The Fermi Thread - Part 3

[Musho]

ok with small work on Excel with the fx AVERAGE
[IMG.]http://img339.imageshack.us/img339/9306/avge.jpg[/IMG]

I can say GTX480 is faster the HD5870 by 50% on average

50% lead would be really great. Would give me the itch to upgrade from my 2x 5870 in crossfire Computers is such an expensive hobby

[ElSel10]

'I can say GTX480 is faster the HD5870 by 50% on average' don't worry if that was true case nvidia would priced it above 5970 lol gonna wait for bettter reviews

Perhaps not if they want to steal ATI sales? After being delayed against HD5000 for ~6 months perhaps Nvidia is coming back with a vengeance?

[aver]

ok with small work on Excel with the fx AVERAGE


I can say GTX480 is faster the HD5870 by 50% on average

However if you look at specific games such as Dirt 2 the benchmarks shows under 30% average advantage.

Breakdown if numbers are true:

Heaven - 193% Average Advantage (780% advantage in one score though)

Neutrone - (-10%) average advantage

Dirt 2 - 27.5% advantage

Battleforge - 36.5% advantage

Stalker - 19.1% advantage

3D Mark Vantage - .3% advantage

[Zed_X]

MAX 96C ahhhhaaa thats hot....

Now i know why the 1.8A fan is installed, in the summers it gonna make your pc fly...

The TDP of 230w does not make sense, i mean max 96C with that huge a cooler and 92C avg on 230w tdp ??

ohhhhhhhhhh, thats horrible temps! Only 5C above HD5970

and 10C above HD 5870, oh GOD, how can be someone too blind and trolly?

[96redformula]

50% lead would be really great. Would give me the itch to upgrade from my 2x 5870 in crossfire Computers is such an expensive hobby

I don't understand how nVidia is not doing HDMI passthrough like ATI. No matter how much better performance it gets I need that feature .

Performance wise those numbers posted would be incredible if true...

[eric66]

Perhaps not if they want to steal ATI sales? After being delayed against HD5000 for ~6 months perhaps Nvidia is coming back with a vengeance?

they still have same ms why sell with loss if you have much better hardware ? physx cuda stuff already makes nvidia products atractive don't know its unbelievable if true then kudos to nvidia they are again my next gc

[mindfury]

ohhhhhhhhhh, thats horrible temps! Only 5C above HD5970

and 10C above HD 5870, oh GOD, how can be someone too blind and trolly?

Noise,5870 fan doesnt speed up until 80C

[ElSel10]

they still have same ms why sell with loss if you have much better hardware ? physx cuda stuff already makes nvidia products atractive don't know its unbelievable if true then kudos to nvidia they are again my next gc

Where is there any concrete evidence that Fermi cards will be sold at a loss?

[Musho]

ohhhhhhhhhh, thats horrible temps! Only 5C above HD5970

and 10C above HD 5870, oh GOD, how can be someone too blind and trolly?

Why is he trolly and blind? He has a perfectly valid point. That _is_ quite hot. And as mentioned before, the ATI cards don't really ramp the fan up all that much when loaded, simply because it doesn't need to dissipate the same amount of heat as Fermi will have to.

[Chumbucket843]

this link has been posted but...
http://www.amazon.com/Zotac-GeForce-...9600472&sr=1-1
if you look closely at the box you will see some specs.

[Manicdan]

they still have same ms why sell with loss if you have much better hardware ? physx cuda stuff already makes nvidia products atractive don't know its unbelievable if true then kudos to nvidia they are again my next gc

this card isnt going to make up its revenue in a month, its going to be across 2-3 years. they lost alot of time and if at 500$ they have a better price/perf ratio, amd will be forced to lower prices and start giving up that great profit margin, or loose market share.

we all want nvidias card to demolish amds, so prices can go down again, which they havnt done in 10 months.

