The Fermi Thread - Part 3

[Mungri]

Let's not forget a 4870 under load:


4870 CF under load:


And before you try and say it's irrelevant because it's from last generation, no it isn't. If you were fine with HD4000 cards then the only reason you wouldn't be fine with GF100 would be unreasonable bias (avoiding the f-word).

Yes this is true. My top 4870 burnt out from excessive temperatures in crossfire mode. They are horrible cards and I wouldnt reccomend them to anyone, and the same goes to the Fermi.

ATI sorted out these thermal issues from the 4890 / 4770 onwards, something that Nvidia forgot to do.

[zalbard]

Sorry, but use a thermal camera with a card with full cover backplate -> FAIL

This backplate is there to act as a heatspreader, though. So no.

[Dami3n]

This backplate is there to act as a heatspreader, though. So no.

The backplate in 5870/5970 isn´t very good as heatspreader,
http://www.overclock.net/ati-cooling...backplate.html
but you cant see the real thermals behind it, so comparison is really useless in this case.

[Olivon]

How is it possible to blog about GPUs for a living and not know the difference between a ROP and a TMU? And even if he were to get that right, texture cache efficiency increases with resolution, not the other way around. I wonder how he feels about Cypress' 32 ROPs

/boggle

17 march

SemiAccurate moles are saying that plans are in the works to tape out a GF104 and a GF106 in short order as well, but they haven't left Dear Leader's workers' paradise yet.

5 april

SemiAccurate has confirmed that the GF104 did tape out a few months ago.

So ... Semi ... Accurate

[annihilat0r]

17 march



5 april



So ... Semi ... Accurate

[flippin_waffles]

What's even more interesting is the part that you seemed to look over:



Sorry if you wanted to leave that part out, Flip.

Actually I did leave that part out on purpose; to prove a dam point. I could have just as easily made a sensationalist thread and simply ask if they are cheating, based on that discovery. You know, just to make a mountain out of a mole hill.....Seems potentially far worse of a cheat than that little shimmering glitch in Crysis.

BTW, as far as the Crysis crisis, has there been any investigation done to try and find the cause, or is it all still just heresay?

[Sn0wm@n]

if GF100 is having problem with heat i cant imagine GF119 ....

[snoro]

gf119 is just a low end variant of gf100. Something like gefore g210,g220 and gt240 but with fermi architecture so heat shouldbe alot less.

[Sn0wm@n]

GF119 is a higher name then the GTX480 wich is GF100 so sure its a low end fermi .....

[Mad1723]

GF119 is a higher name then the GTX480 wich is GF100 so sure its a low end fermi .....

It's the way Nvidia has worked for the last generations at least. G80 was the 8800GTX, G84 was the 8600GT and the G86 was the 8400GT

[Sn0wm@n]

It's the way Nvidia has worked for the last generations at least. G80 was the 8800GTX, G84 was the 8600GT and the G86 was the 8400GT

so GF119 will be a very very very low end part ????

[highoctane]

so GF119 will be a very very very low end part ????

Does it really matter...

[Mad1723]

so GF119 will be a very very very low end part ????

Probably a low end part, nobody knows yet

[annihilat0r]

if GF100 is having problem with heat i cant imagine GF119 ....

i lol'd

[Sn0wm@n]

Does it really matter...

to someone... im sure it does....

[Jowy Atreides]

The GeForce GTX 480 enjoys an average advantage of 30-37%, depending on the resolution,
......over the previous-generation flagship GeForce GTX 285.

ZING!
http://www.xbitlabs.com/articles/vid..._12.html#sect0

[ElSel10]

The GeForce GTX 480 enjoys an average advantage of 30-37%, depending on the resolution,
......over the previous-generation flagship GeForce GTX 285.

ZING!
http://www.xbitlabs.com/articles/vid..._12.html#sect0

I love how they have a performance comparison chart of the 480 vs the 285 and 5970 but not the 5870, which are far more comparable in price.

[highoctane]

The GeForce GTX 480 enjoys an average advantage of 30-37%, depending on the resolution,
......over the previous-generation flagship GeForce GTX 285.

ZING!
http://www.xbitlabs.com/articles/vid..._12.html#sect0

When full-screen antialiasing is turned on, the difference is 46-48%.

Not sure what ZING infers.

Who buys a 285/480 class card and not use AA.

[japamd]

I love how they have a performance comparison chart of the 480 vs the 285 and 5970 but not the 5870, which are far more comparable in price.

They did

http://www.xbitlabs.com/misc/picture...87_big.png&1=1

[Chickenfeed]

Who buys a 285/480 class card and not use AA.

^^ This.

Misleading context much?

[saaya]

The GeForce GTX 480 enjoys an average advantage of 30-37%, depending on the resolution, over the previous-generation flagship GeForce GTX 285. When full-screen antialiasing is turned on, the difference is 46-48%

a 50% boost is nice, but clearly a far cry from the 100% nvidia claimed... they kept saying 2x the performance of a gtx285... which is not true... even if you allow them to backpedal their claims to refer to a 295... thats a cut down gt200, and its sli, and even THAT beats a 480 slightly...
gtx285 sli is about as fast as a 5970 isnt it?

i like xbitlabs review, too bad they are always 2 weeks behind every other site, but at least they do a really good job every time...

about rmas... didnt gt200 cards have pretty high rma rates too? and they didnt run that hot... i think theres more to this than just heat... fan profiles, heatsink design, heatsink mounting and card assembly... packaging and pcb problems...

i think the main reason nvidia uses that backplate is not as a heatsink, its to stabilize the pressure on the gpu packaging during different temperatures, so that they dont run into solder bump issues and gpu packaging issues again. lets see if it works...

GTS450 specs look nice... close to 5850 performance... but 200-230E for that?
in what, 4-6 months from now? if it would be out now that would be an ok-good price/perf point... but in 1-2 quarters?
we will see...

[Mechanical Man]

GTS450 specs look nice... close to 5850 performance... but 200-230E for that?
in what, 4-6 months from now? if it would be out now that would be an ok-good price/perf point... but in 1-2 quarters?
we will see...

Yeah, there is no situation where hd5850 would be over that at the time 450 will hit the stores.. Just look at hd5850 launch price

Fermi is nice if you need power and can keep your house cool enough at summer. For me atleast, even all electronics off, it is too hot in summer..

[highoctane]

a 50% boost is nice, but clearly a far cry from the 100% nvidia claimed... they kept saying 2x the performance of a gtx285... which is not true...

Well there was at least one instance of more than 100%

[Sn0wm@n]

nvidia: one benchmark to rule them all

[Olivon]

King Shamino breaks P31700 Pts with GTX 480 - WR Single GPU Vantage



1190/1350/2380 GPU/RAM/Shaders ~1.35V