[Kaotik]

Originally Posted by UniversalTruth

lol

And where do all these discrete (add-in) videocards for laptops belong to?

They're not considered add-in boards. Don't know which category other than overall marketshare includes them, but Jon Peddies add-in board market share doesn't.

Jon Peddie Research (JPR), the industry's research and consulting firm for graphics and multimedia, announced estimated graphics Add-in Board (AIB) shipments and sales’ market share for Q2’12. The JPR AIB Report tracks computer graphics boards, which carry discrete graphics chips. They are used in desktop PCs, workstations, servers, and other devices such as scientific instruments. They may be sold as after-market products directly to customers or they may be factory installed. In all cases, they represent the higher-end of the graphics industry as discrete chips rather than integrated processors.

Note that laptops aren't mentioned, and it's a category that definitely doesn't fit to "other devices"

[DavidGraham]

Originally Posted by Kaotik

Actually tad under 60% http://jonpeddie.com/press-releases/...-last-quarter/

Thanks for the correction , though I expect NVIDIA's share to increase , now that they have their mainstream GPUs out of the door .

[Dave Baumann]

Originally Posted by Kaotik

Laptops have nothing to do with the discrete GPU marketshare (aka add-in boards)

"Discrete GPU" would be cover Notebook discrete, desktop OEM discrete add-in (via ODM's) and AIB add in - thats whats reported when people list "the discrete GPU market" (particularly the top-line Mercury numbers that are often reported). Peddie's numbers there appear to be looking solely at AiB.

[Kaotik]

Originally Posted by Dave Baumann

"Discrete GPU" would be cover Notebook discrete, desktop OEM discrete add-in (via ODM's) and AIB add in - thats whats reported when people list "the discrete GPU market" (particularly the top-line Mercury numbers that are often reported). Peddie's numbers there appear to be looking solely at AiB.

Yes, you're right that it could mean that too, but in this case (using Peddie numbers) that wasn't the case.

Mercury reports 57% discrete for nVidia in Q2, so slightly lower than JPR, but over 2% units higher than Q1

[Dave Baumann]

I'm not convinced that Peddie is tracking ODM boards there.

[tviceman]

The lack of any new info or rumors is boring. With AMD's new drivers and game bundling, perhaps (more hopeful than anything) Nvidia will surprise launch GK110 as Geforce card early in very limited quantities to gain some bragging rights back.

[UniversalTruth]

What kind of new info would you prefer? Like announcements dates or?

View nVidia Kepler 2013 refresh generation

DRAM Contract Price Falls Below US$16 Due to Weak Demand

2012 was a good year for consumers buying DDR3 memory, I expect them to clean up next year. 2013 will be the year of VERY CHEAP pc parts if you have the patience to wait. Video cards, followed by motherboards are heading for the largest tumble. Best, Liquid Cool

[tviceman]

Originally Posted by UniversalTruth

What kind of new info would you prefer? Like announcements dates or?

Leaked specs of the actual shipping Geforce-based GK110 card, leaked performance slides, leaked release dates...

[Ailuros]

Originally Posted by tviceman

Leaked specs of the actual shipping Geforce-based GK110 card, leaked performance slides, leaked release dates...

Typically you'll see these kind of leaks when any product is fairly close to its release.

[Kaotik]

Originally Posted by Ailuros

Typically you'll see these kind of leaks when any product is fairly close to its release.

Well, they did claim they're on schedule for decemeber release on the GK110 Teslas, which is just over month to 2 months away now, so fairly close

[Ailuros]

Originally Posted by Kaotik

Well, they did claim they're on schedule for decemeber release on the GK110 Teslas, which is just over month to 2 months away now, so fairly close

tviceman said GeForce and not Tesla. As it seems there won't be any desktop variants with GK110 within this year. IF they'll release someting relevant in Q1 2013, then the first leaks should start around December or even January.

[UniversalTruth]

Yup, that's correct. As far as I remember, there was information about possible launch in March 2013.

