GF100 Previews from Digital Experience

[Zstream]

I wasnt sneaking that in. as soon as I typed it I realized that it could certainly be over 225 so i changed it. you can see that even me going back and adding other stuff to my comment was done in less than 2 minutes.

So you believe that Nvidia has come up with a better way to handle power then the ATI 5800 series?

 

[Keysplayr]

Huh? Who said I was claiming my opinion was based on just a 6+8 pin adaptor?

Oh and apparently you are not well informed on the PCIE 2.0 specs.

75+150+75 = ????????????????????? Yeah you guessed it.... 300! 75+150 comes from the 6 & 8 pin alone, add 75 from the slot and well once again it equals 300.

Ok, let me be a bit more clear. The current ATI offerings is hovering around 180-200 Watts, 190W for a surface area of 334mm2. Now IF the 50% more transistor count is accurate then it will certainly creep up to the 250+ range no question.

190W divided by 334mm2 is around .56. If we took the exact same transistor count from ATI and increase the surface area to 530mm2 is around .35.

.35 + .17 (increase of 50%) then we reach around 275watts at full load.

That's all pretty interesting. But I, and maybe some others?, would like to know what exactly a 334mm2 ATI die has to do with an Nvidia 50% larger die? I'm not sure they consume power identically given the same exact die size and transistor count, which is what you're basing this whole thing on I suppose?

May I ask why?

So you believe that Nvidia has come up with a better way to handle power then the ATI 5800 series?

How is he supposed to know that? Do you think it's that far fetched that Nvidia could possibly do something better than ATI can? I wouldn't know yet about this particular aspect. Power handling, but it's pretty easy to see you're writing off Nvidia as having the ability to do so?

:::shrugs:::

 

[toyota]

So you believe that Nvidia has come up with a better way to handle power then the ATI 5800 series?

what?

what is the TDP of the 5870? its 188 watts.
what is the TDP of the 4890? its basically the same at 190watts.

ATI doubles every aspect of thier card and still ends up having the same TDP when going from 55nm to 40nm. why in the heck do you think Nvidia, who is lowering their bus width to 384, is going to be remotely close to 300 watts. just lowering the bus width on the gtx285 would have likely put it around 140 watts even on 55nm. they never did change the TDP rating on the 55nm 216sp gtx260 320bit but it only pulls about 110 watts during gaming.

again, at worst, the top Fermi single gpu card will be around 225watts and I have no doubt about it. I will gladly eat my words if I am wrong so lets just wait and see.

 

[akugami]

Anyone who believes that a card capable of 300W TDP is at 55C is not only silly but incapable of understanding the workings of a GPU. FUD was at the show and stated the FAN was on really high, he stated it was quite noticeable.

The 300W TDP is likely false and I believe from a different article. Likely it's only listed as 300W because that's the maximum power draw for a PCIe part (unless I'm remembering the max power draw wrong). I'm not saying 55C is the correct load temp, again, look at my preface that the stuff I listed from that thread is rumor and not truth. That means take it with a grain of salt. It's just that if true, this would point to a part that would have decent overclocking headroom. People here care about that type of stuff.

no it doesnt.. he can always comeback and say.. we encountered technical diffiulties.. or whatever to cover up the lie.. nobody would know..

True, but if we have to take certain things people say at face value at times. And if you have a history of lies, it tends to bite you in the rear after a while.

 

[SlowSpyder]

Because 6 monitors (2560x1600) 1 or 2 of which require display port are "affordable for the masses.

You are grasping at straws here.

You're purposely acting dense here. Neither AMD nor Nvidia build or sell monitors. You don't game on Quadros. A Quadro that is of about equal horsepower to a given Radeon gaming card will generally cost much, much more. AMD brought the ability to game on three monitors to the mainstream, now it looks like Nvidia might do the same with their next part.

 

[Wreckage]

You're purposely acting dense here. Neither AMD nor Nvidia build or sell monitors.

You do realize you need to buy a monitor to use a video card don't you?

 

[Lonyo]

You're purposely acting dense here. Neither AMD nor Nvidia build or sell monitors. You don't game on Quadros. A Quadro that is of about equal horsepower to a given Radeon gaming card will generally cost much, much more. AMD brought the ability to game on three monitors to the mainstream, now it looks like Nvidia might do the same with their next part.

SLI isn't mainstream.

 

[lopri]

That's all pretty interesting. But I, and maybe some others?, would like to know what exactly a 334mm2 ATI die has to do with an Nvidia 50% larger die? I'm not sure they consume power identically given the same exact die size and transistor count, which is what you're basing this whole thing on I suppose?

No. From what I can see he's basing the whole thing on a simple math that you and some others apparently have trouble to do. A 6-pin plus a 8-pin. Unless there is something else to be explained, that amounts to 225~300W max power.

How is he supposed to know that? Do you think it's that far fetched that Nvidia could possibly do something better than ATI can? I wouldn't know yet about this particular aspect. Power handling, but it's pretty easy to see you're writing off Nvidia as having the ability to do so?

I have no idea why a common sense push some folks' nerves so hard. Again, HD 5870 = 6-pin + 6-pin. A Fermi shown by NV = 6-pin + 8-pin. Unless somehow the world has gone around and 6+6 > 6+8, there is nothing strange to assume the latter being more power-hungry. I am not sure what you mean by "NV could possibly do something better than ATI". Is it something like "Fermi draws more power (6+8) from the PSU yet consumes less power" (?)

