The Larrabee instruction set has been revealed and Michael Abrash gives a first look:
[...] MMX, SSE and Altivec. They supported vector arithmetic, but could only read and write data from contiguous locations in memory, rather than random-access as Larrabee. So SSE was only useful for operations on data that was naturally vector-like: RGBA colors, XYZW coordinates in 3D graphics, and so on. The Larrabee instructions are suitable for vectorizing any code meeting the conditions above, even when the code was not written to operate on vector-like quantities. It can benefit every type of application!
Trying to get video on mobile gadgets can be frustrating. First you need to find an encoder with the right mix of container and codec support, and then you have to suffer the looong encoding times and pray the final file is actually compatible with your device.
Today NVIDIA unveiled a new driver with support for GPU physics and CUDA. Bundled with the driver is an impressive array of demos but a little gem caught my eye: a CUDA accelerated H264/AVC enconder!
It's named Badaboom and, at least on my PC, it's anywhere from 3 to 5 times faster than any other transcoder, on the same videos. And I have one of the slowest supported GPUs (a GeForce 8800GTS). Not only that, but it actually had zero problems with most of the videos I tried to encode. It only choked on an admitely obscure test file (MOV with MPEG2 video), but it failed with a clear error message, instead of hanging or producing an unplayable file. Definitely a sure purchase if you often transcode videos for your gadgets.