NVIDIA Maxwell Speculation Thread

[Svensk Viking]

Sweoverclockers mention in their link though that there's one Denver core in the GPU; since online translators are usually a disaster I wonder if they mean 1 CPU core for all GPU chips irrelevant of size (doesn't make much sense) or if the specific chip (most likely GM107?) will carry 1 Denver core. If it truly is to replace the 650 Ti boost the latter has 768SPs@980MHz. Now 2,3 or 4 clusters?

I'm a swede. Sweclockers says that Maxwell is supposed to have an ARM CPU, but it's uncertain whether all Maxwell-based cards will have it. Nothing specific about the GTX 750.

 

[Ailuros]

I'm a swede. Sweclockers says that Maxwell is supposed to have an ARM CPU, but it's uncertain whether all Maxwell-based cards will have it. Nothing specific about the GTX 750.

I appreciate the help

 

[xDxD]

Not again with that "1" in the middle nonsense. Since we've unfortunately inhereted that fairy tale from the supposed "GK100" that never existed I think we can finally get back to reality, exactly because NV has these days the luxury to re-design entire product families and trash the old ones I guess.



Sweoverclockers mention in their link though that there's one Denver core in the GPU; since online translators are usually a disaster I wonder if they mean 1 CPU core for all GPU chips irrelevant of size (doesn't make much sense) or if the specific chip (most likely GM107?) will carry 1 Denver core. If it truly is to replace the 650 Ti boost the latter has 768SPs@980MHz. Now 2,3 or 4 clusters?

can i ask you: what is your opinion?

 

[Ailuros]

can i ask you: what is your opinion?

How would I know? As of lately I'm officially labelled as an NV hater in this neck of woods

 

[UniversalTruth]

I was thinking of tech journalists. And those who think that you scrap a chip because you can't make the previous one work, but then can make it work by giving it a different name.

Not saying that projects don't get cancelled, it happens all the time, but I can't remember a single case where it happens because someone can't make it work (whatever that means.) Projects get cancelled because market realities change: a competitor comes out with something drastically new, or because the design time is too long during the pre-tape out stage.

I have never said this... What I guess is that it is possible they mad a new revision (thus the new numbering) with improved characteristics which deserves more to be released in the wild... Next time I will avoid the 'scrap' word, perhaps it is what you dislike

 

[xDxD]

How would I know? As of lately I'm officially labelled as an NV hater in this neck of woods

Ok thank you anyway

 

[ams]

http://videocardz.com/48726/nvidia-geforce-gtx-750-ti-maxwell-gpu-february

Tom's Hardware hinted something along these lines 10 days ago:

http://www.tomshardware.com/reviews/tegra-k1-kepler-project-denver,3718-3.html

Moving forward, every GPU architecture Nvidia develops will be mobile-first. That was a decision management made during Kepler’s design and then applied to Maxwell from the start. It doesn’t mean Maxwell will show up in Tegra first (in fact, the first Maxwell-powered discrete GPUs are expected in the next few weeks). But the architecture was approached with its mobile configuration and power characteristics in mind, scaling up from there. Sounds like a gamble for a company so reliant on the success of its big GPUs, right? Nvidia says that the principles applied to getting mobile right will be what help it maximize the efficiency of its discrete products moving forward—and we’ll have the hardware to put those claims to the test once GeForce GTX 750 materializes.

 

[iMacmatician]

Not again with that "1" in the middle nonsense. Since we've unfortunately inhereted that fairy tale from the supposed "GK100" that never existed I think we can finally get back to reality, exactly because NV has these days the luxury to re-design entire product families and trash the old ones I guess.

Now where's the GM100?

About GN1x7 parts, I recall there was a GF117 part with, as far as I am aware, no sign of a GF107 predecessor.

I expected that initial Maxwell parts would be named in the 800 series. Maybe they're reserving that for 20 nm parts?

 

[Ailuros]

Now where's the GM100?

About GN1x7 parts, I recall there was a GF117 part with, as far as I am aware, no sign of a GF107 predecessor.

I expected that initial Maxwell parts would be named in the 800 series. Maybe they're reserving that for 20 nm parts?

I don't think it'll sell less if it's called 750-whatever.

 

[itaru]

Denver tips?

Instruction-optimizing processor with branch-count table in hardware

 

[jlippo]

I'm a swede. Sweclockers says that Maxwell is supposed to have an ARM CPU, but it's uncertain whether all Maxwell-based cards will have it. Nothing specific about the GTX 750.

I would bet that all Maxwell based chips have ARM core or several.
This excludes Kepler re-brands.

 

[A1xLLcqAgt0qc2RyMz0y]

Denver tips?

Instruction-optimizing processor with branch-count table in hardware

Looks like Nvidia's put those ex-Transmeta employee's talents to good use.

http://en.wikipedia.org/wiki/Transmeta

NVIDIA Licenses Technologies from Transmeta Corporation

Nvidia taps Transmeta team for x86 chip, claims analyst
http://www.theregister.co.uk/2009/11/04/nvidia_transmeta_x86

It is interesting that the above patent was filed in May 2012 and just published Nov 2013 so it does look like it is part of the Denver project.

Now onto the good stuff, what non-native instructions do you think it is being referred to?

X86 is specifically mentioned but I thought that the $1.5 billion patent licensing between Nvidia and Intel prevented that.

 

[UniversalTruth]

Now where's the GM100?

Perhaps where the GF100 was in relation with GF110

 

[Ailuros]

Perhaps where the GF100 was in relation with GF110

Pleaaaaaaaaaaaaaaaaaaase don't....mercy

 

[iMacmatician]

I expected that initial Maxwell parts would be named in the 800 series. Maybe they're reserving that for 20 nm parts?

Maybe there's another (unlikely?) possibility. NVIDIA is pulling an AMD in 2014 and reserving the 800 numbers for OEM parts, while the retail parts remain as the 700 series. Then when 20 nm Maxwell shows up late this year or whenever, they'll introduce a new naming system, maybe along the lines of "GeForce M1 [number]" (like "Tegra K1").

 

[eastmen]

There is a rumor going around about Maxwell low end hitting next month. Anyone hear anything about that being true ?

 

[Alexko]

It probably is true: a low-end 28nm chip, apparently codenamed GM107.

 

[McHuj]

Disappointing that's its a low end part, but I'm really excited to see the new architecture and see how well it performs compared to Kepler. We should be able to make at least some rough approximations of what a high end part will be.

 

[boxleitnerb]

Apparently both, the 750 and 750 Ti have a 6-pin connector. Performance is supposed to be impressive, especially at higher resolutions despite 128-bit memory interface. The card (750 Ti?) should be in between the 260x and the 270. The Denver core (I thought it wouldn't be on board on the smaller GPUs or not at all until 16nm FinFet??) supposedly helps with some form of compression.

http://forums.anandtech.com/showthread.php?t=2364264

 

[DSC]

http://videocardz.com/49409/nvidia-geforce-gtx-750-ti-15-slower-gtx-660