[A1xLLcqAgt0qc2RyMz0y]

In the conversion from Jaguar, a Cray XT5, ORNL essentially gutted the existing 200 cabinets and retrofitted them with nearly ten thousand XK7 blades. Each blade houses two nodes and each one of them holds a 16-core Opteron 6274 CPU and a Tesla K20 GPU module. The x86 Opteron chips run at a respectable 2.2 GHz, while the K20 hums along at a more leisurely 732 MHz. But because to the highly parallel nature of the GPU architecture, the K20 delivers around 10 times the FLOPS as its CPU companion (Using the 27 peak PF value for Titan, a back-of-the-envelope calculation puts the new K20 at about 1.2-1.3 double precision teraflops.

http://www.hpcwire.com/hpcwire/2012-...l?featured=top

[Gipsel]

Or to quote myself from the other Kepler thread:

Originally Posted by Gipsel

According to HPCWire, the Teslas run at 732 MHz. Taking the claimed 27 PFlop/s theoretical peak and factoring in the CPU flops (2,63PFLOP/s for base clock, would be 2,99 PFlop/s for all core turbo clock), it is a 14 SMx GK110 version.

14 SMx
732 MHz
6 GB GDDR5 @ 384Bit memory interface (clock still unknown)
1.31 TFlop/s theoretical peak

It's probably the top version of the K20 line if one compares it with the one that was on sale in a shop (which had only 13 SMx and a 320Bit memory interface and only 1.17 TFlop/s peak).

What I find interesting, is that the XK6 nodes obviously gets upgraded to XK7 ones in the process. Originally it was planned that the Tesla cards are just drop in extensions for the new XK6 nodes (it should have worked as Jaguar just got upgraded from XT5 to XK6). What is strange, is that HPCWire claims a max power consumption of 12.7 MW (up from prior 10.8 MW afaik), while the specs from Cray say only 54.1 kW per rack, same as with XK6. No idea what to make out of that.

[tviceman]

14 SMX's with the all the memory controllers sounds like yields are decent and bodes well for the GK110 Geforce card!

[iMacmatician]

Anyone have estimates for TDP?

[tviceman]

Originally Posted by iMacmatician

Anyone have estimates for TDP?

I can't find the link at the moment, but I remember 225 watts being cited.

[iMacmatician]

For the (presumably) 14 SMX variant?

[Blazkowicz]

I would presume 13 SMX get it easily in 225W, while 14 SMX can be in higher power custom designs. The intent may be for you to get rack units made by some companies, while the card is more affordable. Two or more card variants are possible and probable, too.

[Ailuros]

It comes more down to frequencies then enabled/disabled cluster amounts on a chip.

[iMacmatician]

From Bright Side of News: "Titan Comes to Life: 46 Million Nvidia CUDA Cores, 300,000 AMD x86 Cores."

Their numbers work out to 2496 CCs and 5 GB GDDR5 per GPU.

[Gipsel]

Originally Posted by iMacmatician

From Bright Side of News: "Titan Comes to Life: 46 Million Nvidia CUDA Cores, 300,000 AMD x86 Cores."

Their numbers work out to 2496 CCs and 5 GB GDDR5 per GPU.

Just multiplying numbers with the specs of the lesser K20 version coming from that shop doesn't count as source of information in my book. For starters, Titan has definitely the 6 GB version with the full memory interface and HPCWire explicitly claimed a clock frequency of 732 MHz (instead of 705 MHz).

[tviceman]

Originally Posted by iMacmatician

From Bright Side of News: "Titan Comes to Life: 46 Million Nvidia CUDA Cores, 300,000 AMD x86 Cores."

Their numbers work out to 2496 CCs and 5 GB GDDR5 per GPU.

Originally Posted by Gipsel

Just multiplying numbers with the specs of the lesser K20 version coming from that shop doesn't count as source of information in my book. For starters, Titan has definitely the 6 GB version with the full memory interface and HPCWire explicitly claimed a clock frequency of 732 MHz (instead of 705 MHz).

I brought this up on anandtech and the mods there confidently said hpcwire is much more accurate (they take the time to verify their info) than websites like brightsideofnews and semiaccurate. I'll believe the mods and hpcwire until proven otherwise.

[A1xLLcqAgt0qc2RyMz0y]

Originally Posted by Gipsel

Or to quote myself from the other Kepler thread:

14 SMx
732 MHz
6 GB GDDR5
1.31 TFlop/s theoretical peak

Ryan Smith of AnandTech agrees with you.

http://www.anandtech.com/show/6421/i...cores#comments

We're basing our numbers off of the figures published by HPCWire.

http://www.hpcwire.com/hpcwire/2012-...l?featured=top

For a given clockspeed of 732MHz and DP performance of 1.3TFLOPs, it has to be 14 SMXes. The math doesn't work for anything else.

[iMacmatician]

From SemiAccurate: "Nvidia Tesla K20 specs gives hints about 28nm yields."

Posted for discussion purposes only, since he completely ignores the 14 SMX GK110 at ORNL.