Nvidia Geforce 'Maxwell' Thread

Rofflay · 9 May 2014 at 11:09

That will be good news because Nvidia will want to get in on 4K movies too and it makes sense to release a new card in 2014 with both HDMI 2.0 and H.265 is pretty important for some of thier customers too as i was looking at 4K and Bluray or where will you get 4K content? It is going to be a while we need brand new players for H.265 and 4K Bluray discs with multiple layers.

But Red-Ray player is making a grab also basically being a player with a 1TB HDD and H.265 and having no disc drives only an online store that streams in advanced codecs like H.265 and plays from a HDD. You basically buy the films from the store for a fixed price. But then it twigged when i found out it costs around £1100. So basically if you have Maxwell 880 and pc with H.265 you can have your own Redray player using unlimited torrent content and a 1TB drive for about £500 inside your current computer. Until Q1 2015 i do not see there being any players yet alone the discs and the hordes of 4K and 8K transfers of popular films from 1960 and 1980 that people will want to revisit.

Silent_Scone · 9 May 2014 at 13:23

drunkenmaster said:
memory amount doesn't matter, got to go bigger with each new gen, marketing buzzwords. Don't forget that Nvidia and AMD(though Nvidia started it) rebrand the SAME cards frequently just so Dell and the rest can sell "new" computers with better graphics in them.

In terms of compute the more memory the better. In terms of stacked mem, I'm trying to find out what the minimum amount of memory you can get in a stack actually is because currently the HMB is a stack of memory on top of a organic substrate with a 128bit bandwidth(the memory can be MUCH more than that internally but the external connection is currently limited as they haven't got to interposers yet, interposer = silicon = 10 times thinner traces for massively more bandwidth, organic = much thicker, far lower bandwidth possible). Meaning if you can get a minimum of 2gb of mem a stack then to get 512bit bandwidth for a gpu you would need 4 stacks and 8gb minimum and that is a 28nm gpu that needs 512bit bus. If we get roughly speaking two times or maybe closer to 2.5x the cores on 14nm gpu's, 512bit might well not be enough meaning requiring even more memory.

Nvidia doesn't have stacked mem for 28nm and basically no chance till Pascal. But by Pascal they'll want 8-16gb being the range gpu's are using. AMD will be in the same boat. So it's likely we'll see 8gb being pushed at 20nm and 16gb on 14nm potentially.

Memory prices are coming down and density is increasing, on top of that volume for gddr5 will have increased with PS4 production, 8gb gddr5 per console means production volume is up which means price per unit will be down. One of the biggest reasons for not having 8-16gb of memory recently is really the number of chips and number of traces on a pcbm stacked mem(interposer or not) will effectively remove that limitation completely while also reducing power significantly. It's also worth remembering that all bandwidth is not the same, memory controller, the way the memory's own logic accesses it, everything can be improved. HMB over Gddr5 in current speeds is supposed to both be 35% faster and use 40% less power.

Anyway, in marketing terms it's no different to some low end POS card getting the same memory amount as a high end card. For compute, being able to store a lot more information locally means a lot less deleting and fetching across pci-e and ability to use larger data sets which for compute is becoming critical.

Like system memory where people were saying better off with 4 or 8gb of okay memory than 2gb of uber fast memory. Not having enough was significantly worse in every way than losing 5-10% in speed.

I think it's feasible at least. 16GB by Q4 next year at the high end.

RE the memory bus query from before, I think there is some confusion with the speculation regarding what card is what. Mid-range cards can have 256 buses till their blue in the face. in fact they more than likely will

Eurofighter · 23 May 2014 at 09:05

http://www.anandtech.com/show/7900/nvidia-updates-gpu-roadmap-unveils-pascal-architecture-for-2016

Now Maxwell's main feature is DirectX 12 and not unified virtual memory it has to be a die shrink to 20nm. Also 28nm has been with us for two generations now.

Orangey · 23 May 2014 at 16:38

>March 26

Eurofighter · 25 May 2014 at 16:20

I think early next year is a good estimate. I wish we had more info on this Arm core that it's supposed to have.

Jono8 · 25 May 2014 at 17:33

Eurofighter said:
I think early next year is a good estimate. I wish we had more info on this Arm core that it's supposed to have.

Nvidia will certainly release some sort of "next gen" or faster cards this year. To think they won't release anything of that sort in 2014 is silly.

CAT-THE-FIFTH · 25 May 2014 at 17:43

GT740 pictured:

http://www.eteknix.com/first-images...40-revealed-confirmed-to-be-entry-level-gpus/

Eurofighter · 25 May 2014 at 18:02

Jono8 said:
Nvidia will certainly release some sort of "next gen" or faster cards this year. To think they won't release anything of that sort in 2014 is silly.

I can't be sure but I don't think it will be the high end desktop this year.

AshenEdge · 25 May 2014 at 19:30

CAT-THE-FIFTH said:
GT740 pictured:

http://www.eteknix.com/first-images...40-revealed-confirmed-to-be-entry-level-gpus/

ooooh Nvidia please give us something to get excited about

Meanwhile at AMD HQ:

It's not a bash without Flash

PhysicsMan94 · 25 May 2014 at 20:12

CAT-THE-FIFTH said:
GT740 pictured:

http://www.eteknix.com/first-images...40-revealed-confirmed-to-be-entry-level-gpus/

Phwoar look at those sexy 740s! Maxed out solitaire, ooph, think I need to lie down, getting too excited...

