In the latest Green500 supercomputer list published by Green500.org, 18 of 20 fastest supercomputers are powered by IBM. The most efficient cluster is the BladeCenter QS22 Cluster built with PowerXCell 8i 4.0 Ghz processor and Infiniband in University of Warsaw. Actually this cluster is not a computing monster provided the max performance is 18.57 [...]
This 800 teraflop real-time ray tracing system, jointly developed by Toyota and Unisys, consists of 73-core chips and can fit into a desktop computer chassis. This monster GPU system is aimed at the auto industry for efficient prototype body designs and paint combinations. For your information, the NVIDIA Tesla desktop GPU system C1060 featuring 240 [...]
The trend of distributing complicated and heavy computations to the GPU is fast developing. PGI now presents it new product line of C and Fortran compilers with the support of NVIDIA GPU for general purpose computing. The new compiler, PGI Release 9.0 includes the PGI Accelerator™ has been demonstrated at ICS ’09 in Hamburg last [...]
The CUDA, NVIDIA’s extension to C language for general purpose GPU (GPGPU) computation, is reported to be unlikely seen on its competitor’s product lines. It is not surprising to see that AMD drops the support for the CUDA as AMD has its own proprietary flavor of GPGPU, called Stream, although it is [...]
Categories
Recent Posts
- Ge.tt in a nutshell
- File sharing services accessed from China
- Journal Citation Reports 2009 Released by Thomson Reuters
- Apple iPad and iBooks Store: The Next Standard In e-Reader?
- ASUS DR-950 Photos?
- Apple Tablet: Potentially the Best eBook Reader?
- ASUS Into eBook Business with 2 DR Models
- Kindle for PC now available for download
- Amazon Kindle soon available on windows PC
- Windows 7 is so much better
- iRiver Story emerges
- Hanvon D21 ePaper gets reviewed
- Since when everyone is making digital readers?
- Sony’s new Daily Edition Reader
- Sony Reader get free NYPL digital contents
