[brettjv]

The Goals of this Testing Process:
What I am studying here is the following three scenarios, using Crysis:Warhead as my benching application:
1) The effect of upgrading from a P35 with a core2duo, to an x58 with an i7, when using a BFG GTX295, using a common, easily attained overclock on each processor, and
2) The effect of upgrading from a 750 SLI with a core2duo, to an x58 with an i7, when using a SLI GTX280, again using a common, easily attained overclock on each processor, and
3) The effect of upgrading from a GTX295 to an SLI GTX280 setup, on an x58 board with an i7 processor, again using a common, easily attained overclock on the processor. Note that I was primarily interested in the difference between these two GPU setups when using EQUAL GPU (and CPU, equal everything, basically) clocks, so my analysis is done on that scenario.

GPU's Tested (More like a Legend for Tests Below):
GTX 295n (p35, [email protected] 4.0GHz, PCI-Ex 1.1 1x16X, STOCK SPEED)
GTX 295i (x58, [email protected]+turbo=3.466GHz, PCI-Ex 2.0 2x16X STOCK SPEED)
GTX 295ix (x58, [email protected]+turbo=3.466GHz, PCI-Ex 2.0 2x16X, OVERCLOCKED TO GTX280 SPEED)
2xGTX280n (750 SLI, [email protected] 4.0GHz, PCI-Ex 2.0 2x16X STOCK SPEED)
2xGTX280i (x58, [email protected]+turbo=3.466GHz, PCI-Ex 2.0 2x16X STOCK SPEED)

Boards Used:
P35 = Asus p5k-e wifi (GTX295)
750 = Evga 750 SLI FTW (SLI GTX280)
x58 = Gigabyte ex58-ud4p (GTX295, SLI GTX280)

Other Details:
All are DX9 Tests, using Microsoft Windows XP Professional Service Pack 2, using Crysis Warhead (unpatched), and the FBWH Benchmarking Tool. All tests EXCEPT the p35 tests were run using the latest (182.08 WHQL) drivers.

Brief Summary of Results
(in case you lack the stomach for sorting out my results below):

For the GTX295, the move FROM the PCI Express 1.1 P35 board with an e8400 processor at 4.0GHz TO the PCI Express 2.0 X58 board with the i920 processor at 3.466 yielded nearly universally POSITIVE results, showing GAINS in 11 of 15 of the tests. The overall range of gain was between -2.6% and +5.4%, with the preponderance of data showing a delta of somewhere around +4% for this (roughly 700 dollar) platform change.

For the SLI GTX280 setup, the move FROM the PCI Express 2.0 750 nforce-SLI board (2x16 lanes) with the e8400 processor at 4.0GHz TO the PCI Express 2.0 (2x16 lanes) X58 board with the i920 processor at 3.466 yielded nearly universally NEGATIVE results, showing LOSSES in 9 of the 12 of the tests, and improvements in only one of 1 the 12. The overall range of gain was between -2.9% and +9.0%, with the preponderance of data showing a delta of somewhere around -1.5% for this (roughly 700 dollar) platform change.

Once I have installed a better cooler on my system, I will happily investigate whether or not raising the clocks of the i7 will yield further positive results, but based on my fairly extensive benchmark observations over the course of my lifetime, I strongly suspect that improvements due to further overclocking will be minimal, as these are (purposefully) largely GPU-BOUND tests, intended to reflect the kind of gaming settings that those of us who buy $500 graphics solutions are looking to be playing at.

In comparing the performance of the GTX295 and the SLI GTX280 (both at equal clocks, that of the BFG GTX280 OC) on the Core i7 setup, we see a clean-sweep. The SLI setup defeats the dual-card solution in every test. The range of improvement seen by stepping up from the GTX295 to the SLI setup (at equal clocks) was between +6.1% and +13.3%, with the average perf gain for SLI coming in at around +8.5% improvement. As one might expect, given that the primary benefit of the 2x280GTX setup would be it's 15% memory bandwidth advantage, the gains of the SLI setup were more pronounced in the two sets of tests where AA was ON. The ability of the GTX280 SLI setup to overclock slightly higher also produces a slight improvement in it's margin of victory in the 'max overclocked' tests.

