himanshu khandelia wrote:
Hi Carsten,
The benchmarks were made is 1 NIC/node, and yet the scaling is bad.
Does that mean that there is indeed network congestion ? We will try
using back to back connections soon,
Hi Himanshu,
In my opinion the most probable scenario is that the bandwidth of
Hi Carsten,
The benchmarks were made is 1 NIC/node, and yet the scaling is bad.
Does that mean that there is indeed network congestion ? We will try
using back to back connections soon,
-himanshu
maybe your problem is not even flow control, but the limited network
bandwidth which is shared
Hi,
We tried turning on switch control on our local cluster
(www.dcsc.sdu.dk) but were unable to achieve any improvement in scale
up whatsoever. I was wondering if you folks could shed light upon how
we should go ahead with this. (We have not installed the all-to-all
patch yet)
The cluster
Hi Himanshu,
maybe your problem is not even flow control, but the limited network
bandwidth which is shared among 4 CPUs in your case. I also have done
benchmarks on Woodcrests (2.33 GHz) and was not able to scale an 8
atom system beyond 1 node with Gbit Ethernet. Looking in more detail,
the
Hi Carsten,
Thank you very much for the prompt reply.
I know very little about network architecture, and therefore
understand your explanation only partly. Based on what you say,
however, would it be fair to conclude that on the quad core
woodcrests, it will not be possible to improve scaleup
himanshu khandelia wrote:
Hi Carsten,
Thank you very much for the prompt reply.
I know very little about network architecture, and therefore
understand your explanation only partly. Based on what you say,
however, would it be fair to conclude that on the quad core
woodcrests, it will
6 matches
Mail list logo