Well, technically they could even use the same old NVIO core, right ?
Although i doubt they will.
The "dual-chip" part is intriguing. It better not be another 9800 GX2...
NVIDIA GT200 Rumours & Speculation Thread
- Thread starter Arun
- Start date
- Status
Well, technically they could even use the same old NVIO core, right ?
Although i doubt they will.
The "dual-chip" part is intriguing. It better not be another 9800 GX2...
Agreed.
I propose we stop gunking up the thread with fake screenshots and quotes of said screenshots.
If you have to ask "Is this real?", then it's most likely a waste of thread space.
If you have to answer "It's the third fake screenshot in as many pages", perhaps you can not quote the image as well.
Maybe it's just me, but the sudden sprouting of "me too" unverified or fake screenshots is getting silly and rage-inducing.
If you have to ask "Is this real?", then it's most likely a waste of thread space.
If you have to answer "It's the third fake screenshot in as many pages", perhaps you can not quote the image as well.
Maybe it's just me, but the sudden sprouting of "me too" unverified or fake screenshots is getting silly and rage-inducing.
Text doesn't require me to scroll down half the thread page to find out someone got creative with MS Paint.
What's wrong with linking to the image, and not including the image when quoting?
Let's make this easy - no GPU-Z screenshots for either GT200 or R7xx unless accompanied with other plausible-sounding information, and no discussion that assumes the transistor count or the die size to be a perfectly reliable and obvious variable. A ridiculous amount of the noise in both threads is related to these factors so let's give that a try... please? 
How do you know that?
What are the current roumors?
GT200 has 10 clusters with 24 SPs and eight TMUs. 192 SPs = eight clusters ("GTS"). The problem: 48 TMUs, what would mean GT200 is a six clusters GPU with 32 SPs per cluster.
Jawed
Legend
Just for a laugh
http://www.cupidity.f9.co.uk/GTC.xls
It's just a spreadsheet to play with transistor counts, making a rather simplistic assumption.
Jawed
http://www.cupidity.f9.co.uk/GTC.xls
It's just a spreadsheet to play with transistor counts, making a rather simplistic assumption.
Jawed
aaronspink
Veteran
Or said better, if its CPU-Z or GPU-Z, it is FAKE. If it isn't FAKE, everyone will already know about it. No need to post. Its not like the maker of GPU-Z or CPU-Z knows anything that any joe blow out there doesn't already know.
aaronspink
Veteran
There is always a difference between rumors and unmitigated BS.
for example, the GT200 should be awesome cause it has 101010101010101 uber mega cores with a clock speed of 4.98376436 GHz but is having problems cause its melting as soon as power is allied is unmitigated BS just slightly more out there than people posting random images from a problem that has no idea what a GT200 or D10U is or whats its die size is etc.
Does anyone really think that the companies out there just let the people who make CPU-Z and GPU-Z know their whole roadmap and everything else they want to know about unreleased products before even the reviewers?
Aaron spink
speaking for myself inc.
If GPU-Z only IDs GPUs from an internal database, as caboosemoose said (and as this thread seems to indicate--this one, too--and, er, the GPU-Z page mentions a database), then it doesn't seem like a stretch to think it wouldn't have pre-release specs on it. (I don't know if it has in the past.)
Edit: OK, it seems W1zzard has "GT200 detection" in a test build. I dunno if that means he necessarily knows the right specs or even all of them, given my first link.
Not the entire roadmap but a few months before release they usually get in communication...
Example...
I also remember when an AMD employee published results of the RV670 before the card was released.
And yes, W1zzard DOES know more than the average joe blow.
Sure, but if the author of the screenshot just wrote "I have learned from a reliable source..." and wrote down the specs, would that be a rumour or an "unmitigated BS"? Then, if the specs look believable...
Well, but such things *should not* happen.LordEC911 said:
There's another screenshot now circulating which sounds as much BS as the first one. How many hits does it take until someone gets it right or thinks that he's gotten it right?
aaronspink
Veteran
There is a big difference between knowing the PCI-E device ID (which is what detection IS, basically all you need to know is the VID which is the same across all products for a company and the DID which is product specific, aka G80 GTX, ultra, and gts all have different DIDs) and knowing the actual details of the design like number of pipes, die size, etc.
And since a lot of the data GPU-Z uses comes from an internal DB in GPU-Z I highly doubt that it means anything more than him mailing it to someone at nvidia or an OEM and asking if it correctly IDs the product.
Now he probably does have code that reads some of the PCI register space for the device to get some of the parameters like clocks, etc, and assuming nvidia has those registers standardized across their product lines it should work for any new device out there. But that depends on nvidia keeping them standard or giving wizard the config space addresses and the decoder ring for the information. I personally have never been enamored enough with reading nvidia's config space to have taken a look at it (though its easy enough to dump with a wide variety of tools out there, riva tuner will do it for example).
Aaron Spink
speaking for myself inc.
Last edited by a moderator:
- Status