I should also add that there is a fairly consistent trend that shows using an i7 processor DOES raise minimum frame rates, and although it's not universal, I think overall its a phenomenon worthy of some study and consideration.

DETAILED RESULTS:

************ GAMER, 1680x1050, 0xAA TESTS ***************

GAMER 2X @ Map: ambush @ 0 1680 x 1050 AA 0xx
GTX 295n Framerate [ Min: 25.70 Max: 88.25 Avg: 66.41 ]
GTX 295i Framerate [ Min: 37.58 Max: 92.13 Avg: 68.31 ]
% CHANGE P35 ([email protected]) TO X58 ([email protected]) For GTX295 on Stock Clocks = +2.9%

2xGTX280n Framerate [ Min: 26.60 Max: 97.21 Avg: 70.58 ]
2xGTX280i Framerate [ Min: 37.07 Max: 102.84 Avg: 76.95 ]
GTX 295ix Framerate [ Min: 35.68 Max: 96.11 Avg: 72.49 ]
% CHANGE 750SLI ([email protected]) TO X58 ([email protected]) For SLI GTX280 on Stock Clocks =+9.0%
% CHANGE GTX295 TO SLI GTX280 on i7 on Equal Clocks= +6.1%

GAMER 2X @ Map: frost @ 0 1680 x 1050 AA 0xx
GTX 295n Framerate [ Min: 42.66 Max: 75.17 Avg: 59.15 ]
GTX 295i Framerate [ Min: 44.63 Max: 78.79 Avg: 61.53 ]
% CHANGE P35 ([email protected]) TO X58 ([email protected]) For GTX295 on Stock Clocks = +4.0%

2xGTX280n Framerate [ Min: 51.58 Max: 88.94 Avg: 70.30 ]
2xGTX280i Framerate [ Min: 50.60 Max: 88.56 Avg: 69.74 ]
GTX 295ix Framerate [ Min: 41.99 Max: 82.99 Avg: 65.05 ]
% CHANGE 750SLI ([email protected]) TO X58 ([email protected]) For SLI GTX280 on Stock Clocks = -0.8%
% CHANGE GTX295 TO SLI GTX280 on i7 on Equal Clocks= +7.2%

GAMER 2X @ Map: avalanche flythrough @ 8 1680 x 1050 AA 0xx
GTX 295n Framerate [ Min: 36.91 Max: 72.85 Avg: 55.94 ]
GTX 295i Framerate [ Min: 42.73 Max: 76.65 Avg: 58.98 ]
% CHANGE P35 ([email protected]) TO X58 ([email protected]) For GTX295 on Stock Clocks = +5.4%

2xGTX280n Framerate [ Min: 30.47 Max: 89.09 Avg: 70.35 ]
2xGTX280i Framerate [ Min: 45.98 Max: 87.96 Avg: 69.10 ]
GTX 295ix Framerate [ Min: 39.69 Max: 82.88 Avg: 62.96 ]
% CHANGE 750SLI ([email protected]) TO X58 ([email protected]) For SLI GTX280 on Stock Clocks = -1.7%
% CHANGE GTX295 TO SLI GTX280 on i7 on Equal Clocks= +9.7%

************ GAMER, 1680x1050, 4xAA TESTS ***************

GAMER 2X @ Map: ambush @ 0 1680 x 1050 AA 4xx
GTX 295n Framerate [ Min: 25.42 Max: 66.82 Avg: 47.05 ]
GTX 295i Framerate [ Min: 32.85 Max: 70.95 Avg: 45.82 ]
% CHANGE P35 ([email protected]) TO X58 ([email protected]) For GTX295 on Stock Clocks = -2.6%