Boomstick777 · 1 Jun 2014 at 14:21

NVIDIA Maxwell GPUs Could Possibly Feature New Axial-Radial Hybrid Fan Solution

Read more: http://wccftech.com/nvidia-maxwell-...xialradial-hybrid-fan-solution/#ixzz33OS9iYus

NVIDIA recently announced two new graphics card, their high-end GeForce GTX Titan Z and their entry level GeForce GT 740. These graphics card would be the last entry to Kepler series since NVIDIA plans to unveil the next generation Maxwell GPUs later this year.

NVIDIA GM200 (Maxwell Architecture, High-Performance for Telsa/Quadro Arrives later for Cosnumers, Successor of GK110)
NVIDIA GM204 (Maxwell Architecture, High-End Consumer, Successor of GK104, First GeForce 800 Series Products likely to feature)
NVIDIA GM206 (Maxwell Architecture, Performance Minded, Successor of GK206, Mid-Range GeForce 800 Series products to feature)
NVIDIA GM107/207 (Maxwell Architecture, Entry Level, Successor of GK107, Entry Level GeForce 800/700 Series To feature, Already introduced on GTX 750 Ti / GTX 750)

Read more: http://wccftech.com/nvidia-maxwell-...xialradial-hybrid-fan-solution/#ixzz33OSaMHjp

CAT-THE-FIFTH · 1 Jun 2014 at 15:02

PhysicsMan94 said:
Phwoar look at those sexy 740s! Maxed out solitaire, ooph, think I need to lie down, getting too excited...

It seems they are rebadges and not Maxwell based it seems.

CAT-THE-FIFTH · 1 Jun 2014 at 15:03

Boomstick777 said:
NVIDIA Maxwell GPUs Could Possibly Feature New Axial-Radial Hybrid Fan Solution

Read more: http://wccftech.com/nvidia-maxwell-...xialradial-hybrid-fan-solution/#ixzz33OS9iYus

NVIDIA GM200 (Maxwell Architecture, High-Performance for Telsa/Quadro Arrives later for Cosnumers, Successor of GK110)
NVIDIA GM204 (Maxwell Architecture, High-End Consumer, Successor of GK104, First GeForce 800 Series Products likely to feature)
NVIDIA GM206 (Maxwell Architecture, Performance Minded, Successor of GK206, Mid-Range GeForce 800 Series products to feature)
NVIDIA GM107/207 (Maxwell Architecture, Entry Level, Successor of GK107, Entry Level GeForce 800/700 Series To feature, Already introduced on GTX 750 Ti / GTX 750)

Read more: http://wccftech.com/nvidia-maxwell-...xialradial-hybrid-fan-solution/#ixzz33OSaMHjp

It makes me wonder whether the new cards will be 28NM too - a better air cooler would make sense at least for the top end cards in that case.

nashathedog · 1 Jun 2014 at 19:13

PhysicsMan94 said:
Phwoar look at those sexy 740s! Maxed out solitaire, ooph, think I need to lie down, getting too excited...

I was just looking at those myself, So I can now buy a GT 740 with 4 gb's of GDDR5 but I still can't get a top end 780 ti with more than 3 gb's.

There's something wrong with this.

It's like I'm being forced into getting a 6gb 780 if I can't wait, But why do they need to make us wait.

PhysicsMan94 · 1 Jun 2014 at 19:26

To be fair these 740s could be partially aimed at people who want a decent CUDA compatible card but don't want to pay a lot of money given that Maxwell is ushering in big performance boosts for CUDA applications (the 750/Ti, a Maxwell card, has a CUDA "compute capability" of 5.0 compared to high-end cards like the Titan and even the Tesla K20 which has CCs of 3.5).

Boomstick777 · 13 Jun 2014 at 01:34

Nvidia Geforce GTX 880 and 800 Series to be More Powerful But Cheaper than the 700 Series

Read more: http://wccftech.com/nvidia-geforce-gtx-880-800-series-powerful-cheaper-700-series/#ixzz34TVJ4qeX

We have just received some really pleasant news on Nvidia’s Maxwell Architecture, the 20nm side of things to be exact. The new report basically mentions a few things; firstly, that the GTX 880 will be more powerful than the GTX 780, something common sense derives easily. And secondly that the GTX 880 will be cheaper than its 700 Series Counterpart.

Orangey · 13 Jun 2014 at 01:49

So (just following this through, not believing it yet) this means 780 would get a big price drop... hard to believe with it being Nvidia.

Reaper 392 · 13 Jun 2014 at 01:53

That means there are two possibilities. Either the Maxwell cards become cheaper than the previous generation (and the previous generation cards get a price drop to match) or Nvidia puts the price difference in their pocket. It doesn't take a genius to work out which one they will do.

Deceptor · 13 Jun 2014 at 02:06

the fact remains that unless Nvidia brings out intermediary Maxwell cards based on the 28nm node it could loose market share to AMD who is already working on intermediary GPUs

*insert rage here*

Random Guy Esq. · 13 Jun 2014 at 02:52

I'll believe it when I see it.

Are we looking at early 2015 for these?