Designing Our Next GPU Test Suite
by Ryan Smith on March 14, 2010 12:00 AM EST- Posted in
- Ryan's Ramblings
In keeping with our desire to refresh our GPU test suite periodically, we’re going to be redoing our GPU test suite to rotate in some more modern games, along with rotating in some DirectX11 games capable of taking advantage of this generation of GPU’s full capabilities. And while we already have a pretty solid idea of what we’re going to run, we wanted to throw out this question anyhow and see what responses we get.
What games would you like to see in our next GPU test suite, and why?
What we’d like to see is whether our choices line up with what our readers would like to see. We can’t promise that we’ll act on any specific responses, but we have our eyes and ears open to well-reasoned suggestions. So let us know what you think by commenting below.
Facebook
Twitter
RSS
240 Comments
View All Comments
NT78stonewobble - Tuesday, March 16, 2010 - link
Regarding the selection of games: I'd like to see a nice broad spectrum of graphics engines. If I remember them correctly Unreal Engine, farcry 2, crysis and other engines atleast that way we can ourselves atleast correlate a bit to older games using the same engine.Minimum FPS values if at all possible. Especially to discover if these are lower on the dual chip cards and or sli/crossfire setups.
You can still keep running the synthetic tests. There are those who like them. Just don't focus on them in the conclusion.
Keep the focus in graphics card reviews on the graphics.
But maybe you could do a roundup once in a while with different processor / graphics card combinations to discover what processors are needed to drive what graphics cards at what resolutions. Personally I have an e8500@3.8ghz, 4 gb ram and a gtx 260(216 core) but even though my processor is only a dual core and old I'm still graphics card bottlenecked. Since I game at 1920x1200 with as much eye candy on as possible. AFAIK I'd only gain a few extra fps by making a major upgrade to the higher clocked i7's. It would be nice to get a confirmation that I am correct in this.
An additional article with both your old and new games and popular ones and the performance and bug mentions of eg. the last 3 driver sets of either ATI and Nvidia. I know this is a big SOB article with lots of work. But as a consume I am sick and tired of having to switch drivers all the time to get a specific game working and then loosing 30 percent performance in other games. An article to discover who has the best drivers here and now would be nice. For a consumer perspective it doesn't really matter that a card is the worlds best at game xxxxx if it has alot of errors in others.
So basically your graphics card reviews are good (needs those minimum fps numbers) but additional articles to help out buyers to get the best performance for their setup (processor vs. graphics bottleneck) and service/ease of use (state of drivers) would be really really appreciated.
XiZeL - Tuesday, March 16, 2010 - link
AS many replies state i would also like to see Battlefield Bad Company 2 included, it one of the best selling games actually a verry demandig game for PC's unlike CODMW2.XiZeL - Tuesday, March 16, 2010 - link
sorry for the double post but would be also good if you show how each game scales in Crossfire and SLIjive - Tuesday, March 16, 2010 - link
whatever games you end up choosing please note the engine the game is based on. This gives a rough estimate on how other game titles based on the same engine will fare on the similar HW platform.You've probably done this before but don't test two games based on the same engine, unless for some strange reason those would scale differently. If this is the case an in-depth analysis why it is so, would be interesting.
Drazick - Tuesday, March 16, 2010 - link
Please include some Direct Compute / Open CL tests.Thanks.
BoFox - Tuesday, March 16, 2010 - link
Hey, I think that the games tested are great. The more games, the better! (As long as they are popular games, not crappy games, heh!)Anyways, I think it would be very helpful to include minimum frame rates like over at Xbitlabs. Even more useful would be to provide time-based graphs of frame rates like those shown at HardOCP. Some games that average at 30fps for a certain card would be pretty consistent, staying within 2-3 fps of the average.. on another card (of the different "make", be it NV or ATI) it would dip to 1/2 of the average or worse. Would it dip just once throughout a couple minutes of benching, or would it dip several times?
About the Steam poll that shows most of the user base to still be gaming at 1280x1024 or lower--I do not think we should be alarmed at the statistics. Those who play at say, 1024x768 are most likely those who are using a rig that is around 4-5 years old on average, or older. They are most likely the ones who only play games and do not give a slightest amount of feces about the computer hardware/benchmarks. They do not even know what the display configuration settings like Anisotropic or AntiAliasing mean. I have a few friends and a couple of brothers who play computer games but do not even care about turning up the resolution. They are most likely not going to bother reading Anandtech articles ever.
Those who bother reading the articles are definitely going to be interested at 1680x1050 resolution as the minimum, trust me.
I guess that's all I have to say here. Just make sure to keep a couple of DX9 games around.. there will continue to be hugely popular DX9 games until Xbox360 and PS3 are finally replaced by the next-gen consoles (since most PC games are ported from the consoles).
fepple - Tuesday, March 16, 2010 - link
Yeah, no point looking at the steam polls when they can check their web stats to see what people who actually visit the site use :)Though having said that I mainly read the site at work hehe
phuzi0n - Tuesday, March 16, 2010 - link
There is one yet to be released game (~2 months until release) that you absolutely must include: STARCRAFT 2If you don't include SC2 then you will lose merit with a HUGE portion of gamers. You did good by finally including World of Warcraft in the tests, albeit too late to matter, but for SC2 you need to include it from the moment it is released. There will be many people looking to upgrade their cards to handle SC2 so make sure that you have the data for them!
glockjs - Tuesday, March 16, 2010 - link
i really don't care what you use tbh. just has to be somewhat current. the only thing i would say NOT to use is any console port or any game made by blizzard/activision.btw mmorpg's are a bad idea to use..they're made to run on crappy hardware so they can capture a wider audience.
HillBeast - Monday, March 15, 2010 - link
You should probably try to have as few if any NVIDIA The Way It's Meant To Be Played and ATI oriented games. I know this will be hard but there is alot of games out there that prefer specific brands and it makes it kind of unfair. Also seing you will be comparing NVIDIA and ATI, show results of w/ and w/o PhysX enabled and when games and stuff start using DirectCompute show the same thing, w/ and w/o.