2xGTX280n Framerate [ Min: 24.50 Max: 75.90 Avg: 55.84 ]
2xGTX280i Framerate [ Min: 35.39 Max: 81.85 Avg: 55.84 ]
GTX 295ix Framerate [ Min: 35.18 Max: 69.58 Avg: 49.27 ]
% CHANGE 750SLI ([email protected]) TO X58 ([email protected]) For SLI GTX280 on Stock Clocks = 0%
% CHANGE GTX295 TO SLI GTX280 on i7 on Equal Clocks= +13.3%

GAMER 2X @ Map: frost @ 0 1680 x 1050 AA 4xx
GTX 295n Framerate [ Min: 31.54 Max: 55.58 Avg: 44.85 ]
GTX 295i Framerate [ Min: 30.47 Max: 57.67 Avg: 45.69 ]
% CHANGE P35 ([email protected]) TO X58 ([email protected]) For GTX295 on Stock Clocks = +1.9%

2xGTX280n Framerate [ Min: 35.60 Max: 69.85 Avg: 54.38 ]
2xGTX280i Framerate [ Min: 38.96 Max: 70.68 Avg: 54.35 ]
GTX 295ix Framerate [ Min: 34.71 Max: 62.52 Avg: 49.40 ]
% CHANGE 750SLI ([email protected]) TO X58 ([email protected]) For SLI GTX280 on Stock Clocks = 0%
% CHANGE GTX295 TO SLI GTX280 on i7 on Equal Clocks= +10%

GAMER 2X @ Map: avalanche flythrough @ 8 1680 x 1050 AA 4xx
GTX 295n Framerate [ Min: 26.80 Max: 56.97 Avg: 42.70 ]
GTX 295i Framerate [ Min: 34.97 Max: 58.77 Avg: 45.53 ]
% CHANGE P35 ([email protected]) TO X58 ([email protected]) For GTX295 on Stock Clocks = +6.6%

2xGTX280n Framerate [ Min: 34.62 Max: 69.54 Avg: 55.38 ]
2xGTX280i Framerate [ Min: 41.89 Max: 69.83 Avg: 54.70 ]
GTX 295ix Framerate [ Min: 37.03 Max: 64.03 Avg: 49.46 ]
% CHANGE 750SLI ([email protected]) TO X58 ([email protected]) For SLI GTX280 on Stock Clocks = -1.2%
% CHANGE GTX295 TO SLI GTX280 on i7 on Equal Clocks= +11%

************ ENTHUSIAST, 1680x1050, 0xAA TESTS ***************
ENTHUSIAST 2X @ Map: ambush @ 0 1680 x 1050 AA 0xx
GTX 295n Framerate [ Min: 13.40 Max: 60.92 Avg: 47.98 ]
GTX 295i Framerate [ Min: 31.51 Max: 63.57 Avg: 48.85 ]
% CHANGE P35 ([email protected]) TO X58 ([email protected]) For GTX295 on Stock Clocks = +1.8%

2xGTX280n Framerate [ Min: 23.41 Max: 71.92 Avg: 55.80 ]
2xGTX280i Framerate [ Min: 18.53 Max: 71.55 Avg: 54.96 ]
GTX 295ix Framerate [ Min: 34.21 Max: 68.10 Avg: 51.71 ]
% CHANGE 750SLI ([email protected]) TO X58 ([email protected]) For SLI GTX280 on Stock Clocks = -1.5%
% CHANGE GTX295 TO SLI GTX280 on i7 on Equal Clocks= +6.3%

ENTHUSIAST 2X @ Map: frost @ 0 1680 x 1050 AA 0xx
GTX 295n Framerate [ Min: 35.25 Max: 56.76 Avg: 46.27 ]
GTX 295i Framerate [ Min: 35.88 Max: 59.99 Avg: 47.94 ]
% CHANGE P35 ([email protected]) TO X58 ([email protected]) For GTX295 on Stock Clocks = +3.6%

