Archive for May, 2009

Chessmaster XI vs. Chessmaster 9000

Sunday, May 31st, 2009

I play chess on my iPhone while commuting, and would like to analysis these games at home. Unfortunately I am on Mac, and own the latest Chessmaster 9000 (on Mac OS X), that is just stupid when doing analysis. (The chess engine is The King and is strong enough to play well)

So I decided to buy Chessmaster XI to install it on my Windows Vista partition, to have more usefull game analysis.

Las! The Analysis were identical to Chessmaster 9000 and absolutely useless if you are not a beginner…

So my clear advice is: avoid Chessmaster for Analysis!

Next week, I will install Fritz 11 for the same purpose, hope it will solve my problem :-)

GeForce GTX 260M & GTX 280M

Saturday, May 30th, 2009

Dell just presented the new “Alienware” laptop, equipped with GeForce GTX 260M or optionnally GTX 280M, that may be a beast for 3D and CUDA too :-)

In fact the GeForce GTX 260M & GTX 280M are just re-branded versions of GeForce 9800M GT/GTX, themselves being just re-branded of the good old GeForce 8800!!! (I own one on my desktop and appreciate it)

Instead of 200 to 240 Scalar Processor, you end up with 96 to 112 SP at a lower frequency, wrose you will end-up with a CUDA 1.1 device instead of a CUDA 1.3 and finally you will pay for the re-branding on GTX :-(

CUDA 1.4 adds double-precision floating point and some other instructions that may be usefull mainly for thread communication (to simplify). Anyway GTX M series could not support GTX desktop software!

Please nVidia, stop rebranding old-technology (the great 8800) to new names, faking, and ending with GeForce GTX series (M) disabled to run software written for GeForce GTX series (desktop)!

Fujitsu 128 Gflops CPU

Sunday, May 17th, 2009

Fujitsu just introduced a new CPU, with a peak performance of 128 Gflops/s, that is targeted as super-computer clustered architecture, for simulation, scientific calculations, research …

They call it “the fastest processor ever”, with 128 Gflops/s.

nVidia’s G200 processor is rated at 933 Gflops/s at peak use (GTX 280 cards), more than 7X powerful than Fujitsu latest processor, for a fraction of the price, available for more than 1 year, for anyone (just go to BestBuy or any PC reseller!).

This is where CUDA is: to use the fastest processors with the fastest memory (GDDR5), and unleash the real power hidden in a simple PC or Mac, to turn them into personal super-calculatore, wether it is for research, simulation, military use, or even video encoding at blazing speed (fastest Full HD 1920×1080p H.264 encoding), or even chess!

With an old GeForce 8600 GTS or a GeForce 9500 GT, you have more peak computing power than the most powerful un-available processor from Fujitsu, in a matter of minute, for around $50! Able to compete or beat a $10000 computer!

Target Architecture

Wednesday, May 13th, 2009

CUDA is supported by 2 architectures : the 2 years-old GeForce 8xxx/9xxx/GT/GTS architecture and the new GTX architecture.

There’s finally few differences between the new GTX architecture and the old one, such as doubling the register number, and the coalesced memory access algorithm, the double (64bits floating point number) support, and few instructions.

Even on the first architecture, older GeForce 8800 (CUDA 1.0 architecture) may lack some features of the actual ones (or Geforce 9800/GTS250 as they were renamed :-( ), such as atomic operations. Cruel!

I will focus on what is known as CUDA 1.1 devices, such as formet GeForce 8800, 8600, 8500, any 9xxx or GT/GTS. They share the same general architecture and metrics, mainly differenciated by number of SP (Scalar Processors), frequencies, memory bandwidth (bus size & frequency) and naturally memory size.

I have one GeForce 8600M GTS, with 32 SP (Scalar Processors) on my laptop, and a GeForce 8800 GTS, with 96 SP on my home desktop, that is approximately 4X faster, due to increase in frequency as well as having more cores.

I plan to go for a GeForce GTX 260 (CUDA 1.3 device) if I am lucky on CraigsList or find a good sale, in the future, while staying compatible and optimizing for the CUDA 1.1 devices.

Thanks for the patience!

Wednesday, May 13th, 2009

I switched from my Web developer job to a CTO job, really demanding, on a big filesharing website, working 12 hours, 6 days a week, it was really hard at first but now the main problems seems to be solved :-)

So I am back on track, reading Chess book, chess development forums and web sites, chess programming book, and naturally CUDA documentation and nVidia’s forums.

I even bought a new pocket chess computer for the fun, to play while commuting on Montreal’s STM Metro.

CUDA-Chess work is my priority after my work (and my beloved one naturally!)