i thought i'd have to buy a mouse.

Nope. Fortunately the cell phone charging stations have a connector which fits the Genius mouse. It charges in 2-3 minutes and it should be good for all my writing tonight

Genius


There are still sessions from 5-6 PM ... but i am really tired and these computing sessions are making me sleepy .. time for a nap .. then i have Day 1 to write

So is Big Kepler coming out right away for the professional market, or in 6-9 months?

Q4

There is no mention of the big K gaming GPU here at the GTC. So logically, it is after the professional products are out.

What is going to be released shortly is the GK104 dual chip (single precision) for the pro markets - much the same as the GTX 690 was released for gamers.

Was anything said about the GTX 660?

i didn't ask

No one brought it up. However, the information that i got before the GTC was that Nvidia's lineup was being released "normally". And this seems to indicate that it would be sooner rather than later - much like the timeline of GTX 560 Ti being released after the GTX 570/580.

i am working on "GTC - Day 1" now .. with luck i will get it roughed in tonight and published in the early AM.

The first full day of the GTC exceeded my expectations. And i barely got to the exhibitors tonight (in the Networking section). Nvidia provides the food; drinks are ridiculous expensive. Good thing for me that i don't drink alcohol.

However, i think my caffeine habit will be back and firmly entrenched by Friday. It is SO busy and the "classes" (sessions) are scheduled so you almost have to run from one to another (if i may be allowed a complaint; i was late to 3 sessions)

How are u enjoying ur programming sessions man ?
I got a mail that Nsight has been renamed to Nsight Visual Developer edition,got any more info on that?

i almost fell asleep just once. Strong black organic tea worked for the rest of the sessions.
-- it is much better than i thought. Either my understanding of GPU compute has improved over the past 3 years since the 2009 GTC ....
. . . or i am numb


(i like to think it is that i am improving instead of getting better at it)


i got the press release on Nsight

AND I QUOTE>>>

NVIDIA today introduced NVIDIA® Nsight™, Eclipse Edition, the world’s first integrated development environment (IDE) for developing GPU accelerated applications on Linux- and Mac OS-based systems.
NVIDIA Nsight provides powerful debugging and profiling tools that enable high performance computing (HPC) and graphics developers to fully optimize the performance of CPUs and GPUs.
New Nsight, Eclipse Edition for Linux, Mac OS
The new Nsight, Eclipse Edition, enables CUDA® programmers to easily develop, debug and optimize the performance of GPU-accelerated applications within a familiar, highly productive IDE based on the open source Eclipse framework (http://www.eclipse.org). Key features include:
 Automatic code refactoring – Helps convert slow sequential CPU loops into parallel GPU kernels
 Integrated expert analysis system – Provides automated performance analysis and step-by-step guidance to address application performance bottlenecks
 High-productivity development environment – Syntax highlighting and auto-completion for both CPU and GPU code helps developers program more efficiently
 Integrated code samples, online documentation – Makes it easy for developers to quickly get started
“NVIDIA Nsight is the ultimate development platform for heterogeneous computing,” said Ian Buck,
NVIDIA Unveils Nsight Development Platform for Linux, Mac and Windows Page 2
general manager of GPU computing software at NVIDIA. “Whether you’re a graphics or HPC developer, Nsight makes it easy to develop parallel code for GPUs and CPUs using your preferred IDE.”
NVIDIA Nsight Visual Studio Edition for Windows
NVIDIA also announced an updated version of NVIDIA Nsight, Visual Studio Edition for Microsoft Windows developers. Nsight, Visual Studio Edition (formerly known as NVIDIA Parallel Nsight) adds a number of new enhancements and updated features designed to make parallel programming on GPU-based Windows systems faster and easier than ever.
Key among these features is local single GPU debugging, which enables CUDA developers to debug their CUDA C/C++ code natively on the hardware with any system equipped with any CUDA 1.1 or higher capable GPU. Other features include performance improvements to the frame profiler and debugger, and support for DirectX 9 frame debugging, frame profiling and analysis.
“Previously, debugging required dedicated systems that were often expensive and time consuming to configure,” said Tony Tamasi, senior vice president of content and technology at NVIDIA. “Now, any system with an NVIDIA GPU that supports debugging can be used without any additional cost or system upgrades, resulting in significant cost and time savings.”
Availability
Developers can sign up to receive a free preview of NVIDIA Nsight, Eclipse Edition or Nsight, Visual Studio Edition by joining the NVIDIA GPU Computing Registered Developer program at http://www.nvidia.com/paralleldeveloper.
For more information about Nsight development tools, visit the Nsight web page. To learn more about CUDA or download the latest version, visit the CUDA website. More NVIDIA news, company and product

/end quote


Did that help?

i too have heard that we would know a lot more about bigK Q4. it was not feasible to have it anytime soon. i also know they got plenty of time to tweak the design which sits between the 680 and the 690. with the extra time they can squeeze more performance from it and this is good. perhaps they can get it up a notch, take advantage of this delay which was completely out of their control. If they can get more performance and more per watt then it might pay off in the long run.

Its absolute, nvidia had a great planB this round. They learned hard lessons from fermi. Nvidia has a good looking year ahead with a solid plan. they seem to have it together beautifully. The gk104 is their lemon-aid and its paid off big time!!

Thanks a lot,it was a big help

Strong black organic tea is a miracle i know that first hand

Tomorrow is the "deep dive" into the Kepler architecture featuring the GK110. Unfortunately i have to miss Jensen's fireside chat with the emerging companies .. but then someday, ABT will be on the stage presenting our vision


Well, we figured it out before the GTC. CUDA is absolutely taking off. Nvidia is putting a big emphasis on GPU computing and obviously BigK is going to the pro market first. The GK-104 has done a great job and GTX 689/690 should be the top gaming cards until they are replaced in 6-9 months.

Performance per watt is the mantra. It not only make gaming GPUs work, it works wonderfully when there are hundreds of GPUs working together in a super-computer - or in Nvidia's new grid. It looks like they solved the latency issues and cloud computing - including gaming - is on the way. OnLive was not mentioned but they are also exploring using Kepler.

The other thing that the AMD-friendly forum doesn't quite comprehend is that Kepler is much better than Tahiti at compute (what do you expect from AMD employees?).
OpenCL is optimized for 79x0 which is the only reason why the benches look better. And there is no way to run CUDA on 7970 or it would be a dog in comparison (from what i heard)