<p dir="ltr"><br>
On Aug 8, 2014 1:23 PM, "Eduardo Menendez" <<a href="mailto:eariel99@gmail.com">eariel99@gmail.com</a>> wrote:<br>
><br>
> Hi, <br>
> Thanks for the replies to my question. Some answers were devoted to solid state devices (SSD), but I had asked about SAS (Serial attached SCSI) in comparison to SATA disks. I know that SAS are better for high I/O load. The point is how critical ist that for QE. <br>
><br>
> I have to choice between (better CPU with SATA) vs (worse CPU with SAS).<br>
><br>
><br>
> I have the feeling (from the times of IDE hard disks), that under heavy I/O load (using swap memory), the use of CPU drops to 1% and calculations are impractical. I guess it is still true, i.e., heavy I/O are to be avoided in ab initio calculations. <br>
</p>
<p dir="ltr">There is no simple answer to your choice that is always valid. Many other factors can have a big impact, bigger than SAS vs sata. E.g. High capacity disks have higher throughput. Thus your approach is not very useful. You need to find what the weakest point in your configuration is. That can be ram, cpu clock or type, disk, whether to add an SSD or not, a dedicated scratch disk. And more.<br>
</p>
<p dir="ltr">><br>
><br>
><br>
><br>
><br>
> Eduardo Menendez Proupin<br>
> Departamento de Fisica, Facultad de Ciencias, Universidad de Chile <br>
> URL: <a href="http://www.gnm.cl/emenendez">http://www.gnm.cl/emenendez</a><br>
><br>
> “Science may be described as the art of systematic oversimplification.” Karl Popper<br>
><br>
><br>
> _______________________________________________<br>
> Pw_forum mailing list<br>
> <a href="mailto:Pw_forum@pwscf.org">Pw_forum@pwscf.org</a><br>
> <a href="http://pwscf.org/mailman/listinfo/pw_forum">http://pwscf.org/mailman/listinfo/pw_forum</a><br>
</p>