[jaredpace]

ok with small work on Excel with the fx AVERAGE

snip

I can say GTX480 is faster the HD5870 by 50% on average

You could say it's 880% faster in Heaven, and 10% slower in Crysis too!

@Jared,

where did you find that? is it at newegg?

Found it on hwupgrade.it forums, dunno where it's from. Andrea could probably tell you

[Musho]

You could say it's 880% faster in Heaven, and 10% slower in Crysis too!

He shouldn't include the 880 number. It isn't really a good representation, but will greatly skew the results because it's such a big number.

[tajoh111]

The driver gains are going to be insane with fermi. I have never seen a card with such inconsistent performance. Forget 10% performance gains, we are going to be seeing 20-30 percent performance gains. Especially with titles like Quake and crysis, which are typically NV favoured, the heaviest gains are going to be seen here. I have guessed that when making the latest drivers NV focus on their weakness first and will later focus on their strengths. With strong performances in games like dirt 2, hawx, battleforge and unreal, which are usually AMD favoured, this card should be killing in games like crysis and quake wars. Its not those and at this point one has to suspect drivers(assuming the recent leaked charts are true).

[kache]

this link has been posted but...
http://www.amazon.com/Zotac-GeForce-...9600472&sr=1-1
if you look closely at the box you will see some specs.

Nice find!

[eric66]

Why is he trolly and blind? He has a perfectly valid point. That _is_ quite hot. And as mentioned before, the ATI cards don't really ramp the fan up all that much when loaded, simply because it doesn't need to dissipate the same amount of heat as Fermi will have to.

that heat thing is what disturbs me 96 is pretty high and in warm climates it will be even higher making card unoverclockable same problem which i had with my ultra hope 470 is more cooler seems thats my only choice

[jmke]

^mindfury

Are those percentages for % faster than the 5870? If so, then WOW.

not really ; if you see 100% it as fast as HD 5870, if you see 170%, it's 70% faster than HD 5870. Check the Crysis % compare for proof ;-)

[trinibwoy]

Well that would be a problem if you were considering buying one? Early numbers don't look too bad, certainly not anywhere near the R600/NV30 predictions we've been hearing for 6 months. Would be nice to know how AMD feels about it - who knows what they have waiting in the wings.

[mattlef]

The only thing I'm really wondering about is the 2d downclocking.
Do you think nVidia put this feature into the 4series chips like the ati 5 series? this is a huge selling feature for me.

[cegras]

Percentages are so stupid. If your test is 10 fps and the other card gets 20, that's a factor of 2.0, but clearly both are unplayable. Furthermore, why are we factoring synthetics into 'real-world' tests?

[trinibwoy]

The only thing I'm really wondering about is the 2d downclocking.
Do you think nVidia put this feature into the 4series chips like the ati 5 series? this is a huge selling feature for me.

GT200 already downclocks really low in 2D. Don't know why it would be any different with Fermi. Idle numbers still look very high though. Leakage is a biatch.

[mindfury]

The only thing I'm really wondering about is the 2d downclocking.
Do you think nVidia put this feature into the 4series chips like the ati 5 series? this is a huge selling feature for me.

According to Chiphell benchmark,GTX 480 and HD5870 have 45W difference at idle.

[Musho]

Won't chiphell get sued for these benchmarks if they turn out real? I mean, didn't they sign a NDA which probably means you have to pay 200k+ if you break it?

[Manicdan]

According to Chiphell benchmark,GTX 480 and HD5870 have 45W difference at idle.

ouch, that will hurt.

[jmke]

Percentages are so stupid. If your test is 10 fps and the other card gets 20, that's a factor of 2.0, but clearly both are unplayable. Furthermore, why are we factoring synthetics into 'real-world' tests?

if you remove all unplayable results (<30fps) from the charts and keep only the game benchmarks; you end up with ~28.7% increase in performance over HD 5870.

granted that this card is overkill for 1680x1050, and also 1920x1200 for most games, even with 8xAA.