Sign in with
Sign up | Sign in

Benchmark Results: WinRAR

Updated: Tuning C'n'Q: Maximize Power And Performance, Part 2

For this test, we've selected a 616MB folder full of files (the installer for Adobe Photoshop CS4), compressed it into an .RAR file under the “Best” compression profile, and then uncompressed it to the desktop. Let's check out the processor utilization for these two tasks.



Just like its internal benchmark, WinRAR can fully utilize two cores and put heavy load on more than two cores. But if you're more likely to simply decompress files, a dual-core processor is mostly enough. Processor utilization only went above 50% on few occasions.

This method of benchmarking does have its drawbacks--higher variability in the results. Even with repeated runs, some of our numbers will be slightly slower or faster. Overall, enabling power management exacts a measurable and certain penalty on all three processors. Unlike SuperPi, changing p-state transition time has more of an impact here, helping us regain most of the performance.

Although it seems large at first glance, the real performance difference between these three processors when decompressing is roughly three seconds. A larger, more noticeable difference can be seen when compressing files. So, this is where we will focus. Look at how the field is spread out. The dual-core processors are about 100 seconds slower than the Phenom II X3 710, and the Phenom II X3 710 is about 20-30 seconds slower than the Phenom II X4 945 and 955.

We saw a similar situation with the 3DMark Vantage earlier. The Athlon II X2 250 has the lowest average power consumption, followed by the Phenom II X3 710 at default and Optimized settings. But that’s just looking at average power consumption.

Looking at total power consumption, the Phenom II X4 945 finishes better than the others. It’s the second-fastest processor, but it consumes the least amount of power. Much like SuperPi, this benchmark should give us a rough idea of actual power consumption with multi-threaded applications that can scale to four or more cores. The moral of the story so far: don’t judge power consumption by looking at average numbers. At default settings, even the Phenom II X4 955 can consume less power than the Athlon II X2 250 in this benchmark.

React To This Article