MathematicaMark8 Benchmark Report for GPGPU, Compiled Code & Multi-Core Systems
- To: mathgroup at smc.vnet.net
- Subject: [mg117042] MathematicaMark8 Benchmark Report for GPGPU, Compiled Code & Multi-Core Systems
- From: Syd Geraghty <sydgeraghty at me.com>
- Date: Tue, 8 Mar 2011 05:37:03 -0500 (EST)
Bruce & MathGroup regulars,
There have been some requests for performance data in MathGroup recently in order to make informed decisions regarding the purchase of hardware to optimize use of Mathematica 8.
Are there any variations of the MathematicMark8 ( plus results ) that are optimized for:
1 CUDA/OpenCL with Midrange or Professional ATI / nVidia graphics cards
2 Compiled Code
3 Multi-Core Systems (up to 12 core say)
which would give us some better information in selecting our next hardware configurations?
I ran my MacBook Pro as a 4-node homogenous cluster and considering it is a 4 year old machine I am pleasantly surprised that the highest rated system only beats my setup by 1 : 0.67 (see below).
{
{"Machine Name:", "4-node homogeneous cluster"},
{"System:", "MacOSX-x86-64"},
{"Date:", "March 7, 2011"},
{"Mathematica Version:", "8.0.0"},
{"Benchmark Result:", "0.67"}
}
vs top system in standard benchmark comparison
3.07 GHz Core i7-950 (8 Cores) 1.00
Windows 7 Pro (64-bit) Desktop
Thanks in advance for any helpful data.
Cheers .... Syd
Syd Geraghty B.Sc, M.Sc.
sydgeraghty at mac.com
Mathematica 8.0 for Mac OS X x86 (64-bit) (November 6, 2010)
MacOS X V 10.6.5 Snow Leopard
MacBook Pro 2.33 GHz Intel Core 2 Duo 2GB RAM