Tuesday, November 11th 2014

NVIDIA GeForce GTX TITAN II Detailed

Riding on the success of its GM204 silicon, it looks like NVIDIA won't wait for the 20 nm silicon fab process to build its next big GPU, which powers its enthusiast-segment graphics cards. The GM200 silicon will be built on the existing 28 nm silicon fab process. Among other SKUs, NVIDIA's next GeForce GTX TITAN product, the GTX TITAN II, could be based on this chip. A curious-looking data entry was submitted from an anonymous source to SiSoft's hardware database, which gives away some rather glaring details of the GM200.

To begin with, the GM200 will be built on the existing 28 nm node, and will feature a die-area of 551 mm². The chip will be based on the "Maxwell" architecture, and feature 3,072 CUDA cores, cushioned by a 3 MB L3 cache. The chip will likely feature a 384-bit wide GDDR5 memory interface, with lossless texture compression algorithms, which work to step up memory bandwidth. The standard memory amount is a staggering 12 GB, double that of the first-generation GTX TITAN. Given how AMD recently gave 8 GB variants of its Radeon R9 290X a coordinated launch, GPU vendors could be seeing a utility in giving their products such massive amounts of video memory, to cope with resolutions such as 4K Ultra HD, and perhaps even 5K 16xHD. The chip features reasonably high clock speeds, with the core running at 1100 MHz, and a staggering 1390 MHz GPU Boost. The memory, however, is clocked at 6.00 GHz.
Source: Expreview
Add your own comment

32 Comments on NVIDIA GeForce GTX TITAN II Detailed

#1
64K
12 GB VRAM? Well, I wasn't expecting that and running at 6 GHz?
Posted on Reply
#2
RCoon
Not 100% convinced 12GB VRAM is remotely useful for 4K gaming. Even 8GB is excessive. 50% increase in core count is interesting for sure, but bearing in mind what happened with the previous Titan, I don't forsee these selling very many to previous buyers.
Posted on Reply
#3
GhostRyder
Interesting, I am actually a bit surprised they are sticking with 28nm silicon but this could point to an earlier than expected release if they are willing to stick with the previous process. Might also indicate they are planning on holding off on the next lineup of GTX products for quite some time (By that I mean like next August ish).

I do not think they are intending that amount of VRAM for the 4K gaming public, more likely to me the same as before the render crowd and cuda devs.

64K
12 GB VRAM? Well, I wasn't expecting that and running at 6 GHz?
Well remember that is also on a higher bus speed.
Posted on Reply
#4
15th Warlock
GhostRyder
Interesting, I am actually a bit surprised they are sticking with 28nm silicon but this could point to an earlier than expected release if they are willing to stick with the previous process. Might also indicate they are planning on holding off on the next lineup of GTX products for quite some time (By that I mean like next August ish).

I do not think they are intending that amount of VRAM for the 4K gaming public, more likely to me the same as before the render crowd and cuda devs.


Well remember that is also on a higher bus speed.
Yup, when the original Titan came up, 4K gaming wasn't even a thing, and that didn't stop Nvidia from straping 6GBs of VRAM to it.

Remember that for a lot of researchers and renderers, the Titan is really a poor man's Tesla; with its fully functional FP64 performance, Nvidia is not just targeting the gaming crowd.
Posted on Reply
#5
BorisDG
Die size just 551mm² compared to 561mm² (GK110)?

Yeah, because GM204 is bigger than GK104 and I expected same here. :)
Posted on Reply
#6
the54thvoid
Well, with 12GB of Ram they can sell it on Video editing and the like. I think this memory amount, if true will dictate it's price point. I figure it's going to be another £800/$1000 card, unfortunately. And as RCoon says, once bitten twice shy - how long after this do they release a 6GB 980Ti version for 2/3 the price?

Will wait to see what Daddy Maxwell is going to be.
Posted on Reply
#7
64K
Stacking this card up against the original Titan it looks like a significant increase in performance.

Titan
2,688 Cores
GPU Clock 836 MHz
Boost Clock 876 MHz

Titan II
3,072 Cores
GPU Clock 1100 MHz
Boost Clock 1390 MHz

Very impressive. I wonder what the Titan Black II will do if they release one.
For me, I will wait on GM210 as I only use my cards for gaming.
Posted on Reply
#8
Dbiggs9
I will wait for the 20nm
Posted on Reply
#9
GhostRyder
15th Warlock
Yup, when the original Titan came up, 4K gaming wasn't even a thing, and that didn't stop Nvidia from straping 6GBs of VRAM to it.

