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
Barneyk - Sunday, March 14, 2010 - link
I really wanna see old games tested more, especially for mid- and low-range cards.I would also like to see a combination graph where you add all the numbers together, or some other sort of graphic that shows the combines results.
Its not really what you're asking for, but that is things I wanna see. :)
And also, when it comes to CPU-testing I really feel that the lack of multitasking testing is hurting the overall picture I get.
f0d - Sunday, March 14, 2010 - link
i believe bad company 2 would make a great addition to the next GPU test suitealso i would like to see different cpus to be included not just clock speeds on the same designed cpu
like
i7 960
i7 870
q9550
x4 965
x2 255
e8600
but that maybe more to do with a cpu benchmark
maybe a game test suite is what we need? test a single game over multiple resolutions with multiple video cards and multiple cpu's
maybe its just me but i would like to see how the different hardware scales especially at higher resolution
ive been so fascinated over it since i saw this http://techreport.com/discussions.x/18095">http://techreport.com/discussions.x/18095
Paulman - Monday, March 15, 2010 - link
+1 agreeAlthough, to be realistic I'm sure Anandtech would have to carefully choose how many additional CPU's they'd want to test, and for which specific games and GPU's they'd test it, since the possible combinations for testing will easily skyrocket.
But it would be very valuable if Anandtech tried - especially so we can see the impact of dual-, tri-, and quad-core scaling with particular games and at particular GPU performance levels.
deathbycomputer - Sunday, March 14, 2010 - link
Ah. My thoughts exactly. Different CPUs should be used because not everyone has a i7 975. A lot of the results on GPU benchmarks don't reflect those of which the consumer will realistically achieve.Rebel44 - Sunday, March 14, 2010 - link
New AvPS.T.A.L.K.E.R.: Call of Pripyat
Please show us graf of FPS in time - just like HardOCP.
StormyParis - Sunday, March 14, 2010 - link
Specifically, most tests answer the question "Which card should I buy ?", but not the seminal question: "Should I upgrade now, and what ?".So I'd trade a handful of game tests for a handful older cards in the tests, and a RAM / CPU impact test.
TGressus - Sunday, March 14, 2010 - link
As others have surmised, average frame rate is not half the story. Other sites already use a line graph of their recorded frame rate dataset and it does give a great overview of that particular benchmark. The problem is that more than 2 cards on one chart becomes a mess to readYou could improve upon the idea by using something similar to the Frequency function in Excel and graph the number of times frame rate was recorded at 10fps increments. This would clean up the graph and still reveal trends like, "Average frame rate is 84, but you will be spending most of your firefights 10-20 FPS south of that."
Example:
http://img229.imageshack.us/img229/3137/frequencyi...">http://img229.imageshack.us/img229/3137/frequencyi...
At very least make sure to report both minimum and MEDIAN frame rate if you have the data.
makin moneys - Sunday, March 14, 2010 - link
Bad Company 2 - It's a new, high production value FPS, many have voted for this alreadySupreme Commander 2 - Should provide a good RTS benchmark; Blizzard is notorious for making their games able to be ran on low end hardware so I think this would be a better benchmark choice than StarCraft 2
Batman: Arkham Asylum - Another hardware taxing game, like BC2
Bioshock 2 - A new, impressive looking RPG/FPS
(when it's out) Crysis 2 - Needs no explanation
I also roll my eyes whenever I see MMO benchmarks. There are way too many variables to get any reliable runs. Same for other games that don't have a benchmark built in, but they aren't as unpredictable as MMOs.
Ryan Smith - Wednesday, March 17, 2010 - link
I actually had a chance to talk to the GPG guys at GDC last week. At this point we're not planning on using SupCom2; they told us straight-up that it's not going to be graphically intensive (so they're pulling a Blizzard here).Dzban - Sunday, March 14, 2010 - link
Since when Batman is hardware taxing? It runs great on whatever GPU you have. In full HD it plays great even on 9600GT. It's a waste of time to test it on newer GPUs.