Thursday, October 21st 2010

GeForce GTX 580 Expected to be 20% Faster than GeForce GTX 480

NVIDIA's next enthusiast-grade graphics processor, the GeForce GTX 580, based on the new GF110 silicon, is poised for at least a paper-launch by end of November, or early December, 2010. Sources in the video card industry told DigiTimes that the GTX 580 is expected to be 20% faster than the existing GeForce GTX 480. The new GPU is built on the existing 40 nm process, NVIDIA's 28 nm GPUs based on the Kepler architecture are expected to take shape only towards the end of 2011. Later this week, AMD is launching the Radeon HD 6800 series performance graphics cards, and will market-launch its next high-end GPU, codenamed "Cayman" in November.Source: DigiTimes
Add your own comment

98 Comments on GeForce GTX 580 Expected to be 20% Faster than GeForce GTX 480

#1
FordGT90Concept
"I go fast!1!11!1!"
20% hotter and more power draw too? XD
Posted on Reply
#2
assaulter_99
by: FordGT90Concept
BTW, the "to" doesn't sound right in the title.
GeForce GTX 580 Expected to be 20% Faster than GeForce GTX 480

Is that ok? :laugh:

Ok, so you owe me one BTA! ;)
Posted on Reply
#3
wahdangun
maybe what they mean was the "wooden" version will be launching at December
Posted on Reply
#4
HXL492
This is interesting, wonder how the cayman would fare.
Posted on Reply
#5
983264
Before NV says that, they must beat the 5970 in dual gpu segment...
Posted on Reply
#6
motasim
... and exactly how many nuclear power stations are required to power it ... :rolleyes:

... but look at the bright side; you won't need any gas cooker after you get this GPU ... ;)
Posted on Reply
#7
Atom_Anti
Haha, the new GrillForce is coming! Chefs going to like it:laugh:.
Posted on Reply
#8
Fourstaff
Hmmm, If you would consider GTX480 with 10% higher clock speed and all 512 cores.....
Posted on Reply
#9
entropy13
by: Fourstaff
Hmmm, If you would consider GTX480 with 10% higher clock speed and all 512 cores.....
And it's an entirely new generation!:laugh:
Posted on Reply
#10
Red_Machine
by: btarunr
poised for at least a paper-launch by end of November, or early December, 2010.
What? THAT soon?
Posted on Reply
#11
20mmrain
So let me get this straight Nvidia is going to launch a Architecture early that wasn't supposed to be launched until 2011 and they couldn't get Fermi out on time.... yeah right!
Posted on Reply
#12
HardSide
by: Red_Machine
What? THAT soon?
It's just a paper launch, they need to counter ATI new card that they announced. Been a nvidia fan boy for a while now, but if they dont fix the problems that were in the 400 series, they going to lose a lot of fan base.
Posted on Reply
#13
wahdangun
by: Red_Machine
What? THAT soon?
yeah just like fermi, they just paper dragon until six month later
Posted on Reply
#14
DanTheMan
And as the story goes ... After the 580 series, 1 month later AMD will answer back with a 15% increase, then NVidia will promise a 680 series (another 10% gain) then AMD .........

Same story .... over and over ..... It never ends folks!
Posted on Reply
#15
derwin75
This is good but it's only words. Action speaks louder than words. All I can do is wait and see when it comes out.
Posted on Reply
#16
Bjorn_Of_Iceland
by: DanTheMan
And as the story goes ... After the 580 series, 1 month later AMD will answer back with a 15% increase, then NVidia will promise a 680 series (another 10% gain) then AMD .........

Same story .... over and over ..... It never ends folks!
Its a half empty / half full conundrum :D
Posted on Reply
#18
Benetanegia
GF104 = 384 SP = 2 x GPC

GF110 = 3 x GPC = 576 SP

576/480 = 1.2 = +20% :eek:

Power consumption on Furmark

GTX460 = 155 W = 336 SP

GF110 = (155 W x 576 SP) / 336 SP = 265 W



iGame GTX460 = 141 W
GF110(?) = (141 x 576) / 336 = 241 W
Posted on Reply
#19
SNICK
oh my god only 20% faster than gtx 480!:)
there were also speculation about doubled gf100:laugh:

now there is no way nvidia can steal AMD's thunder
Posted on Reply
#21
CharlO
Wait, 20% faster than a 480 is like a 5870 oc'd (well something more) but it is definetly slower than a promised 6870... They are not even getting even on paper now?
Posted on Reply
#22
SNICK
by: charlo
wait, 20% faster than a 480 is like a 5870 oc'd (well something more) but it is definetly slower than a promised 6870... They are not even getting even on paper now?
i think you mean 6970
Posted on Reply
#23
Benetanegia
by: SNICK
fyi . Tdp is dependent on several factors other than ONLY shaders count
FYI, Fermi cards are based only on 2 type of clusters, GPC and ROP partitions, every card is just a conbination of the two. .

GPC = SPs + TMUs + rasterizers + setup engines + tesselators (Polymorph) + L1 cache

ROP partition = ROPs + memory controllers + L2 cache

When something increases everything increases by the same amount as long as the GPC is equal. GF100 and GF104 GPCs are different.

You want me to speculate power consumption based on ROP partitions?

GF104 = 2 ROP partitions = 155 W
GF110 = 3 ROP partitions = (155 x3) / 2 = 232 W

Other than silicon increase the only thing that affects power consumption is clock speed which I speculated is going to be the same more or less, and that it won't affect too much: as can be seen on the chart I provided @820 Mhz (quite an OC) the power is 176 W.
Posted on Reply
#25
SNICK
by: Benetanegia
FYI, Fermi cards are based only on 2 type of clusters, GPC and ROP partitions, every card is just a conbination of the two. .

GPC = SPs + TMUs + rasterizers + setup engines + tesselators (Polymorph) + L1 cache

ROP partition = ROPs + memory controllers + L2 cache

When something increases everything increases by the same amount as long as the GPC is equal. GF100 and GF104 GPCs are different.

You want me to speculate power consumption based on ROP partitions?

GF104 = 2 ROP partitions = 155 W
GF110 = 3 ROP partitions = (155 x3) / 2 = 232 W

Other than silicon increase the only thing that affects power consumption is clock speed which I speculated is going to be the same more or less, and that it won't affect too much: as can be seen on the chart I provided @820 Mhz (quite an OC) the power is 176 W.
your whole mathemetics based on unitary method is not supposed to be used on electricals components.TDP of FULL gf100 is around 204 watts more than gf100.based on your mathematics can you explain why the TDP OF FULL GF100 IS QUITE MORE AS EXPECTED?
Posted on Reply
Add your own comment