ATI's X8xx CrossFire Graphics Arrive
by Derek Wilson on September 26, 2005 1:00 PM EST- Posted in
- GPUs
Mode and Hardware Scaling
There are two scaling factors we want to pay attention to here. First, we will be looking at how CrossFire scales with AA mode. As 8x and 10x share the same performance numbers (as do 12x and 14x), we will be looking at: no AA, 4xAA, 6xAA, 10xAA, 12xAA modes. We will be able to see just how steeply Super AA modes drop off in this test.We will also be looking at single card performance. It will be important to compare these numbers to the Super AA numbers, as each card in Crossfire mode should ideally be able to render the entire scene at single card speeds. Any difference between the Super AA modes and a single card running at either 4x or 6x AA account for the overhead of the communication system and compositing engine. The performance drop in these modes is more pronounced than in the other modes, as each card needs to render the entire scene at full frame rate. Any latency or bandwidth limitations will show up more easily here.
Our graphs show frames per second on the y-axis and AA mode across the x-axis.
Doom 3 shows a good performance increase due to CrossFire at standard AA settings. Both the single card and CrossFire setup are similarly GPU limited as AA increases. The drop from 4xAA on a single card to 10xAA on CrossFire is nearly 50%.
When we take a look at Half-Life 2, we see that a single card is fairly GPU limited with increasing AA while CrossFire handles it well. Going from single card AA modes to CrossFire Super AA modes shows more than a 50% drop in performance. This is likely due to the fact that as framerate increases, the impact of latency and bandwidth restrictions on transferring fully antialiased scenes from the slave to the master is proportionally greater.
Here's a look at how going from a single to a two card configuration scales under NVIDIA and ATI hardware under the games we tested.
| Multi GPU % Performance Improvement (16x12 noAA) | |||
| Doom 3 | Battlefield 2 | Half-Life 2 | |
| CrossFire | 41.3 | 53.8 | 15.8 |
| SLI | 34.0 | 89.3 | 22.2 |
| Multi GPU % Performance Improvement (16x12 4xAA) | |||
| Doom 3 | Battlefield 2 | Half-Life 2 | |
| CrossFire | 69.2 | 63.6 | 35.5 |
| SLI | 81.3 | 89.3 | 68.8 |
While Crossfire performs reasonably well without AA, the 4xAA modes show that SLI provides better scaling. More importantly, though, the lack of support for higher resolutions - as well as 1600x1200 at higher refresh rates - becomes a serious limitation of Crossfire. 7800 GTX SLI running at 2048x1536 has no competition from any ATI configuration for the time being.
Facebook
Twitter
RSS
76 Comments
View All Comments
waldo - Tuesday, September 27, 2005 - link
wow, that commment sure brought a lot of attention.I would agree that Thom's appears to be "outright-bought" in this article as they don't post the limitation of hte card at 1600x1200 @ 60hz, or at least not as clear as Anand, but they do make some valid comments that Anand's article didn't post either.
Perhaps they are targeting different audiences? No idea, but why are their numbers on par with the 7800 SLI in many situations? That sounds fishy somewhere.
DerekWilson - Tuesday, September 27, 2005 - link
Just to cover what others haven't yet --Adding an X8xx card to an R520 CrossFire card would either not perform well at all or would not work.
Personally, dual-GPU as an upgrade solution is not really a plus unless you are just deffereing purchase for a couple months while prices drop and your wallet heals from the first GPU purchase. If your personal upgrade cycle is a year or more, you'd be much better off just buying a new card from a new architecture.
Everyone else has pretty well hit it on the head. The 1600x1200 limit is a killer. We also have no availability and we are butted right up against the launch of ATIs next gen parts.
