aren't as detailed when it comes to cooling, overclocking and overclocked performance.

Awesome review! Lots of games there and I like how you refuse to turn off PhysX for Batman because the game looks so bland without smoke and stuff. However, the game developer is neglecting owners of ATI cards when using PhysX instead of Havok or Infernal or some other algorithm for physics that yield similar-looking physics, if not better-looking and better-performing for both parties (ATI and NV)--see the Ghostbusters game for PC.

I am surprised at how close the 5850 performs to a 5870 when overclocked to equal speeds. In many of the benchmarks here, the performance is actually identical. This phenomenon supports a growing popular theory that 5870 chips are severely bottlenecked by the memory bandwidth. It is so badly bottlenecked that an additional 160 shaders and 8 units hardly does any good for a 5870 compared against an equally clocked 5850. Since the memory bandwidth is a rather critical issue for 58xx cards of today (as the 5870 chip should perform exactly 100% greater than a 4890 plus some DX11 features, but has only 23% greater memory bandwidth than a 4890), I will try to clear up a few more things about it.

Whenever a 5870 loses out to a dual 4890 in Crossfire (in identical settings), the memory bottleneck (and possibly L1/L2 cache bottleneck also) is to be blamed rather than the drivers. Crossfire scaling is never 100% efficient due to additional CPU overhead and some other factors. If the drivers could scale two 4890's so that it beats a 5870 by a wide margin in some games, the drivers should be able to scale all of the shaders/cores in a 5870 chip just fine since it is a very similar chip architecture that actually looks like two units of 4890 chips infused into one chip. ATI is not known for introducing a new driver set that brings such incredible performance boosts like Nvidia has done a couple times in the past several years, so this is not to be expected.

An excellent test would be to reduce the memory bandwidth of each 4890 to exactly half of a 5870, and then crossfire the 4890's together to see how it compares against a 5870. If the 5870 is now at least equal to 2x 4890's in all of the games in which the 5870 was previously beaten by 2x 4890's (let alone a 4870X2), it would confirm this issue for once and for all. The first hardware-review site to do this test could gain quite a bit of popularity and respect.

Great article. I especially like the game selection and resolution/setting choices. Kudos.

Can you shed any light on the problems people are having here:
http://forums.overclocker...showthread.php?t=18063487

There seem to be quite a few people with a variety of manufacturers having this issue. This is the only reason I can find not to buy.

Why everybody is testing the graphic cards with the most expensive CPU, remains a BIG mystery for me!! Just check the latest STEAM surveys, that more than 70% still run dual-cores, and cheap quad-cores. So what's the purpose for top line procs???? To show off??? Don't tell me is because CPU bottleneck, because that's a BIG B.S.!!! (

Personaly I am more interested to see how those video cards react on normal processors, like an E8500, Q6600, or AMD's equivalents. There you can see perfectly if is worth buying an ultra expensive video-card or not.

Every-time I see the numbers from the tests, I have to think, "Well if this card can output 250fps on an I7 ultra-super-extreme, how many FPS will it output on a E8500 o.c. ????" For once, I would really REALLY wish to see how those cards do on real-life situations with normal components, not on the most expensive ones!