2xGTX280n Framerate [ Min: 41.86 Max: 68.68 Avg: 55.46 ]
2xGTX280i Framerate [ Min: 39.72 Max: 68.38 Avg: 54.77 ]
GTX 295ix Framerate [ Min: 39.04 Max: 65.70 Avg: 51.20 ]
% CHANGE 750SLI ([email protected]) TO X58 ([email protected]) For SLI GTX280 on Stock Clocks = -1.2%
% CHANGE GTX295 TO SLI GTX280 on i7 on Equal Clocks= +6.9%

ENTHUSIAST 2X @ Map: avalanche flythrough @ 8 1680 x 1050 AA 0xx
GTX 295n Framerate [ Min: 29.53 Max: 56.18 Avg: 42.34 ]
GTX 295i Framerate [ Min: 32.84 Max: 57.93 Avg: 44.38 ]
% CHANGE P35 ([email protected]) TO X58 ([email protected]) For GTX295 on Stock Clocks = +4.8%

2xGTX280n Framerate [ Min: 36.29 Max: 66.68 Avg: 53.03 ]
2xGTX280i Framerate [ Min: 35.83 Max: 65.52 Avg: 51.53 ]
GTX 295ix Framerate [ Min: 36.90 Max: 62.21 Avg: 47.96 ]
% CHANGE 750SLI ([email protected]) TO X58 ([email protected]) For SLI GTX280 on Stock Clocks = -2.8%
% CHANGE GTX295 TO SLI GTX280 on i7 on Equal Clocks= +7.4%


************ ENTHUSIAST, 1680x1050, 4xAA TESTS ***************

(includes max overclocking tests for GTX295 & GTX280 SLI on i7)

ENTHUSIAST 2X @ Map: ambush @ 0 1680 x 1050 AA 4xx
GTX 295n (576/1242/1998) Framerate [ Min: 3.86 Max: 49.85 Avg: 36.75 ]
GTX 295i (576/1242/1998) Framerate [ Min: 24.76 Max: 50.50 Avg: 36.75 ]
% CHANGE P35 ([email protected]) TO X58 ([email protected]) For GTX295 on Stock Clocks = 0%

2xGTX280n (615/1350/2214) Framerate [ Min: 22.65 Max: 57.30 Avg: 43.15 ]
2xGTX280i (615/1350/2214) Framerate [ Min: 29.84 Max: 57.44 Avg: 42.78 ]
GTX 295ix (615/1350/2214) Framerate [ Min: 25.50 Max: 52.84 Avg: 39.06 ]
% CHANGE 750SLI ([email protected]) TO X58 ([email protected]) For SLI GTX280 on Stock Clocks = -1%
% CHANGE GTX295 TO SLI GTX280 on i7 on Equal Clocks= +9.5%

GTX 295n (666/1458/2376) Framerate [ Min: 5.90 Max: 57.95 Avg: 42.03 ]
GTX 295i (666/1458/2376) Framerate [ Min: 29.23 Max: 59.00 Avg: 42.03 ]
2xGTX280i (684/1458/2376) Framerate [ Min: 30.58 Max: 63.19 Avg: 46.45 ]
% CHANGE P35 ([email protected]) TO X58 ([email protected]) For GTX295 at Max Overclock = 0%
% CHANGE GTX295 TO SLI GTX280 at Max Overclock = +10.5%

ENTHUSIAST 2X @ Map: frost @ 0 1680 x 1050 AA 4xx
GTX 295n (576/1242/1998) Framerate [ Min: 26.49 Max: 43.31 Avg: 35.28 ]
GTX 295i (576/1242/1998) Framerate [ Min: 26.12 Max: 47.92 Avg: 36.66 ]
% CHANGE P35 ([email protected]) TO X58 ([email protected]) For GTX295 on Stock Clocks = +3.9%

2xGTX280n (615/1350/2214) Framerate [ Min: 33.27 Max: 55.89 Avg: 44.12 ]
2xGTX280i (615/1350/2214) Framerate [ Min: 30.43 Max: 55.32 Avg: 43.25 ]
GTX 295ix (615/1350/2214) Framerate [ Min: 27.52 Max: 50.30 Avg: 39.61 ]
% CHANGE 750SLI ([email protected]) TO X58 ([email protected]) For SLI GTX280 on Stock Clocks = -1.9%
% CHANGE GTX295 TO SLI GTX280 on i7 on Equal Clocks= +9.2%

