T5450 VS T9300 Crysis (Vista) Benchmarks

[hydra]

Lots of interest in the first generation Penryns. Today I got mine so posting this
for any interested and to consolidate various benchmarks.

This proves to me that Crysis is somewhat GPU limited, not the reason I bought the T-9300.

For Crysis fans I would venture that the stock P6831FX is a bargain @ $1250

Crysis 1.1 internal benchmark_gpu DX9. Vista Home Premium

Stock P6831FX with T5450
Auto settings medium with anti-aliasing off, no OC. 1440X900

==============================================================
TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s)
!TimeDemo Run 0 Finished.
Play Time: 66.51s, Average FPS: 30.07
Min FPS: 17.30 at frame 1940, Max FPS: 40.91 at frame 74
Average Tri/Sec: 25425868, Tri/Frame: 845573
Recorded/Played Tris ratio: 1.08
!TimeDemo Run 1 Finished.
Play Time: 68.81s, Average FPS: 29.06
Min FPS: 17.30 at frame 1940, Max FPS: 52.42 at frame 67
Average Tri/Sec: 24815264, Tri/Frame: 853792
Recorded/Played Tris ratio: 1.07
!TimeDemo Run 2 Finished.
Play Time: 68.96s, Average FPS: 29.00
Min FPS: 17.30 at frame 1940, Max FPS: 52.42 at frame 67
Average Tri/Sec: 24774954, Tri/Frame: 854205
Recorded/Played Tris ratio: 1.07
!TimeDemo Run 3 Finished.
Play Time: 69.11s, Average FPS: 28.94
Min FPS: 17.30 at frame 1940, Max FPS: 52.42 at frame 67
Average Tri/Sec: 24712958, Tri/Frame: 854018
Recorded/Played Tris ratio: 1.07
TimeDemo Play Ended, (4 Runs Performed)
==============================================================
This run with the T9300 with same settings as above.
==============================================================
TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s)
!TimeDemo Run 0 Finished.
Play Time: 56.65s, Average FPS: 35.30
Min FPS: 26.24 at frame 1957, Max FPS: 43.90 at frame 80
Average Tri/Sec: 29854914, Tri/Frame: 845697
Recorded/Played Tris ratio: 1.08
!TimeDemo Run 1 Finished.
Play Time: 61.83s, Average FPS: 32.35
Min FPS: 18.39 at frame 154, Max FPS: 55.29 at frame 89
Average Tri/Sec: 27615084, Tri/Frame: 853699
Recorded/Played Tris ratio: 1.07
!TimeDemo Run 2 Finished.
Play Time: 53.83s, Average FPS: 37.15
Min FPS: 18.39 at frame 154, Max FPS: 55.29 at frame 89
Average Tri/Sec: 31742490, Tri/Frame: 854397
Recorded/Played Tris ratio: 1.07
!TimeDemo Run 3 Finished.
Play Time: 53.82s, Average FPS: 37.16
Min FPS: 18.39 at frame 154, Max FPS: 55.29 at frame 89
Average Tri/Sec: 31778262, Tri/Frame: 855082
Recorded/Played Tris ratio: 1.07
TimeDemo Play Ended, (4 Runs Performed)
==============================================================

Gain Min:1FPS - Average:8FPS - Max:3FPS
This works out to about $100 a frame per second on the high end on
the second run

[crash]

Interesting benchmarks! A T5450 isn't particularly powerful compared to the T9300 so it doesn't surprise me that the T9300 achieved a few more fps. I think you'd see very little (if any) difference between a T7500 or T7700 and a T9300. Crysis is still very much limited by the GPU.

[lewdvig]

Now we have to figure out how to unlock those extra pipes and make GTXs out of our GTSs.

I winder if there are some easy to identify laser cuts on the GPU die.

Unlikely that a pencil trick will work here, but worth a look when I tear mine down.

[Snowsurfer]

World in Conflict.

• T5450 Avg fps 26 Lowest fps 12 Highest fps 56
  • T9300 Avg fps 36 Lowest fps 14 Highest fps 77
    • T9300+OC Avg fps 40 Lowest fps 16 Highest fps 92

[crash]

World in Conflict.

• T5450 Avg fps 26 Lowest fps 12 Highest fps 56
  • T9300 Avg fps 36 Lowest fps 14 Highest fps 77
    • T9300+OC Avg fps 40 Lowest fps 16 Highest fps 92

Keep in mind World in Conflict utilizes the CPU much more than a FPS game like Crysis does. Most RTS games, in fact, do. Supreme Commander is very largely dependent on the CPU.

Therefore, you'll notice a much greater increase in fps for RTS games than FPS games because RTS games use the CPU much more.

[brainer]

that proves what most people are saying... an 2.4 ghz C2d proccesor is Overkill for games

[B2TheEYo]

Now any FPS gamer that bought a pPenryn just for the extra power needs to ask theirself - is it really worth $100 bucks a frame? LMFAO

Great results.

Definitely you would see gains in CPU intensive applications though.

Really, I think Penryn is just a more optimized C2D chip, nothing more.

[hydra]

Actually it was $102 per frame, lol. But this thing flies doing video encoding!

I can, at this point, only bench my ancient T2400. Almost three times as fast encoding a DVD to a PSP.

The cost of a T7700 and the T9300 are both around $320 so what the heck The better chips are coming later this year?

[Cinner]

Really, I think Penryn is just a more optimized C2D chip, nothing more.

I think that's true, but that optimization comes with less heat and longer batterylife, so it's all good

[Ihaveworms]

If you have team fortress 2 or something, could you do a test there, that game seems rather CPU limited and I would like to see how big a difference the processor upgrade is.