HN Explorer

Ask HN: What is your entry:"What would you do with 48 cores ?"

3 points by nuitblanche ↗ HN
I am curious. Anybody else care to share their entries foe the 48 cores giveaway AMD contest( http://blogs.amd.com/work/2010/03/03/48-cores-contest/ ). The deadline is passed.

Here is mine: http://nuit-blanche.blogspot.com/2010/03/what-would-you-do-with-48-cores.html

7 comments

[ 3.7 ms ] story [ 28.2 ms ] thread
jacquesm
Interesting topic, but why would an X number of cores be the reason of approaching this particular problem ? It seems as though 'more is better' is the way to go here, so a cluster of rented boxes would be one way of getting there, a couple of high end GPGPU cards would be another. The number of cores in a box is just another multiplicand under the bar to get to the time-to-a solution.
nuitblanche
Let me rephrase the question to see if I understand it correctly:

Why do you go for a single piece of hardware when in fact, you could already do this on several GPGPUs and one CPU or use several CPUs on the cloud.?

It could probably be done using these techniques but the answer resides in the fact that if the computation of the phase transition is done very fast for a particular "measurement matrix" (i.e. a particular "sensor") then we could foresee its use in near-real time in instances where the "measurement matrix" would be time dependent (such as the atmosphere....). This would be important in that it would allow to figure out if these new "sensors" can be actually used (as opposed to being mere lab constructions).

jacquesm
I figure a GPGPU implementation would be much faster on hardware you can buy in the corner computer store today as compared to that 48 core machine.

From my limited understanding of that type of problem GPGPUs are well tailored to doing those computations, for the exact same reason a multi-core box would be a good match.

But price-performance wise the GPGPU wins hands down.

nuitblanche
The small problem with GPGPU is that one aspect of this project is to look at several reconstruction solvers (as they are being called in compressed sensing). Not all of these solvers are clearly amenable to a GPU approach. The idea is really to make it easy for new reconstruction algorithm to be tested on these phase transition maps as opposed to a dedicated effort at re-thinking the algorithm so that it can fit the GPU architecture.
boyter
Id take it to a LAN and lord it over everyone else.

Seriously though... I dont need that much CPU power. I would prefer 4 cores and enough disk bandwidth to keep them busy.

nailer
Spawn worker threads.
[dead] losethos
The real question is "What would you do when all your clients have 48 cores."