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

SiSoftware Sandra Cryptography Benchmark

Intel paved the way for dedicated cryptography instructions with Westmere-EP's AES-NI instructions. Since the Westmere-EP article, I've been using Sandra's cryptography benchmark as a way to test hardware encryption speed.

You can see that the additional cores and memory bandwidth more than make-up for Sandy-Bridge-EP's lack of clock speed compared to Westmere-EP. Those are some pretty big performance increases across the board, but especially in the AES256 test.

TrueCrypt 7.1a Benchmark

TrueCrypt is a software system for establishing and maintaining an on-the-fly-encrypted volume (data storage device). On-the-fly encryption means that data is automatically encrypted right before it is saved and decrypted right after it is loaded, without any user intervention. No data stored on an encrypted volume can be read (decrypted) without using the correct password/keyfile(s) or correct encryption keys. Entire file system is encrypted (e.g., file names, folder names, contents of every file, free space, meta data, etc).

While I always like to run synthetic benchmarks to establish expectations for a new processor or platform, I much prefer real-world application tests when I can find them. I recently got turned-on to TrueCrypt and, luckily, it has its own built-in benchmark.

It's pretty obvious here that Sandra's synthetic tests translate pretty well to TrueCrypts "real-world" tests.

SPECjbb 2005 v1.07

SPECjbb2005 (Java Server Benchmark) is SPEC's benchmark for evaluating the performance of server side Java. Like its predecessor, SPECjbb2000, SPECjbb2005 evaluates the performance of server side Java by emulating a three-tier client/server system (with emphasis on the middle tier). The benchmark exercises the implementations of the JVM (Java Virtual Machine), JIT (Just-In-Time) compiler, garbage collection, threads and some aspects of the operating system. It also measures the performance of CPUs, caches, memory hierarchy and the scalability of shared memory processors (SMPs). SPECjbb2005 provides a new enhanced workload, implemented in a more object-oriented manner to reflect how real-world applications are designed and introduces new features such as XML processing and BigDecimal computations to make the benchmark a more realistic reflection of today's applications.SPECjbb2005 is a widely used, industry standard benchmark.

In a nutshell, each "warehouse" spawns an independant thread which determines the concurrency of the benchmark run. Systems tested have a an expected peak number of warehouses (X) that correspond to the total number of "hardware threads" in the machine. Scores are output as "Business Operations per Second (BOP/s)", and are based on average throughputs of X, X+1, X+2, X+3, etc, up to and including 2X.

I wasn't trying to set a SPECjbb record here. I just wanted to run the workload as a comparison of the two platforms being evaluated here. I used the Oracle JRokit 64bit JVM (jrockit-jdk1.6.0_29-R28.2.2-4.1.0-windows-x64). I ran the tests with a single JVM and the following commandline options: -Xms3700m -Xmx3700m

On both systems, "hardware prefetch" and "adjacent cache line prefetch" options were disabled in the BIOS.

Even with my non-optimized SPECjbb configuration, the E5-2687W system turns in quite the spectacular performance. Maybe one of these days I'll get the multi-JVM runs and commandline options figured out for this benchmark and REALLY get some stellar numbers!

Printed from (,3.html)