Remember that for a lot of researchers and renderers, the Titan is really a poor man's Tesla; with its fully functional FP64 performance, Nvidia is not just targeting the gaming crowd.
Well yea but my problem is the card has always been a gaming card first and a work horse second. Its essentially what the GTX 580 once was and my problem with that was we used to get that as just part of the package.

the54thvoid
Well, with 12GB of Ram they can sell it on Video editing and the like. I think this memory amount, if true will dictate it's price point. I figure it's going to be another £800/$1000 card, unfortunately. And as RCoon says, once bitten twice shy - how long after this do they release a 6GB 980Ti version for 2/3 the price?
I still am doubtful that a 980ti will ever come out based on what we have seen so far. I believe that Titan II will come and then eventually the GTX 10XX series (Or the next naming scheme) instead but that is just my guess based on the information I have right now and the way NVidia tends to do things.

64K
Stacking this card up against the original Titan it looks like a significant increase in performance.

Titan
2,688 Cores
GPU Clock 836 MHz
Boost Clock 876 MHz

Titan II
3,072 Cores
GPU Clock 1100 MHz
Boost Clock 1390 MHz

Very impressive. I wonder what the Titan Black II will do if they release one.
For me, I will wait on GM210 as I only use my cards for gaming.
Oh yea its going to be nice but at a price, not sure what's going to happen with Titan II black (or if it will come out) mostly because I think it will come down to the die used and if this is similar to last rounds Titan. I think more information is going to be needed before we can safely say what the future is going to bring with this card.
Posted on Reply
#10
Casecutter
Yeah, Nvidia has to do something because TSCM is going to be behind on 16nm, and there wasn't enough gain for Nvidia if Maxwell on 28nm is good enough, cheap and not much if any risk... why do 20nm.

Interestingly will they release professional cards on this GM200, keep the pressure on AMD and their professional aspirations? If they aren't looking to get it into the professional markets this will be tough to get volumes that give "enthusiast" good pricing, but I can see this really shocking the market. It abets Nvidia in substantiating their premium price, should be a good response to whatever AMD brings, with little risk. If AMD does 20nm and HBM their pricing won't great on chips or memory so, Nvidia knows they can stack this with lots of lower cost memory and it makes the professionals happy and gamer will say... we've got ram!

Nvidia has the safe play, AMD a crap shoot (if 20nm/HBM) and they'll need to triumph with the next go around. Even if thier 390X comes with great perf/w and perf/$, if it does or doesn’t best this GM200 in say GTX 980Ti form, Nvidia can just work AMD over on pricing (if they have it in high volumes for professional makets). Kind of where Nvidia had been when they held with 65mn, and AMD went 55nm and 4870.
Posted on Reply
#11
Bytales
Finally a GPU thats worth stuffing into my computer. I wonder if they will ever make a titan-Z II out of it.
Posted on Reply
#12
MxPhenom 216
ASIC Engineer
64K
Stacking this card up against the original Titan it looks like a significant increase in performance.

Titan
2,688 Cores
GPU Clock 836 MHz
Boost Clock 876 MHz

Titan II
3,072 Cores
GPU Clock 1100 MHz
Boost Clock 1390 MHz

Very impressive. I wonder what the Titan Black II will do if they release one.
For me, I will wait on GM210 as I only use my cards for gaming.
Not only that, but Maxwell cores are able to do a lot more then Kepler's.
Posted on Reply
#13
xorbe
MxPhenom 216
Not only that, but Maxwell cores are able to do a lot more then Kepler's.
I thought someone tested this, and found they were fairly close clock-for-clock.
Posted on Reply
#14
HumanSmoke
64K
12 GB VRAM? Well, I wasn't expecting that and running at 6 GHz?
GK 110 initially emerged as the Tesla range, so it isn't beyond the realms of possibility that GM 200 does also. Also quite possible that the qualification samples have wider options for board power limits / frequency range for validation purposes across a number of possible applications as has been seen previously with test/qualification samples.
The other factor is obviously as marketing bullet point, but the 6 GHz memory (rather than the 7 GHz now standard for high end GeForce) would tend to support the hypothesis that the test board is being validated for GPU operation across a range of speeds. Dialling down the memory removes it as a factor for testing the system for stability etc.
xorbe
I thought someone tested this, and found they were fairly close clock-for-clock.
I sincerely doubt it. Architecture to architecture, the closest analogue to the GM 204 is GK 104, so you'd be comparing the GTX 770 (1536 core, 32 ROP, 128 TMU, 256-bit, 1046MHz core, 1085MHz nominal boost) to the GTX 970 (1664 core, 32 ROP, 104 TMU, 256-bit, 1050MHz core/1178MHz nominal boost). The difference in performance (32% at 1440p) far outstrips the hardware comparison, because the salient differences are amongst other things, core-to-compute unit ratio improvement ( 128:1 instead of 192:1), increased and better organized L2 cache structure, and the use of colour compression.
Posted on Reply
#15
Ferrum Master
Boney, you still alive.

