Andrew,

Regarding

On Sun, Aug 16, 2015 at 12:54 AM, J. Andrew Rogers <[email protected]>
wrote:

> The only modern barrel processor is Intel's Xeon Phi, particularly the new
> version that will be coming out. Again, if you design codes like for a CPU,
> the performance will be unremarkable. And unfortunately, that is how Intel
> is positioning it to developers — like a CPU with a lot of threads — when
> it really isn’t. So most programmers will design terribly inefficient
> algorithms for it.
>
> ...
>
...


> Also, idiomatic C/C++ etc has a CPU-targeted orthodoxy built into it and
> is all programmers know. C++ works fine for barrel processors but idiomatic
> code looks different than for CPUs and no one teaches idiomatic C/C++ for
> barrel processors.
>

I am interested in optimizing OpenCog's innards for the new Xeon Phi, so I
am curious for any pointers or further insights you can offer on this.

Suppose one has a large graph with heterogeneous structure and various
complex access patterns -- what are some links or resources giving good
information on how to do efficient implementation for e.g. the new Xeon
Phi....   The core code in question is C++ currently...

thx
Ben


-- 
Ben Goertzel, PhD
http://goertzel.org

"The reasonable man adapts himself to the world: the unreasonable one
persists in trying to adapt the world to himself. Therefore all progress
depends on the unreasonable man." -- George Bernard Shaw



-------------------------------------------
AGI
Archives: https://www.listbox.com/member/archive/303/=now
RSS Feed: https://www.listbox.com/member/archive/rss/303/21088071-f452e424
Modify Your Subscription: 
https://www.listbox.com/member/?member_id=21088071&id_secret=21088071-58d57657
Powered by Listbox: http://www.listbox.com

Reply via email to