NVIDIA GTX295 Preview/Review

[Stukov]

If will be vertical axis from ZERO, you didnot to recognise difference, point will be at same place ... because this in some games are vertical axis above zero ... dont read axis but NUMBERS!

dinos: Vantage 3D Perf: with Core i7 3600MHz

X2 - 16 927
280 - 12 276
295 - 18 971
260 - 10 715
4870 - 10084

Orb, with respect, the graph is very misleading.

The mathmatical difference in FPS is 1.1% yet your graph spacing shows a 20% distance in space. (~63.5 pixels of space in a ~317 pixel graph) That is a 19% divergence from what the actual spacing should be and what you show in your graph.

You can't just say "You're stupid if you don't read the numbers" as the graph is misleading no matter what.

[X.T.R.E.M.E_ICE]

Orb, with respect, the graph is very misleading.

The mathmatical difference in FPS is 1.1% yet your graph spacing shows a 20% distance in space. (~63.5 pixels of space in a ~317 pixel graph) That is a 19% divergence from what the actual spacing should be and what you show in your graph.

You can't just say "You're stupid if you don't read the numbers" as the graph is misleading no matter what.

Graphs look fine to me.

[moogle]

LOL lonog time didn't see this much lame graph (NHF OBR):

Lol, most misleading graph I've seen in a while. Which is why you always have graphs start from 0.

Graphs look fine to me.

Misleading Graphs

Others disagree

[dan7777]

im very surprised that the gtx295 only just beats 4870x2 by a few fraps i think people with there 4870x2 cards will be happy to see this....

[spursindonesia]

Yeah, just ONE FPS faster -which is within benching error margin- and the graph makes it seems got beaten to a pulp and left 4 dead -pun intended.

[moogle]

Yeah, just ONE FPS faster -which is within benching error margin- and the graph makes it seems got beaten to a pulp and left 4 dead -pun intended.

It's alright we'll all get branded ATi fans

[Shintai]

289w Tdp Card!

[moogle]

289w Tdp Card!

Might not be. GTX280 is 230W~ but when testing a system on load with one of these (3DMark) the wattage from the power meter is suprisingly lower than expected.

[tictac]

LOL lonog time didn't see this much lame graph (NHF OBR):

Graph... fail

[X.T.R.E.M.E_ICE]

It's alright we'll all get branded ATi fans

With a few more driver revision x2 will get flogged.

Graph... fail

Only if you didn't go to school.

[tictac]

With a few more driver revision x2 will get flogged.


Only if you didn't go to school.

Thanks

[Stukov]

Only if you didn't go to school.

So how do you explain 20% difference in graph space with 1.1% difference in performance?

[maroon1]

Did any one check this
http://www.hardwarecanucks.com/forum...preview-3.html

They were using ATI 8.12 WHQL for ATI HD4870 X2, yet GTX 295 performed better in almost all cases

results are different in some games for example gurus crysis review and hardwarecanucks are totally opposite

Gurus was using 180.88 Beta while hardwarecanucks was using 180.87 for GTX 295

[moogle]

With a few more driver revision x2 will get flogged.

It'd be nice to see.

So how do you explain 20% difference in graph space with 1.1% difference in performance?

Yeah, if the graphs are gonna be drawn like that might as well have put it in a table.
I don't look at graphs for the values and work the difference out myself I expect the graph to show it, and fairly

[gallag]

Whats the big deal with the graph, The numbers are clear to see so its not lying and if you are worrying that "average joe" will be "tricked" then don't, Just because average joe is not a computer geek does not mean he cannot read a graph and "average joe" will not even be looking at these graphs anyway because he is not a computer geek.

Some of you guys need to get over yourselves.

[maroon1]

Yeah, just ONE FPS faster -which is within benching error margin- and the graph makes it seems got beaten to a pulp and left 4 dead -pun intended.

Did you say 1 fps ?

[Stukov]

Whats the big deal with the graph, The numbers are clear to see so its not lying and if you are worrying that "average joe" will be "tricked" then don't, Just because average joe is not a computer geek does not mean he cannot read a graph and "average joe" will not even be looking at these graphs anyway because he is not a computer geek.

Some of you guys need to get over yourselves.

A graph exists to see a graphical representation of the data. If you say "well only the numbers matter" and are completely removing the accuracy of the graph then there is no need to make a graph in the first place.

That is like saying when documenting a crime scene that only the data is important so instead of taking a picture with a digital camera you finger paint the crime scene.

[moogle]

Now theres some people who know how to make a graph

Nice looking one too

[Solus Corvus]

Graph... fail

I agree that the graph is horrible. There is a mere 1 fps difference and yet the lines are miles apart when they should be practically on top of each other in a realistic visual representation of the performance difference.

But with that said, it's kinda hard to entertain complaints of intentional misleading. He not only clearly labeled the axes but labeled the individual data points as well. Seriously, who only looks at the colored lines and bars without looking at the words and numbers? Idiots?

Did you say 1 fps ?

[2560x1600 result]

[2560x1600 result]

[2560x1600 result]

[1600x1200 result]

lol.

[Stukov]

I never accused anyone of making graphs for whatever reason, others may have, I was merely saying they are very misleading.

[labs23]

Yep, misleading for no apparent reason. Well, seems to me they put nV on top by one point-only for that specific graph though. I just hope that 1 point is really worth it, but its obviously on the margin of error zone, as others stated.

[AbelJemka]

Did you say 1 fps ?

[2560x1600 result]

[2560x1600 result]

[2560x1600 result]

[1600x1200 result]

Most powerful card on the market but no maxumum game settings?
Bit-tech review
Fallout 3 : 2560x1600 8AA+16AF Max Detail
4870X2 73.6 avg 23.0 min
295GTX 48.3 avg 14.0 min
Farcry 2 : 2560x1600 4AA DX10 Ultra High (versus Very high for Hcanucks)
4870X2 41.4 avg 29.0 min
295 GTX 44.8 avg 29.0 min
Left For Dead : 2560x1600 8AA+16AF Max Detail
4870X2 67.8 avg
295GTX 72.3 avg

Funny you use that particular graph for Crysis Warhead
I can see why

[Jowy Atreides]

Only if you didn't go to school.

Correction; Only if you attended an inferior school.

The business school where I learned Quantative Analysis would cry at these graphs

[jasonelmore]

Most powerful card on the market but no maxumum game settings?
Bit-tech review
Fallout 3 : 2560x1600 8AA+16AF Max Detail
4870X2 73.6 avg 23.0 min
295GTX 48.3 avg 14.0 min
Farcry 2 : 2560x1600 4AA DX10 Ultra High (versus Very high for Hcanucks)
4870X2 41.4 avg 29.0 min
295 GTX 44.8 avg 29.0 min
Left For Dead : 2560x1600 8AA+16AF Max Detail
4870X2 67.8 avg
295GTX 72.3 avg

Funny you use that particular graph for Crysis Warhead
I can see why

Not many of us have a 30" Monitor. Not because we can't afford it but because of it's shear size and lag. 1920X1080 is the most important resolution. Those of you who want the 30" Monitor res's need to opt for something different.

The whole reason nvidia kept the frame buffer down instead of going with the 280's memory size and buffer, was to keep the voltage down. The extra memory would have put this card at over 320 watt TDP

[SimBy]

This card would be nice 5 moths ago. Way too late and not a whole lot better than X2. So basically it's a fail. I'll wait for the next gen from NV and ATI thank you.