GF106 die shot and 512sp GF100

**kaktus1907** · 08-07-2010, 07:58 AM

more benchies vs GTX 480 480sp

Source: EXP

**kaktus1907** · 08-09-2010, 08:08 AM

MADNESS!

http://en.expreview.com/2010/08/09/w...-480/9070.html

**AKM** · 08-09-2010, 08:12 AM

SO the full chip sucks 204 watts more Holy Cow!

**K404** · 08-09-2010, 08:20 AM

ROFL BS it has to be a typo

7% more SP is not gonna add 200W more onto the power consumption. No HSF could handle it.

Y'know what... say for the sake of discussion it IS real..... IMHO there should be a law against bringing out such inefficient electronic equipment.

**Manicdan** · 08-09-2010, 08:29 AM

Originally Posted by K404

ROFL BS it has to be a typo

7% more SP is not gonna add 200W more onto the power consumption. No HSF could handle it.

Y'know what... say for the sake of discussion it IS real..... IMHO there should be a law against bringing out such inefficient electronic equipment.

typo no, user error seems more likely (unless the guy wrote down the wrong number and never thought twice about his findings)

Under full load the GPU voltage of reference GTX 480 was 1.0V, while 512SP edition was 1.056V. Surprisingly, the full spec’ed GTX 480 sucked 644W power, which was 204W higher than 480SP GTX 480!

the fact they talk about it using so much more shows thats what they are seeing, but then comes the question of why. just by knowing the C/W of whatever cooler they are using, and that the load temps are 94C, its clearly not that much more than 480SPs

but their temp chart is a little funny looking too,

notice how the temp goes up almost linear to 94C, and the fan speed tries to follow in an almost liner fashion, and does not get much faster after reaching 94C was first reached. and that somehow 94C is a perfectly flat line

im worried its way past 94C

**K404** · 08-09-2010, 09:30 AM

What about 464W... the 644 is a mis-type?

Another 24W I can just about stomach.

**Russian** · 08-09-2010, 10:30 AM

Originally Posted by saaya

well the latest rumors is that nvidia is reserving LESS capacity at tsmc for the coming months... odd huh?

Heh, I know why.

**ElSel10** · 08-09-2010, 10:37 AM

As previously mentioned... the HSF would not be able to handle an additional 200W. It would just die... Probably 464, not 644.

**cowie** · 08-09-2010, 11:32 AM

Originally Posted by Manicdan

but their temp chart is a little funny looking too,

notice how the temp goes up almost linear to 94C, and the fan speed tries to follow in an almost liner fashion, and does not get much faster after reaching 94C was first reached. and that somehow 94C is a perfectly flat line

im worried its way past 94C

Depending on polling rates a "flat line" can vary 2-3c or more.
The whole thing looks odd to me,i'll wait a few weeks to find out what the card really has.
At this point we know it lives lol

**cold2010** · 08-27-2010, 09:02 PM

GTS 450

**El Mano** · 08-27-2010, 09:10 PM

Originally Posted by Russian

Heh, I know why.

Can you share your thoughts with us?

***
I'm very curious about the full 512 GF100. Specially temps and power.

**570091D** · 08-27-2010, 10:30 PM

hrm, i expected the gts450 to be right behind the gtx460... what's with the cheap-o power regulation components?

btw, the 512 gf100 (as we know it now) will never see the light of day. the power figures shown in the "review" aren't representative of a consumer part and i don't believe that is the reason we can't buy it. the performance of the 512 core chip simply doesn't justify the cost to produce it, if there is only 5% advantage over the 480 core chip and it costs an extra $100 to make it... who would buy such a thing (other than xfx 5970 BE customers...)? nvidia isn't going to loose money to bring a part to market that can't even bring them the performace crown.

**STEvil** · 08-28-2010, 12:18 AM

its the same part and pcb as a 480, why would it cost more to make it? They can slap $50 on it and sell it as the GTX+ model.

**Sn0wm@n** · 08-28-2010, 12:43 AM

Originally Posted by Oliverda

GF108

it looks photoshopped :S well the lettering at least

**kaktus1907** · 08-28-2010, 01:22 AM

Originally Posted by cold2010

GTS 450

hmmm 2 missing memory chips and KA code on gpu.. maybe a crippled version again ? i am not sure it is GTS450

**Sn0wm@n** · 08-28-2010, 01:40 AM

anyway .. the whole ferbi saga is looking more and more ridiculous ... fake 512 shader gpu news + pics of suposedly ready lower end model wich looks photochopped ....

