Well.. even if they perform "only" like a Fermi CC 2.0 with 1024 or even 768 Shaders: that would still be great, considering they accomplish it with just 3.5 billion transistors instead of 3.2 billion for 512 CC 2.0 shaders. Thats significant progress anyway.S

Agreed! Don't fotget power consumtion too. I want a chip not a stove!
Industry will never make a huge jump. They have to put in value the research investment. It's always more profitable two small steps than a big one.

---

HOW TO - Full installation Ubuntu 11.10

Otherwise people will be disappointed the next you "only" make a medium step.. ;)

MrS

---

Scanning for our furry friends since Jan 2002

Pre-Order Site in Holland - 500 Euros

http://www.guru3d.com/news.html#15424

3DMark 11 benchmark, which if verified is interesting. I am being cautious about games claims until I know about any emdedded PhysX code. The 3DMark 11 bench is however more interesting. If that translates into the Compute side as well as it indicates .... could be interesting. Still, lets await reality, but I hope it is as good as the 3DMark 11 result, competition is sorely needed out there.

http://www.guru3d.com/news/new-gtx-680-benchmarks-surface/

Regards
Zy

I was thinking about what could be the architectural bottleneck, which results the under utilization of the CUDA cores in the CC2.1 product line.
The ratio of the other parts versus the CUDA cores in a shader multiprocessor is increased compared to the CC 2.0 architecture, except the load/store units.
While the CC2.0 has 16 LD/ST units for 32 CUDA cores, the CC2.1 has 16 LD/ST units for 48 CUDA cores.
And what do I see in the latest picture of the GF104 architecture?

There are 32 LD/ST units for 192 CUDA cores. (there were 64 LD/ST units on the previous 'leaked' picture)
If these can utilize only 64 CUDA cores here at GPUGrid, then only 512 of the 1536 shaders could be utilized here.
Now that's what I call a bad feeling.
But I'm not a GPGPU expert, and these pictures could be misleading.
Please, prove me that I'm wrong.

I wouldn't expect things to work straight out of the box this time. I concur with Zoltan on the potential accessibility issue, or worsening of. I'm also concerned about potential loss of cuda core function; what did NVidia strip out of the shaders? Then there is a much speculated reliance of PhysX and potential movement onto the GPU of some functionality. So, looks like app development might keep Gianni away from mischief for some time :)
The memory bandwidth has not increased from the GTX580, leaving space for a GTX700 perhaps, and there is no mention of OpenCL 1.2, or DirectX 11.1 that I can see of. In many respects NVidia and AMD have either swapped positions or equilibration this time (TDP, die size, transistor count). Perhaps NVidia will revert to type in a future incarnation.

---

FAQ's

HOW TO:
- Opt out of Beta Tests
- Ask for Help

The problem with CC 2.1 cards should have been the superscalar arrangement. It was nicely written down by Anandtech here. In short: one SM in CC 2.0 cards works on 2 warps in parallel. Each of these can issue on instruction per cycle for 16 "threads"/pixels/values. With CC 2.1 the design changed: there are still 2 warps with 16 threads each, but both can issue 2 instruction per clock if the next instruction is not dependent on the result of the current one.

Load/Store units could also be an issue, but I think this is much more severe.

MrS

---

Scanning for our furry friends since Jan 2002

680 SLI 3DMark 11 benchmarks (Guru3D via a VrZone benching session)

http://www.guru3d.com/news/geforce-gtx-680-sli-performance-uncovered/

Regards
Zy

The problem with CC 2.1 cards should have been the superscalar arrangement. It was nicely written down by Anandtech here. In short: one SM in CC 2.0 cards works on 2 warps in parallel. Each of these can issue on instruction per cycle for 16 "threads"/pixels/values. With CC 2.1 the design changed: there are still 2 warps with 16 threads each, but both can issue 2 instruction per clock if the next instruction is not dependent on the result of the current one.

Load/Store units could also be an issue, but I think this is much more severe.

MrS

The Anandtech's article you've linked was quite enlightening.
I missed to compare the number of warp schedulers in my previous post.
Since then I've find a much better figure of the two architectures.
Comparison of the CC2.1 and CC2.0 architecture:


Based on that Anandtech article, and the picture of the GTX 680's SMX I've concluded that it will be superscalar as well. There are twice as many dispatch units as warp schedulers, while in the CC2.0 architecture their number is equal.
There are 4 warp schedulers for 12 CUDA cores in the GTX 680's SMX so at the moment I think GPUGrid could utilize only the 2/3 of its shaders (1024 of 1536), just like of the CC2.1 cards (there are 2 warp schedulers for 6 cuda cores), unless nVidia built some miraculous component in the warp schedulers.
In addition, based on the transistor count I think the GTX 680's FP64 capabilities (which is irrelevant at GPUGrid) will be reduced or perhaps omitted.

They cannot afford to separate gaming and computing. The chips will still need to be the same for economy of scale and there is a higher and higher interest in computing within games.

