<div dir="ltr"><div dir="ltr">John,<div><br></div><div>Did you see this? <a href="https://www.anandtech.com/show/12828/intel-launches-optane-dimms-up-to-512gb-apache-pass-is-here">https://www.anandtech.com/show/12828/intel-launches-optane-dimms-up-to-512gb-apache-pass-is-here</a></div><div><br></div><div>It's the DDR4 compatible DIMM slot version of Optane as persistent memory. I have a set of Diablo DIMMs still. These are supposed to act in the same general way. Livable speeds (compared to DDR4 DIMMs), up to 512GB per DIMM with the ability to protect and persist memory ranges through reboots/POST and power cycles. I haven't played with the Optane version firsthand so I'm going off of what I've heard. </div><div><br></div><div>--Jeff</div></div></div><br><div class="gmail_quote"><div dir="ltr">On Tue, Sep 25, 2018 at 8:39 PM John Hearns via Beowulf <<a href="mailto:beowulf@beowulf.org">beowulf@beowulf.org</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Over on the SLURM list we recently had a very good discussion on swap<br>
space and job suspension. I commented that for large memory systems we<br>
should be seeing tiered RAM devices by now.<br>
<br>
<br>
I had a look at the Intel pages on Optane memory. It is definitely<br>
being positioned as a fast file cache, ie for block<br>
oriented devices.<br>
<a href="https://www.intel.com/content/www/us/en/architecture-and-technology/intel-optane-technology.html" rel="noreferrer" target="_blank">https://www.intel.com/content/www/us/en/architecture-and-technology/intel-optane-technology.html</a><br>
<br>
I mentioned Diablo Technologies, who I think have gone out of<br>
business. They had available a DRAM form-factor high capacity memory<br>
device using flash. This worked by tiering RAM memory - ie a driver in the Linux<br>
kernel would move little used pages to the slower but higher capacity<br>
device.<br>
I though the same thing would apply to Optane, but it seems not.<br>
Can anyone comment? I do not expect anyone under NDA to comment of course.<br>
<br>
I have probably said this on many forums - we are seeing the trend to<br>
wards higher and higher memory systems, which enable more detailed<br>
meshes, bigger models and in-memory databases. But do we need huge<br>
amounts of expensive and fast DRAM to do this?<br>
_______________________________________________<br>
Beowulf mailing list, <a href="mailto:Beowulf@beowulf.org" target="_blank">Beowulf@beowulf.org</a> sponsored by Penguin Computing<br>
To change your subscription (digest mode or unsubscribe) visit <a href="http://www.beowulf.org/mailman/listinfo/beowulf" rel="noreferrer" target="_blank">http://www.beowulf.org/mailman/listinfo/beowulf</a><br>
</blockquote></div><br clear="all"><div><br></div>-- <br><div dir="ltr" class="gmail_signature" data-smartmail="gmail_signature"><div dir="ltr"><div><div dir="ltr">------------------------------<br>Jeff Johnson<br>Co-Founder<br>Aeon Computing<br><br><a href="mailto:jeff.johnson@aeoncomputing.com" target="_blank">jeff.johnson@aeoncomputing.com</a><br><a href="http://www.aeoncomputing.com" target="_blank">www.aeoncomputing.com</a><br>t: 858-412-3810 x1001 f: 858-412-3845<br>m: 619-204-9061<br><br>4170 Morena Boulevard, Suite C - San Diego, CA 92117<div><br></div><div>High-Performance Computing / Lustre Filesystems / Scale-out Storage</div></div></div></div></div>