Intel E5-2600 Series "Sandy Bridge-EP" Xeons
Posted on: 03/06/2012 05:41 PM

Black Scholes Kernel

In 1973, Black and Scholes developed a model for estimating the value of a stock option, which has been refined over the years to remove several assumptions, thus making techniques based on the model very accurate. Today, financial analysts rely on algorithms based on the Black and Scholes technique to determine the price of a stock option.

This benchmark constitutes of a kernel that implements a derivative of the Black and Scholes technique. The code was developed at SunGard, and utilizes a continuous fraction technique, which is more accurate than the more traditional polynomial approximation technique.

The workload for this benchmark comes in the form of loop iterations internal to the code. The number of steps used in calculating option price, is set to 1e8 (100,000,000) by default. This value can be changed via command line parameter. The number of threads to use can also be specified as a command line parameter.

For my tests, I ran a specially compiled version of the Black Scholes Kernel with 24 threads for Westmere-EP, 32 threads for Sandy Bridge-EP, a problem size of 10,000,000,000 and processor affinity enabled.

We've shown over and over again that Black Scholes likes clock speed and cores, and these results are no exception. The difference between the two platforms correlates almost perfectly to the speed and number of cores/threads present in the system.

Black Scholes Kernel Power

I've covered the whole power usage thing many times in previous articles, but if you're not familiar, let me recap. First up, I will present the actual wattage used over a predefined test period, as logged by the Exotech power meter. Remember, only the machine itself is plugged in to the meter.

The "test period" here, and in all subsequent graphs, is a period that starts five seconds before the machine leaves the idle state and starts the workload. The period ends five seconds after the slowest machine in the test completes the workload.

In this graph, I break down the power usage into Joules or, Watt-seconds. "Total Joules" represents the amount of energy used by the machine over the entire test period (as previously defined above). The "Workload Joules" represent the amount of power required to complete the actual benchmark, from the second the machine leaves the idle state until the second it completes the test.

From here on out I will not be adding commentary to any of the power graphs unless something really jumps out as being odd. I take these readings and make these graphs to aid you in deciding what is going to work best for you based on your own usage model. Because of this, the same numbers will mean different things to different people.

Euler3d CFD Benchmark v2.2

The benchmark testcase is the AGARD 445.6 aeroelastic test wing. The wing uses a NACA 65A004 airfoil section and has a panel aspect ratio of 1.65, taper ratio of 0.66, and a quarter-chord sweep angle of 45 deg. This AGARD wing was tested at the NASA Langley Research Center in the 16-foot Transonic Dynamics Tunnel and is a standard aeroelastic test case used for validation of unsteady, compressible CFD codes.

The CFD grid contains 1.23 million tetrahedral elements and 223 thousand nodes . . . . The benchmark executable advances the Mach 0.50 AGARD flow solution. A benchmark score is reported as a CFD cycle frequency in Hertz.

Again the E5 Xeons fail to disappoint with a 160% performance advantage in Euler3d.

Euler3d CFD Benchmark v2.2 Power




Printed from 2CPU.com (http://www.2cpu.com/contentteller.php?ct=articles&action=pages&page=121,4.html)