To be fair, he said HIS research. No one one knows what they will release for server hardware in this class yet but if a desktop sandy bridge helps him now, lovely. If a server sandy bridge comes along in a few months, even better
Not everything I do is easily parallelizable (so a server farm isn't super useful). I use matlab to design and test my algorithms (some of which can use GPGPU effectively, some of which cannot, some of which the use of GPGPU would add significant coding/debug time just to see that an idea doesn't work). Often times, I can end up crashing the machine I'm working on, so I try to keep my code off of a cluster, as well.
Sandy Bridge looks nice because it runs cooler and faster than the i7 and phenom 2 processors. I couldn't care less about the integrated gpu...but for my research workstation, it would probably be sufficient.
When/if the research I do is deployed to a real system, then it would be worth spending the time to make it uber parallel and/or gpgpu code-compatible. For now, I just want a workstation with a fast(er) processor.
It seems that you were being slightly patronizing by telling me that these processors were not what I wanted, assuming that I didn't know what else was out there.
Comments
So I should add my PhD, publications, and job credentials along with my resume here as well? What a crappy post.
No, but you should probably stop being so dismissive and patronizing. That's a good place to start.
Sandy Bridge looks nice because it runs cooler and faster than the i7 and phenom 2 processors. I couldn't care less about the integrated gpu...but for my research workstation, it would probably be sufficient.
When/if the research I do is deployed to a real system, then it would be worth spending the time to make it uber parallel and/or gpgpu code-compatible. For now, I just want a workstation with a fast(er) processor.
ITT: metapatronizing