Nehalem: Xeon Gets Core i7 Upgrade
Posted on: 03/29/2009 05:00 AM

FlamMap FSPRO
FlamMap is a fire behavior mapping and analysis program that computes potential fire behavior characteristics (spread rate, flame length, fireline intensity, etc.) over an entire FARSITE landscape for constant weather and fuel moisture conditions.

The software models forest fire. Basically, the program calculates fire behavior for multiple weather scenarios based on historic weather observations, then runs a Monte Carlo simulation of an existing fire to get burn probabilities across a landscape.
I was originally turned-on to this application (and its use as a benchmark) in our very own forums. Using real-world applications as testing workloads is always nice, and FlamMap is a great example.

For this article I have a whole new FlamMap/FSPRO workload that seems to make more efficient use of all the threads/core available in the test machines.



I have it on pretty good authority (the author of the benchmark) that the W5580 munbers above are very good. In fact, they are even better than current generation, four socket quad-core boxes.

Power Useage



As the test ran and the inside of the Harpertown machine got warmer, the fans started spinning faster and faster. I point that out because on the above graph you can see the total power useage for the Harpertown machine increase as the test progresses. And who says fans don't draw that much power?!

Performance-per-Watt



MyriMatch
MyriMatch is a tool designed to take experimental data from shotgun proteomics experiments and compare those spectra against sequences in a known database of proteins. Whether the program is being run in a single-computer environment or across an entire cluster of processing nodes, it is able to optimally divide work in a much more efficient way than many other database search programs. This is because it only generates candidate sequences from the known database once for the entire set of spectra instead of once for every spectrum. Thus, for each candidate sequence generated, it is compared against every spectrum. The spectra keep a certain (user-defined) number of candidate sequences that had the highest scores.

MyriMatch is designed to take advantage of (symmetric) multiprocessor systems by multithreading the database search. A search process on an SMP system will spawn one worker thread for each processing unit (where a processing unit can be either a core on a multi-core CPU or a separate CPU entirely). The main thread then generates a list of √?¬¨worker numbers√?¬ģ which is equal to the number of worker threads multiplied by this parameter. The worker threads then take a worker number from the list and use that number to iterate through the protein list.
MyriMatch is one of the newest workloads in my bag of tricks. It is a pretty complex test that I'm not sure if I even understand. In a nutshell, MyriMatch is designed to take experimental data from shotgun proteomics experiments and compare those spectra against sequences in a known database of proteins.



I was especially curious to see how this test would turn out because the last time I ran it, a dual-core Woodcrest system beat a quad-core Clovertown system. Back then I postulated that memory bandwidth and bus contention were the cause of the quad-core slow-down, so I was curious to see how Nehalem would handle things with its abundance of memory bandwidth and physical/logical cores.

It didn't disappoint, but I suspect that turning off SMT would have yielded even better results.

Power Useage



Performance-per-Watt




Printed from 2CPU.com (http://www.2cpu.com/contentteller.php?ct=articles&action=pages&page=nehalem_xeon_gets_core_i7_upgrade,6.html)