topic Many-cores hit the memory wall in Intel® Moderncode for Parallel Architectures

Many-cores hit the memory wall

aminer10 — Mon, 28 Oct 2013 18:16:14 GMT

Many-cores hit the memory wall

http://storagemojo.com/2008/12/08/many-cores-hit-the-memory-wall/

Thank you,
Amine Moulay Ramdane.

"The performance is

aminer10 — Mon, 28 Oct 2013 18:44:52 GMT

"The performance is especially bad for informatics applications
—data-intensive programs that are increasingly crucial to the labs’
national security function."

http://spectrum.ieee.org/computing/hardware/multicore-is-bad-news-for-supercomputers

This is what i have said about memory bound applications and/or disk
bound applications such as database engines etc.

Amine Moulay Ramdane.

Hello,

aminer10 — Mon, 28 Oct 2013 23:17:16 GMT

Hello,

I have asked you about the technologies to scale and to speed up the hardisk system and the memory ...
here is how you can do it with your hardisk system:

To get much more performance from hardisk system, choose a RAID 10 called RAID 0+1, it combines features of
RAID 0 + RAID 1, that means it provides both stripping and mirroring, in RAID 1-0 you start with a base configuration of 4 disks, and by striping volume data across multiple disks you will speed up the reads and writes, and by mirroring the disks this will make your hardisk system fault tolerant and will then higher the availability.

I think that the memory system inside the computer is using also the same technic as the stripping in RAID 10 to speed up memory, we call it interleaved memory.

Or you can use the stacked memory as in the future NVEDIA...

http://www.forbes.com/sites/jasonevangelho/2013/03/19/nvidias-volta-gpu-launches-in-2016-delivers-1tbs-of-memory-bandwidth/

Thank you,
Amine Moulay Ramdane.

Hello,

aminer10 — Mon, 28 Oct 2013 23:17:16 GMT