Hi,
I am intrigued by the nMP, but was wondering whether it will be a good fit for developers and running virtual machines?
A 6 to 12 core Xeon should be fine for heavily multi-threaded apps such as Java application servers, but I'm not sure whether the twin GPUs will really add much for this usage, and it sounds like they could be an expensive overhead that doesn't add any real benefit.
I'm aware that there is a tendency towards using GPUs for non-graphical tasks such as running Java apps, but I get the impression that this is not yet available in the real-world. Does Mac OS or XCode make any use of OpenCL running on the GPUs?
Would the nMP be a good Virtual Machine server? Xeons support the usual VT-x, VT-d extensions, and fast SSD and Thunderbolt storage arrays are great for VMs. But VMs also like lots of cores, so sacrificing CPU cores for (unused) GPU processing ability might not a good deal.
What do you think?
John.
I am intrigued by the nMP, but was wondering whether it will be a good fit for developers and running virtual machines?
A 6 to 12 core Xeon should be fine for heavily multi-threaded apps such as Java application servers, but I'm not sure whether the twin GPUs will really add much for this usage, and it sounds like they could be an expensive overhead that doesn't add any real benefit.
I'm aware that there is a tendency towards using GPUs for non-graphical tasks such as running Java apps, but I get the impression that this is not yet available in the real-world. Does Mac OS or XCode make any use of OpenCL running on the GPUs?
Would the nMP be a good Virtual Machine server? Xeons support the usual VT-x, VT-d extensions, and fast SSD and Thunderbolt storage arrays are great for VMs. But VMs also like lots of cores, so sacrificing CPU cores for (unused) GPU processing ability might not a good deal.
What do you think?
John.