Well, obviously that doesn't make sense. And that's what you're saying. And as for power handling, NV's powermizer has never been as capable as AMD's powerplay since RV670. So it isn't surprising that one, especially a consumer, maintains a skeptical viewpoint until Fermi proves otherwise.

 

[Keysplayr]

No. From what I can see he's basing the whole thing on a simple math that you and some others apparently have trouble to do. A 6-pin plus a 8-pin. Unless there is something else to be explained, that amounts to 225~300W max power.


I have no idea why a common sense push some folks' nerves so hard. Again, HD 5870 = 6-pin + 6-pin. A Fermi shown by NV = 6-pin + 8-pin. Unless somehow the world has gone around and 6+6 > 6+8, there is nothing strange to assume the latter being more power-hungry. I am not sure what you mean by "NV could possibly do something better than ATI". Is it something like "Fermi draws more power (6+8) from the PSU yet consumes less power" (?)

Well, obviously that doesn't make sense. And that's what you're saying. And as for power handling, NV's powermizer has never been as capable as AMD's powerplay since RV670. So it isn't surprising that one, especially a consumer, maintains a skeptical viewpoint until Fermi proves otherwise.

:::facepalm:::
He is basing his calculations on what ATI's 5800 series is.

"I am not sure what you mean by "NV could possibly do something better than ATI"

Zstream asked if toyota believed that NV's power handling was better than ATI's 5800.
Our wires got crossed somewhere Lopri. Just take a step back and lets straighten it out before it turns into something stupid.

 

[Janooo]

There are 2x8 and 1x6 cables coming into the card. Is it usual?

 

[Keysplayr]

There are 2x8 and 1x6 cables coming into the card. Is it usual?

No, there is 1x8 and 1x6 going into the card, and there is probably a 6+2 pin dangling in the air down below. My Silverstone PSU has 2 PCI-e power connectors per line.

It's not "usual" because most PSU's have only 1 PCI-e connector per line.

 

[Janooo]

No, there is 1x8 and 1x6 going into the card, and there is probably a 6+2 pin dangling in the air down below. My Silverstone PSU has 2 PCI-e power connectors per line.

It's not "usual" because most PSU's have only 1 PCI-e connector per line.

That's interesting. All the Fermi pictures show dual 8 pin connector.

 

[Keysplayr]

That's interesting. All the Fermi pictures show dual 8 pin connector.

Got any linkys? Because the guy who actually took the pictures in the OP's first post link, states there is 1x 6 pin and 1x 8pin.

Quote from under the pics:

"From the image above we can see that it requires a 6-pin connector as well as an 8-pin, so I'm guessing this is the high-end flagship model. The card is not too big either, about the same size as the motherboard. The demo system was running the Unigine Heaven benchmark to show off the DX 11 tessellation effects. According to the reports, its expected to hit the markets sometime in March, which means we should be able to get our hands on a review sample before. However, there are no details on pricing or performance of the card yet."

 

[Janooo]

Got any linkys?

... and from twitter

 

[Keysplayr]

Here is a close up:

1 six pin, and 1 8 pin. Even by just looking at the solder points on the PCB. In your picture as well. Look at the solder points.

 

[Janooo]

Here is a close up:

...

1 six pin, and 1 8 pin.

Well, just that there are 2 power cables coming into the 8 pin connector. That's the same card as my first picture only a different angle.

 

[toyota]

that is one 6 pin and one 8 pin.

 

[Janooo]

that is one 6 pin and one 8 pin.

... but it's clear that 2 power cables are hooked to the 8 pin.

 

[Keysplayr]

... but it's clear that 2 power cables are hooked to the 8 pin.

I need to take a picture of my PCI-e power cable to show you what I'm talking about, don't I.....

 

[toyota]

... but it's clear that 2 power cables are hooked to the 8 pin.

the pic just looks odd but that is a 6+2 cable going to the 8 pin. to be clear yes it is 2 cables because if you didnt need the extra 2 pin then that cable would just hang down.

 

[Janooo]

I need to take a picture of my PCI-e power cable to show you what I'm talking about, don't I.....

I know what you mean. Still I find it strange that there are 3 sets of pictures with Fermi and all of them show the same thing.

 

[Keysplayr]

I know what you mean. Still I find it strange that there are 3 sets of pictures with Fermi and all of them show the same thing.

What is it that you find "strange" exactly?

This cable I'm holding has 1x 6+2, and 1x 6 pin PCI-e power connectors. Single cable.

You'll notice that because they are on the same cable, the "8pin" connector seems to have twice as many wires going "into" it.

But, the wires are actually going in, and out, to the next connector on the line.

 

[Janooo]

What is it that you find "strange" exactly?

That in all 3 cases they used PS with such cables.

Your cable has no rubber around the 6+2 that goes out.
The cables in the other pictures clearly show insulation on all of them that would indicate that the power goes in.

 

[tviceman]

For the love of god I cannot wait until the benchmarks and exact specs (core clock, TDP, etc.) start leaking out - presumably in the next couple of weeks - so we can move this discussion in a more useful direction.

 

[toyota]

That in all 3 cases they used PS with such cables.

Your cable has no rubber around the 6+2 that goes out.
The cables in the other pictures clearly show insulation on all of them that would indicate that the power goes in.

so do you think this is some trick where 2 8pin cables magically connect to a single 8 pin connector? of course not.