GTX 295n (666/1458/2376) Framerate [ Min: 31.48 Max: 53.19 Avg: 43.52 ]
GTX 295i (666/1458/2376) Framerate [ Min: 31.36 Max: 54.99 Avg: 42.67 ]
2xGTX280i (684/1458/2376) Framerate [ Min: 31.06 Max: 60.91 Avg: 47.51 ]
% CHANGE P35 ([email protected]) TO X58 ([email protected]) For GTX295 at Max Overclock = -1.9%
% CHANGE GTX295 TO SLI GTX280 at Max Overclock = +11.3%

DirectX 9.0c ENTHUSIAST 2X @ Map: avalanche flythrough @ 8 1680 x 1050 AA 4xx
GTX 295n (576/1242/1998) Framerate [ Min: 22.65 Max: 43.22 Avg: 32.87 ]
GTX 295i (576/1242/1998) Framerate [ Min: 26.99 Max: 44.01 Avg: 34.08 ]
% CHANGE P35 ([email protected]) TO X58 ([email protected]) For GTX295 on Stock Clocks = +3.8%

2xGTX280n (615/1350/2214) Framerate [ Min: 31.70 Max: 52.26 Avg: 41.76 ]
2xGTX280i (615/1350/2214) Framerate [ Min: 32.70 Max: 51.34 Avg: 40.52 ]
GTX 295ix (615/1350/2214) Framerate [ Min: 28.87 Max: 47.97 Avg: 37.12 ]
% CHANGE 750SLI ([email protected]) TO X58 ([email protected]) For SLI GTX280 on Stock Clocks = -2.9%
% CHANGE GTX295 TO SLI GTX280 on i7 on Equal Clocks= +9.2%

GTX 295n (666/1458/2376) Framerate [ Min: 22.25 Max: 50.74 Avg: 37.69 ]
GTX 295i (666/1458/2376) Framerate [ Min: 31.15 Max: 51.67 Avg: 39.41 ]
2xGTX280i (684/1458/2376) Framerate [ Min: 29.21 Max: 56.75 Avg: 43.50 ]
% CHANGE P35 ([email protected]) TO X58 ([email protected]) For GTX295 at Max Overclock = +4.6%
% CHANGE GTX295 TO SLI GTX280 at Max Overclock = +10%

 

[TFL Replica]

Good job, that's a lot of work.
It would be nice if you could use that data to generate some charts.

 

[mr.derp]

Wow, you put a lot of work into that +rep. I just can't look through it with out getting a headache

. You should graph the data.

 

[tweaker123]

so basically the core i7 is not a big change in fps wise from the core2duo am i right?

 

[brettjv]

Yeah ... I know it's a lot of numbers, sorry

Anyone got a suggestion for an easy Graphing program (that translates to this board easily)?

I'm already into this for so much time, I gotta tell ya ... but I'd do it ... if enough people show interest.

 

[brettjv]

Quote:

Originally Posted by tweaker123 so basically the core i7 is not a big change in fps wise from the core2duo am i right?

Basically, that is correct, although the accuracy of that statement is largely dependent upon whether Crysis:Warhead, at higher graphical settings, and using DX9, is generally reflective of gameplay experiences across the board.

I personally am guessing that it IS, that's why I used that test

If you have other reasons for upgrading, the i7 can be an awesome improvement in performance for a lot of different things people use computers for. But if you're talking about gaming, pushing the envelope, using newer games with high graphical settings, you will remain, as has always been typical, largely GPU-bound, regardless of which platform you choose.

Or so the evidence suggests to me ...

