<div><span class="gmail_quote">On 3/23/07, <b class="gmail_sendername">Michael Hannon</b> <<a href="mailto:email@example.com">firstname.lastname@example.org</a>> wrote:</span>
<blockquote class="gmail_quote" style="PADDING-LEFT: 1ex; MARGIN: 0px 0px 0px 0.8ex; BORDER-LEFT: #ccc 1px solid">Peter St. John wrote:<br>.<br>.<br>.<br>> I wish I know more about the SAGE (machine) that hosts the SAGE
<br>> (software) that was used for this</blockquote>
<blockquote class="gmail_quote" style="PADDING-LEFT: 1ex; MARGIN: 0px 0px 0px 0.8ex; BORDER-LEFT: #ccc 1px solid">See the appended blurb from one of our sales guys.<br><br> - Mike<br>
<br>I thought you might find this interesting. The 8-way opteron system is a<br>Western Scientific FusionA8 with 64GB RAM. It ran some high profile<br>calculations, which have appeared in the national news today ("Sage
<br>Supercomputer" refers to the machine):</blockquote>
<div>Thanks Mike. I drilled down and the datasheet is <a href="http://www.wsm.com/datasheets/fusiona8_ds_030507.pdf">http://www.wsm.com/datasheets/fusiona8_ds_030507.pdf</a></div>
<div>Seems like some nice mission-criticality characteristics, e.g. the redundant power supplies built into the case, but if it were me, nutty mathematicians doing something nutty, would I prefer a large number of disposalble nodes? Not that they need to go faster than 3 days (compared to 4 years of design) but they could futz around with it throughout the four years for cheap. But I'm preaching to the Choir?