Updated NVIDIA Benchmarks On Ubuntu With Team Fortress 2

[blackout23]

I find it a bit strange that the GTX 680 doesn't get higer frame rate on 1080p. I own a GTX 580 OC and get more like 200 FPS @ 1080p with everything at max, at all times in CS:S, which should be about as demanding as TF2. Also if you compare the GTX 680 to GTX 650 there isn't that much of a difference, despite the GTX 680 being way more poweful.

[Weegee]

Can't wait for the first nouveau benchmarks with reclocking support.

[verde]

i play tf2 with a gtx 460 with 331 drivers in fullhd without problems, something here is very wrong/strange

Confirmed.

I have a GTX 460 768mb too and I can play maxed out TF2 and L4D2 with vsync at 60FPS. 319.32 drivers.

[Andrecorreia]

something wrong

Can't wait for the first nouveau benchmarks with reclocking support.

this benchs are buggy, some error, or recording benchs destroy the perfomance

[tuke81]

I find it a bit strange that the GTX 680 doesn't get higer frame rate on 1080p. I own a GTX 580 OC and get more like 200 FPS @ 1080p with everything at max, at all times in CS:S, which should be about as demanding as TF2. Also if you compare the GTX 680 to GTX 650 there isn't that much of a difference, despite the GTX 680 being way more poweful.

Fps cap maybe? max_fps 130 Would give you such fps if your graphics card is powerful enough. Don't know what is the default fps cap in tf2(never played that game), but test results are most likely capped to 130 fps.

[Veerappan]

I find it a bit strange that the GTX 680 doesn't get higer frame rate on 1080p. I own a GTX 580 OC and get more like 200 FPS @ 1080p with everything at max, at all times in CS:S, which should be about as demanding as TF2. Also if you compare the GTX 680 to GTX 650 there isn't that much of a difference, despite the GTX 680 being way more poweful.

Either a framerate cap in the engine (as previously mentioned), or possibly a CPU limitation. The 1080p and higher res scores were identical, which could indicate a cpu bottleneck.

[darkbasic]

I bet an fps cap in the engine. Again: please benchmark Metro 2033 Last Light

[enfocomp]

Phoronix kicks ass.... thanks Michael! These benchmarks are exactly what we need to see right now, a popular source engine game compared between multiple cards on linux. Very ineteresting results, surprised to see the gtx 460 and gtx 680 do so poorly; I can't wait to see AMD thrown into the mix.

I hope you are in contact with NVIDIA so they can see the results... I bet they could benefit nicely from Phoronix Linux benchmarks.

Thanks again!

[STrRedWolf]

I'm itching for the Intel benchmarks myself.

[ElderSnake]

Confirmed.

I have a GTX 460 768mb too and I can play maxed out TF2 and L4D2 with vsync at 60FPS. 319.32 drivers.

+1, same card/vram here.

Something can't be right, the 460 is better than that. I can play TF2 @ 1080p no problems. Occasionally FPS dips to 30 or so in real intense outdoor areas, but I'm also kind of CPU limited with my old Intel E5200 Dual Core, and from memory the game is also fairly CPU hungry.

I will say though, I adore TF2 as a game, but even when I was on Windows I've never felt it was a great performer (especially in comparison to it's Source-powered siblings like Portal, HL2 etc). Always felt like there's a lot of metaphorical duct tape and bubble gum holding bits of it together (probably wrong) as it's got a lot of stuff going on and new things added all the time.