**Olivon** · 08-28-2010, 01:47 AM

Originally Posted by Sn0wm@n

anyway .. the whole ferbi saga is looking more and more ridiculous ... fake 512 shader gpu news + pics of suposedly ready lower end model wich looks photochopped ....

Waowww, too bad nVidia, nasty boy !!!

**Sn0wm@n** · 08-28-2010, 02:07 AM

so you think that the numbers and pics about the 512shader gf100 are 100% true and the pics of the pcb for the gf106 the truth ???? both of them look a bit fairy tale like too me ... so that is the reason why i made that post ...

**JohnZS** · 08-28-2010, 04:27 AM

I think these are all the latest GF1xx parts fresh out of WHQL Testing...

NVIDIA_DEV.0DC0.01 = "NVIDIA GeForce GT 440"
NVIDIA_DEV.0DC1.01 = "NVIDIA D12-P1-35"
NVIDIA_DEV.0DC4.01 = "NVIDIA GeForce GTS 450"
NVIDIA_DEV.0DC5.01 = "NVIDIA GeForce GTS 450"
NVIDIA_DEV.0DDF.01 = "NVIDIA GF106-INT"
NVIDIA_DEV.0DE1.01 = "NVIDIA GeForce GT 430"
NVIDIA_DEV.0DE2.01 = "NVIDIA GeForce GT 420"
NVIDIA_DEV.0DFF.01 = "NVIDIA GF108 INT"
NVIDIA_DEV.0E21.01 = "NVIDIA D12U-25"
NVIDIA_DEV.0E22.01 = "NVIDIA GeForce GTX 460"
NVIDIA_DEV.0E23.01 = "NVIDIA GeForce GTS 455"
NVIDIA_DEV.0E24.01 = "NVIDIA GeForce GTX 460"
NVIDIA_DEV.06C0.01 = "NVIDIA GeForce GTX 480"
NVIDIA_DEV.06C3.01 = "NVIDIA D12U"
NVIDIA_DEV.06C4.01 = "NVIDIA GeForce GTX 465"
NVIDIA_DEV.06CD.01 = "NVIDIA GeForce GTX 470"

**The Coolest** · 08-28-2010, 09:31 AM

Originally Posted by cold2010

GTS 450

http://img192.imageshack.us/img192/3143/89820703.jpg

http://img834.imageshack.us/img834/3188/40742991.jpg

http://img829.imageshack.us/img829/4602/36406083.jpg

Is it just me or does that heatsink look like an old P4 cooler?

**cold2010** · 08-28-2010, 09:43 AM

Originally Posted by The Coolest

Is it just me or does that heatsink look like an old P4 cooler?

**highoctane** · 08-28-2010, 09:49 AM

Originally Posted by The Coolest

Is it just me or does that heatsink look like an old P4 cooler?

Actually it looks like the same cooler that came with all my Intel chips since core2, minus the copper slug.

**spursindonesia** · 08-28-2010, 12:31 PM

well, the die sizes of nVidia GF 10x chips are not exactly small relative to its competitor, so cost must be made up & reduced somewhere.

**570091D** · 08-28-2010, 01:39 PM

Originally Posted by STEvil

its the same part and pcb as a 480, why would it cost more to make it? They can slap $50 on it and sell it as the GTX+ model.

the pcb shouldn't be the same, the extra cluster should include a wider mem bus and more mem.... i just don't think the 5% performance gain justifies adding another card to the top of thier line-up, unless it's going to beat the 5970.

**spursindonesia** · 08-28-2010, 02:06 PM

Originally Posted by 570091D

the pcb shouldn't be the same, the extra cluster should include a wider mem bus and more mem.... i just don't think the 5% performance gain justifies adding another card to the top of thier line-up, unless it's going to beat the 5970.

Incorrect bro, because GTX 480 is already equipped with all of GF 100 available ROP & buswidth (48 ROP & 384 bit respectively). If any, it's the power phase sector of the PCB that would need to be strengthened, quite considerably so if the last leak info from expreview can be trusted (considering nVidia's sh1tty GF 100 yield, if they want to maximize the amount of perfect GF 100 available, we could expect 325-350 w TDP of a 512 SP GF 100 part, and that might be conservative !

).

Thread: GF106 die shot and 512sp GF100

Thread Tools

Search Thread

Rate This Thread

Display

Bookmarks

Bookmarks

Posting Permissions