Changes are good, after all there are more shaders, we just have to learn how to use them. As it is the flagship product we are prepared to invest a lot on it.

gdf

I remember the events before the release of the Fermi architecture: nVidia showed different double precision simulations running much faster in real time on Fermi than on GT200b. I haven't seen anything like that this time. Furthermore there is no mention of ECC at all in the rumors of GTX 680.
It looks to me that this time nVidia is going to release their flagship gaming product before the professional one. I don't think they simplified the professional line that much.
What if they release a slightly modified GF110 made on 28nm lithography as their professional product line? (efficiency is much more important in the professional product line than peak chip performance - of course it would be faster than the GF110 based Teslas)

Glad to hear it was the right information for you :)

I think there's more going on. Note that in CC 2.1 they had 3 blocks of 16 shaders, which are arranged in 6 columns with 8 shaders each in the diagram. In the GK104 diagram, however, there are columns of 16 shaders. If these were still blocks of 16 shaders, there would be 12 of the blocks, which in turn would require 12 dispatch units - much more than available.

This wouldn't make sense. What I suppose they did instead is to arrange the shaders in blocks of 32, so that all threads within a warp can be scheduled at once (instead of taking 2 consecutive clocks). In this case there'd be "only" 6 of these blocks to distribute among 4 warps with 8 dispatch units.

Worst case we should stil see 2/3 of the shaders not utilized. However, there are 4 warps instead of 2 now. Still (as in CC 2.1) every 2nd warp needs to provide some instruction suitable for parallel execution, but load balancing should be improved.

And there's still the chance they increased the "out of order window", which is the amount of instructions that the hardware can look ahead to find instructions suitable for superscalar execution. As far as I understand this had only been the next instruction in CC 2.1.

I too suppose it's not going to be a DP monster - and it doesn't have to be as a mainly consumer / graphics oriented card. Leave that for GK100/GK110 (whatever the flag ship will be).

MrS

---

Scanning for our furry friends since Jan 2002

Looks like some guys over at XS managed to catch tom's hardware with their pants down. Apparently, tom's briefly exposed some 680 performance graphs on their site and XS member Olivon was able to scrape them before access was removed. Quote from Olivon:

An old habit from Tom's Hardware. Important is to be quick

LOL!

Anyways, the graphs that stand out:





Other relevant (for our purposes) graphs:

Release date is supposed to be today!
I expect Europe has to wait for the US to wake up, before the Official reviews start. Until then tweaktown's unofficial review might be worth a look, but no CUDA testing, just games.

There is an NVidia Video here.
The card introduces GPU Boost (Dynamic Clock Speed), and 'fur' fans will be pleased!

LegitReviews posted suggested GK110 details, including release date.

---

FAQ's

HOW TO:
- Opt out of Beta Tests
- Ask for Help

2304 is another fancy number, regarding the powers of 2.
Probably the next generation will contain 7919 CUDA cores. :)

Interesting and tantalizing numbers. Can't wait to see how they perform.

it appears that they are actually available for real at least in the UK.
So it is not a paper launch.
gdf

nvidia-geforce-gtx-680-review by Ryan Smith of AnandTech.

Compute performance certainly isn't great and FP64 is terrible (1/24)!

They can be purchased online from around £400 to £440 in the UK, though the only ones I can see in stock are £439.99! Some are 'on order'. So yeah, real launch, but somewhat limited and expensive stock. Also, they are the same price as an HD 7970. While AMD launched both the HD 7970 and HD 7950, NVidia had but one, as yet... This is different from the GTX480/GTX470 and the GTX580/GTX570 launches.
We will have to wait and see how they perform when GPUGrid get's hold of one, but my expectations are not high.

Other Reviews:

Tom’s Hardware

Guru 3D

TechSpot

HardOCP

Hardware Heaven

Hardware Canucks

TechPowerUp

Legit Reviews

LAN OC

Xbit Labs

TweakTown

Phoronix

Tbreak

Hot Hardware
Link Ref, http://news.techeye.net/hardware/nvidia-gtx-680-retakes-performance-crown-barely#ixzz1psRj9zuD

---

FAQ's

HOW TO:
- Opt out of Beta Tests
- Ask for Help

Here in Hungary I can see in stock only the Asus GTX680-2GD5 for 165100HUF, that's 562.5€, or £468.3 (including 27% VAT in Hungary)
I can see a PNY version for 485.5€ (£404), and a Gigabyte for 498€ (£414.5) but these are not in stock, so these prices might be inaccurate.

So its compute power has actually decreased significantly from the GTX 580?! The Bulldozer fiasco continues. What a disappointing year for computer hardware.

It's build for gaming, and that's what it does best. We'll have to wait a few more months for their new compute monster (GK110).

MrS

---

Scanning for our furry friends since Jan 2002

So far we have no idea of how the performance will be here.
I don't expect anything super at start (gtx580 like performance), but we are willing to spend time optimizing for it.

gdf