AMD’s Radeon HD 6870 & 6850: Renewing Competition in the Mid-Range Market
by Ryan Smith on October 21, 2010 10:08 PM ESTThe last new game in our benchmark suite is Civilization 5, the latest incarnation in Firaxis Games’ series of turn-based strategy games. Civ 5 gives us an interesting look at things that not even RTSes can match, with a much weaker focus on shading in the game world, and a much greater focus on creating the geometry needed to bring such a world to life. In doing so it uses a slew of DirectX 11 technologies, including tessellation for said geometry and compute shaders for on-the-fly texture decompression.
It’s also one of the few games banned at AnandTech, as “one more turn” and article deadlines are rarely compatible.
Civ 5 has given us benchmark results that quite honestly we have yet to fully appreciate. A tight clustering of results would normally indicate that we’re CPU bound, but the multi-GPU results – particularly for the AMD cards – turns this concept on its head by improving performance by 47% anyhow. The most telling results however are found in the GTX 460 cards, where there’s a clear jump in performance going form the 768MB card to the 1GB card, and again from the 1GB card to the EVGA card. The 1GB GTX only improves on memory, memory bandwidth, and ROPs, greatly narrowing down the factors. No one factor can explain our results, but we believe we’re almost simultaneously memory and geometry bound.
With that in mind, this is clearly a game that benefits NVIDIA’s GPUs right now when we’re looking at single-GPU performance. This likely comes down to NVIDIA’s greater geometry capabilities, but we’re not willing to rule out drivers quite yet, particularly when a partially CPU-bound game comes in to play. In any case NVIDIA’s advantage leads to their wiping the floor with AMD here, as even the mere GTX 460 768MB can best a 5870, let alone the 6800 series.
Crossfire changes things up, but only because NVIDIA apparently does not have a SLI profile for Civ 5 at this time.
Facebook
Twitter
RSS
197 Comments
View All Comments
krumme - Friday, October 22, 2010 - link
No it more looks like NV have been playing around with rivatuner.Their cards at stock speed is a toaster. Overclocking again just before an AMD release does not make it better.
If they can sell cards at higher speeds do it. Release it. Dont talk about it, or coverdly hide behind factory overclocked cards. Pathetic and weak.
The sad thing about it is they made Anand change their methology.
What a sad result when the start was just bad gfx cards.
MrSpadge - Friday, October 22, 2010 - link
Not sure it was mentioned before: Cypress and RV770 run FP64 at 2/5th the FP32 speed, not 1/5th as written on the first page.MrS
Quantumboredom - Friday, October 22, 2010 - link
It's 2/5 for addition, but 1/5 for multiplication and FMA as I recall it.krumme - Friday, October 22, 2010 - link
I you want real world testning for solid buying knowledge and no idiotic mixing of oc and non oc cards go to HardocpAnand needs to do a solid review of his own gfx reviews. This is stoneage methology with a fishy smell.
For the ssd there was an excellent revision with the real world bm suite introduction, - a revision far deeper is needed here.
chillmelt - Friday, October 22, 2010 - link
I agree. Comparing a non-OC'd card to an OC'd one is in no way fair. The 6870 clearly clubbered 460 gtx in virtually all benchmarks, at stock speeds.krumme - Friday, October 22, 2010 - link
Yes. Think about this scenario: For this review anand (kyle) decided to use an overclocked Sapphire 6870 and compare it to a stock speed 460.What would happen? How would that look?
Its the same thing happening here.
mindbomb - Friday, October 22, 2010 - link
if they are in the same price range, they should be compared.it gives the consumer more information on his purchasing decision.
idk why this arbitrary oc/non-oc rule comes into play. It was a factory overclock, and required nothing on the part of the end user.
krumme - Friday, October 22, 2010 - link
Okey. Now take an OC 6850, 900Mhz, and compare it to a far more expensive stock 460 1G?Where does this end. AMD and NV suplying the sites with factory overclocked cards for reviews. Its going to be a f..king mess. No one knows about availability, time period and so on.
Regarding pruchasing decision all the methology is plain wrong. Go to Hardocp if you consider buying a card. Here you have real world testing. Not like this old bad - and now fishy - methology of anand.
Why does he change methology just when he is getting the new cards. Hmm. Thats simply very bad science.
NV should concentrate on making cards that performs instead of using ressources putting pressure on sites like anand. If they have faster 460 models release it, or stop spinning.
Its a pain to read and watch. And its a shame for a site, that have the ressources to do proper testing, but dont have the balls any more.
We dont like AMD, Intel nor NV to influence the reviews. Right now there is an obvious bad influence from especially NV. And you, and everyone can see it. Dont underestimate the consequenses.
bji - Friday, October 22, 2010 - link
You COMPLETELY missed the point. *IF* the OC 6850 was a factory overclocked card, sold at a price that puts it in the range of cards reviewed, then it *WOULD* make sense to compare it.So your example just further makes the point.
krumme - Friday, October 22, 2010 - link
Well as it stands now they dont release oc cards at launch day, but that will soon change..., if Anand and others dont stick to their policy