Put it to you this way ... in comparing a rig with a P35/e8400/4GB dual-channel ddr2 (@4.0GHz) combo, with a GTX295 (at GTX280 speeds), against an x58/i920/6GB tri-channel DDR3 (@3.466GHz) combo with SLI GTX280's, the BIGGEST difference across this battery of Warhead tests was 13.3%, in favor of the x58/SLI rig. And usually the difference was more like 9%.

Granted this was in XP32, and it's still only one application, but ... I'm pretty sure it means something ...

 

[brettjv]

Quick update on the overclocking front ...

Turns out the GTX295 I got today has a bit more 'in it' than the one I had last month (the one I tested on the p35 board, but returned cause it was open box and had some issues I wasn't comfortable with).

Max clocks on that one were 666/1458/2376, but I'm humming along with this new one at 684/1512/2520. Might have to revisit the 'max overclocked' numbers on the above test. These gains might be enough to pull my maxed out 295 to within a couple percentage points of the maxed out SLI rig, we'll have to see. My 3dMark06 difference between the two is now down to around 100 points (22700 vs 22800, roughly).

I have to say I am truly enamored with the GTX295, in general. When you hold it in your hand, it pretty much screams out 'I am Quality' in every sense of the word. The outside is all, like, rubberized and stuff, and it seems to weigh like 5 lbs. The thing is just ... badass, I'm telling ya.

You want one, believe me

 

[Kelethar]

Quote:

Originally Posted by brettjv Yeah ... I know it's a lot of numbers, sorry Anyone got a suggestion for an easy Graphing program (that translates to this board easily)? I'm already into this for so much time, I gotta tell ya ... but I'd do it ... if enough people show interest.

Microsoft Excel has some "meager" charts that might manage.

Great data nonetheless!

 

[brettjv]

So, I spoke a little too soon. Looks like my card isn't really happy at the speeds I quoted above. I'm looking at 675/1476/2484 to be 100% stable. Oh well, it's still a small improvement over the last one of these I had ...

 

[almighty15]

They just confirmed what i've always thought. Unless you are planing on running 2-3 high end graphics card that require alot of CPU though-put you can manage with a C2D or C2Q with no problems.

 

[grunion]

Graph, easy peasy

 

[brettjv]

Yeah, I hear you guys about the graphs, but here's the thing ...

The reality is, all these numbers are SO friggin close to one another ... all you're really gonna see on the graphs is that there's only very small differences between a GTX295 and 2xGTX280, regardless of whether you run them on a 775 or a i7. That's the bottom-line. All the graphs are just gonna be a bunch of lines that all look nearly identical.

So doing the work seems a little pointless, at least for this particular comparison

Tell ya what, I'll make ya a graph, real quick. Here's what all these numbers say, on average:

GTX295/e8400 (P35, 4.0GHz)
************************************************** *****(50)
GTX295/i7 (X58, 3.467GHz)
************************************************** *******(52)
SLI280/e8400 (750SLI, 4.0GHz)
************************************************** **********(55)
SLI280/i7 (X58, 3.467GHz)
************************************************** *********(54)

There ya go. That's a pretty accurate representation of the average of all those numbers, in a graphical chart.

 

[TFL Replica]

Quote:

Originally Posted by brettjv Yeah, I hear you guys about the graphs, but here's the thing ... The reality is, all these numbers are SO friggin close to one another ... all you're really gonna see on the graphs is that there's only very small differences between a GTX295 and 2xGTX280, regardless of whether you run them on a 775 or a i7. That's the bottom-line. All the graphs are just gonna be a bunch of lines that all look nearly identical. So doing the work seems a little pointless, at least for this particular comparison

Easily resolved by setting the right min/max values.