Concerning all this day of the LEAKS...

It is all speculations... Seriously, I won't believe nvidia will release anything on this 28nm tech node again...
Posted on Reply
#16
HumanSmoke
Ferrum Master
It is all speculations... Seriously, I won't believe nvidia will release anything on this 28nm tech node again...
So by your reasoning the test and validation boards seen in transit in July from the first GM 200 tape out in April, are what? Are you saying that Nvidia not only had access to 16nm before anyone else, but they built a large die GPU before anyone else got ARM or FPGA's out the door? Or are you saying that Nvidia built a monolithic GPU using a CLN20SOC process that can possibly be used for high performance or low power but not both? (IIRC, 20nm production of small silicon didn't actually start until after GM 200 was laid down)
[IMG]http://www.guru3d.com/index.php?ct=news&action=file&id=7515[/IMG]
Posted on Reply
#17
TRWOV
OMG now I feel inadequate :banghead:
Posted on Reply
#18
Casecutter
Ferrum Master
I won't believe nvidia will release anything on this 28nm tech node again...
Do you think they can hold-out on what appears to be a year at least, because that's the best I’ve been hearing for 16nm, or they've gone 20nm, or? I'd also believe when they have 16nm it will certainly be going to professional application and HPC first. At best a big 16mn for gaming enthusiasts is like summer 2016. They'll bring the big Maxwell 28nm they’re not going to be twiddling their thumbs for the next year, they’ve a plan to offer the fastest graphic card right after AMD... or before? I figure they'll let AMD go first if the can't best them they can then price it to hurt.
Posted on Reply
#19
qubit
Overclocked quantum bit
It's a shame that physical constraints of power consumption, heat, transistor budget, physical size and yields prevents NVIDIA from creating the full-blown power of two chip with 512-bit bus, 8GB RAM etc. Such a chip would be considerably more powerful than the current top end GPUs. For balance, these constraints affect AMD equally.

It's still gonna be awesome though. My 780 Ti is an amazing card and this is gonna top it by a considerable margin. Hopefully the latest games at 4K and 120Hz refresh will be a reality with just one card.

I can play the latest games at 1080p 120Hz (often with no dropped frames) on my current PC (see specs) but pushing 4K using DSR really tanks performance, although it may still stay well above 60fps depending on the game and quality settings.
Posted on Reply
#20
Ralfies
Assuming this leak is accurate, the specs are kind of disappointing. GK110 had almost 90% more cores than GK104, while GM200 is only a 50% increase over GM204. It appears it will be clocked close to GM204 though.
Posted on Reply
#21
HumanSmoke
Ralfies
Assuming this leak is accurate, the specs are kind of disappointing. GK110 had almost 90% more cores than GK104, while GM200 is only a 50% increase over GM204. It appears it will be clocked close to GM204 though.
Has the 3072 cores actually been confirmed as the full complement available to the GPU?
Posted on Reply
#22
bhaalkc
in the future, you are gonna have more vram than ram.....
Posted on Reply
#23
RejZoR
Main problem with high res is not the amount of VRAM, but shaders ability to process all the stuff. And currently no graphic card is fast enough to be adequate.
Posted on Reply
#24
Prima.Vera
What video applications out there are confirmed to use excessive amount of VRAM?
Posted on Reply
#25
HisDivineOrder
I think I'm looking forward to when they make the 6 GB version of this and call it the 990 Ti or whatever they devise as their new numbering scheme.

Since last time, they went from four digit numbers to three digit numbers, I guess this will be the Geforce 18? ;)

Eventually bringing us back to the Geforce 1 after a few more generations.
Posted on Reply
Add your own comment