Intel Xeon Phi release announced

[Tangled-Universe]

As the title suggests; yesterday Intel announced the release of it's Larrabee based Xeon Phi co-processor product family:

http://www.tomshardware.com/reviews/xeon-phi-larrabee-stampede-hpc,3342.html

According to Intel this technology process parallell code 2-3x faster than a regular Xeon CPU.
Given the fact that one Xeon Phi card contains 60 cores, 1GHz each, this would be equivalent to 120-180GHz of raw CPU power.
That's roughly 10 i7-2600K's (or more?) for 2500 dollars.

According to Intel any current x86 code can run on it and doesn't need specific recompiling.

If this is true(??) then this would be a (relatively) dirt-cheap rendering option???

[pokoy]

Thanks, totally missed that one.

The graphs are showing roughly a factor of 3 compared to a E5@2.6 GHz for most tasks. Also, 8GB of memory sounds a bit disappointing.
We'll have to wait until some more testing has been done, but it's some very interesting development.

[Tangled-Universe]

Yes that's what I also read after posting this. The 3 fold increase is over dual Xeon I believe. So that's still almost 6 times over a single Xeon.

It seems that in some cases the benefit can be much bigger, like in financial monte carlo modelling. I don't know if it's remotely connected to Monte Carlo GI but what it does show is that it's quite dependant on the task.

8GB may seem low, but that's still roughly 150MB per core/thread. TG2, for instance, can use 50MB as lowest amount of RAM allocated to each thread. The RAM is only for execution, not for storage.
Scene data would be stored into the system/node's RAM, which would obviously be bigger.