 

[grunion]

Quote:

Originally Posted by brettjv Yeah, I hear you guys about the graphs, but here's the thing ... The reality is, all these numbers are SO friggin close to one another ... all you're really gonna see on the graphs is that there's only very small differences between a GTX295 and 2xGTX280, regardless of whether you run them on a 775 or a i7. That's the bottom-line. All the graphs are just gonna be a bunch of lines that all look nearly identical. So doing the work seems a little pointless, at least for this particular comparison Tell ya what, I'll make ya a graph, real quick. Here's what all these numbers say, on average: GTX295/e8400 (P35, 4.0GHz) ************************************************** *****(50) GTX295/i7 (X58, 3.467GHz) ************************************************** *******(52) SLI280/e8400 (750SLI, 4.0GHz) ************************************************** **********(55) SLI280/i7 (X58, 3.467GHz) ************************************************** *********(54) There ya go. That's a pretty accurate representation of the average of all those numbers, in a graphical chart.

I wonder where a Kentsfield at 3.6 would fit in?

All the tests in XP, correct?

 

[brettjv]

Oooh, thanks for addin teh cullurz, grun. That looks real purdy now

I'm betting that a [email protected] would look exactly the same as an [email protected], on these tests, anyways.

Yup, XP32, correct ...

 

[grunion]

Here are some numbers for you to toss around.
Seems you have some strange minimum fps, never once did my fps drop into the teens.

Code:

//////////// Summary \\

03/25/2009 23:06:38 - Microsoft® Windows Vista™ Ultimate  64-bit Service Pack 1

DirectX 9.0c GAMER 2X @ Map: ambush @ 0 1600 x 1200 AA 0x 
==> Framerate [ Min: 33.38 Max: 93.31 Avg: 67.50 ]

DirectX 9.0c GAMER 2X @ Map: frost @ 0 1600 x 1200 AA 0x 
==> Framerate [ Min: 54.42 Max: 89.28 Avg: 72.25 ]

DirectX 9.0c GAMER 2X @ Map: avalanche flythrough @ 8 1600 x 1200 AA 0x 
==> Framerate [ Min: 38.16 Max: 84.03 Avg: 68.08 ]

DirectX 9.0c GAMER 2X @ Map: ambush @ 0 1600 x 1200 AA 4x 
==> Framerate [ Min: 32.03 Max: 80.50 Avg: 53.99 ]

DirectX 9.0c GAMER 2X @ Map: frost @ 0 1600 x 1200 AA 4x 
==> Framerate [ Min: 35.67 Max: 62.91 Avg: 50.14 ]

DirectX 9.0c GAMER 2X @ Map: avalanche flythrough @ 8 1600 x 1200 AA 4x 
==> Framerate [ Min: 37.76 Max: 60.93 Avg: 50.32 ]

DirectX 9.0c ENTHUSIAST 2X @ Map: ambush @ 0 1600 x 1200 AA 0x 
==> Framerate [ Min: 31.84 Max: 74.39 Avg: 58.46 ]

DirectX 9.0c ENTHUSIAST 2X @ Map: frost @ 0 1600 x 1200 AA 0x 
==> Framerate [ Min: 42.73 Max: 67.07 Avg: 55.62 ]

DirectX 9.0c ENTHUSIAST 2X @ Map: avalanche flythrough @ 8 1600 x 1200 AA 0x 
==> Framerate [ Min: 39.39 Max: 63.41 Avg: 51.86 ]

DirectX 9.0c ENTHUSIAST 2X @ Map: ambush @ 0 1600 x 1200 AA 4x 
==> Framerate [ Min: 28.81 Max: 57.37 Avg: 41.76 ]

DirectX 9.0c ENTHUSIAST 2X @ Map: frost @ 0 1600 x 1200 AA 4x 
==> Framerate [ Min: 29.77 Max: 50.51 Avg: 39.76 ]

DirectX 9.0c ENTHUSIAST 2X @ Map: avalanche flythrough @ 8 1600 x 1200 AA 4x 
==> Framerate [ Min: 30.09 Max: 45.90 Avg: 37.74 ]

 

[brettjv]

I did have a few oddly low min-fps runs with the old gtx295, when I did tests on the P35. Not sure if they were card-related, driver-related (this was like 1.5 months back, the gtx was brand new), or platform-related ... but I did take that card back, and now this one doesn't have those really low dips. THanks for sharing, btw