Differences

This shows you the differences between two versions of the page.

--- cluster:123 [2013/10/12 08:01]
hmeij
+++ cluster:123 [2013/10/23 14:52]
hmeij [Summary]
@@ Line 47: / Line 47: @@
 The Libert family rejoices.  The Dell family moves out. The end.
+==== Update ====
+The table below contains data for a cluster whose nodes are all on the Infiniband switch (and also ethernet switch for provision and data).  They also contain a 15K 300 GB SAS drives each for access to local fast disk (Gaussian users). It still deployes the 8-core CPUs, thus 16 pysical cores per node, 32 hyperthreaded cores per node and in both cases 256 GB of memory.
+^  Tnodes^  Tcores^  THcores^  Tmem gb^  Watts^  %of Dell^  TEnergy^  TEnergy $/Y^  TEsavings $/Y^  Quote $^   ROI Y^  Gflops^
+|  10|  160|  320|  2,560|  3,650|  29|  7,300|  7,994|  19,495|  76,866|  3.9|  3,328|
+|  9|  144|  288|  2,304|  3,285|  26|  6,570|  7,194|  20,295|  69,290|  3.4|  2,995|
+|  8|  128|  256|  2,048|  2,920|  23|  5,840|  6,395|  21,094|  61,714|  2.9|  2,662|
+|  7|  112|  224|  1,792|  2,555|  20|  5,110|  5,596|  21,893|  54,138|  2.4|  2,329|
+==== Summary ====
+The Dell racks were bought in 2006. They contain 30 compute nodes, two UPS units, two disks arrays and two switches. Measurements of 2/3rds of the compute nodes with a Kill-A-Watt meter yields an average consumption of 418.4 watts (if the nodes are computing or not).  That totals to 109,956 KwH/year for power, a low water mark. Measurements at the utility panel for one of the racks yields a consumption of 126,000 KwH/year. Cooling requirements (not measured) are assumed to be equal to that.
+Using the low water mark, and a cost per KwH (inclusive of cogen generation costs, maintenance, CP&L power imports, heat reclamation costs, etc), the total cost for both  power and cooling consumption is estimated at $27,489 per year for the Dell racks.
+New hardware could replace the Dell's functionality and reduce power/cooling needs in the data center while yielding significant savings.  An 8-node cluster with each node comprised of dual 8 core CPUs, each with 256 GB of memory and a 15K 300 GB hard disk, all nodes connected to a high throughput/low latency switch, would match or exceed key parameters such as gigaflops of computational power and number of job slots provided (with hyperthreading enabled).
+Such a cluster would consume 77% less energy generating $21,094 in saving per year (after accounting for energy needs of that 8-node cluster).  This implies that in 2.9 years the cost of acquiring that 8-node cluster ($61,714) will be recouped. Based on the low water mark.
 \\
 **[[cluster:0|Back]]**

DokuWiki

User Tools

Site Tools

Differences

Page Tools