I wouldn't recommend SLI either -- as I said -- unless you want absolute maximum preformance. My recommendation may change to CrossFire after the R520 comes along. But who knows what the results of that comparison will be :-)
Mixed modes would perform slightly lower than the dual x850 xt setup and still at most 1600x1200@60 ... Yes, the X850 XT CrossFire does well in performance, but if I'm not going to recommend the X850 XT CrossFire, I'm certainly not going to recommend a mixed solution that will perform worse.
waldo - Tuesday, September 27, 2005 - link
I wasn't saying that you would recommend the other, but it would have been interesting for the readers (who attempt to be congiscent, autonomous beings, and only act at the whim and will of god Anand!) to be able to compare for themselves, per chance see what a mixed solution looks like as that is a selling point of ATI over Nvidia.I don't think SLI from NVIDIA is much of a solution. If you have $1k to shell out for graphics out of the starting gate, great. But you have to get the same manufacturer, the same card, and then the motherboard to match. But I won't be buying an ATI Crossfire setup just yet either.
JarredWalton - Monday, September 26, 2005 - link
Here's my personal take:1) If the 1600x1200@60 Hz problem doesn't bother you, that continue. For me, it's a deal-breaker.
2) Do you already own an X8xx card of some form?
3) Do you have a motherboard with two PCIe X16 slots?
4) Using the ATI Crossfire chipset?
If all of those are true, X800 Crossfire is worth consideration. Personally, 1, 3, and 4 eliminate it from contention. That said, this is current X8xx Crossfire we're looking at. We're not reviewing R520 Crossfire yet, and it will address at least the first point in some fashion.
Brian23 - Monday, September 26, 2005 - link
pwnedfishbits - Monday, September 26, 2005 - link
Yeah, accusing a site of corruption could seem like an attack.
1600 res limited to 60 hz is a deal-breaker for me right away. I can't tolerate playing at 60 hz for any length of time. I'm not a snob about this, it makes me physically feel ill. Pairing two powerful and pricey cards in one system should offer better, easily. I would not plunk down that kind of money to not be able to play at 16x12 when others do it at liveable refresh rates. Nvidia has better price-to-performance single and dual card solutions available right now. ATI should have the same shortly. The mega AA sounds great, but if as soon as you turn it on you say "I need to upgrade now," then what's the point? While the reviewed Crossfire is certainly nice in performance, there's better to be had for the money from both GPU suppliers. It would be irresponsible to recommend it at this time at the current price.
Now if there's a confluence of specifics where this setup makes gaming and financial sense to a handful of people out there, more power to them, they should enjoy this. For the teaming masses of readers though, you shouldn't be suprised by the lack of recommendation.
TrogdorJW - Monday, September 26, 2005 - link
LOL... attack an enthusiast site? Say it isn't so!How about the THG article? They review the platform as a whole, so that's a bit different. I won't comment much on their review, but consider a few points.
They have a page entitled, "Advantages Of CrossFire Over SLI" that reads like marketing hype, and yet they make no mention of the resolution limitation or "Advantages of SLI Over Crossfire". Clearly, they know the limitation exists - ATI hasn't tried to hide this fact, and the lack of any benches at higher than 1600x1200 is telling in and of itself. You do the rest of the math. (Also, some of the results are at best suspect.)
IMO, the writing of the THG article was a bit higher quality, but the content was far more suspect. They come off making everything sound rosey for ATI, and only a fool or a marketing department would believe that. ATI isn't dead yet by any means, but Crossfire is doing little for me right now. Did you realize that it's still not available for purchase at retail? Hmmmm.....
Oh yeah, Catalyst Control Center is pure garbage. Slow, clunky UI, memory hog, and causes as many problems as it fixes. Anyone that tries to tell me how great CCC is (i.e. THG) is immediately under suspicion.
fishbits - Monday, September 26, 2005 - link
Oh, I wasn't making a blanket statement that a review site couldn't be accused of corruption. Just that if you're going to do it, don't then wuss out and pretend it's not an attack :)My experience with THG is mostly second-hand, so I don't say much about it. That they chose to not mention the 16x12 refresh rate limitation, especially with all the debate over it before today, well... that's scandalous.
erinlegault - Monday, September 26, 2005 - link
The TechReport review was the best. Anand seems to give poor Graphics reviews, and definately not up to par with their CPU, motherboard and memory reviews. It must be the authors differences of opinion.overclockingoodness - Monday, September 26, 2005 - link
Could you clarify why Anand's video reviews suck? Just because Anand doesn't benchmark and show 100 diff. graphs of the cards based on the same architecture doesn't mean AnandTech's video reviews suck. TR is quite redundant from one article to another, if you didn't notice.