Intel Westmere 32nm Launch & Clarkdale Core i5-661 CPU Review

Synthetic Benchmarks: Memory

Lavalys Everest Ultimate v5.02

Everest Ultimate is the most useful tool for any and all benchmarkers or overclockers. With the ability to pick up most voltage, temperature, and fan sensors on almost every motherboard available, Everest provides the ability to customize the outputs in a number of forms on your desktop. In addition to this, the memory benchmarking utility provides a useful tool of measuring the changes to your memory sub-system.


As we predicted earlier in this article, the relocation of the integrated memory controller from the CPU die to the GPU die has had some serious impact on the overall memory bandwidth. Despite having a dual-channel DDR3-1333 memory interface, and a theoretical 21.2GB/s of memory bandwith, the Clarkdale chip has up to 75% lower bandwidth than the Lynnfield models. Having said that, it still has better overall memory bandwidth than any of the non-Nehalem-based chips. But will this bandwidth shortage hamper it in real-life applications? We will find out shortly.



If you thought the bandwidth numbers looked bad, the memory latency is downright atrocious. The i5-661 posted worst latency numbers than the Core 2 processors, which have their memory controller on the northbridge and have to communicate through the ancient front-side bus.


Will ScienceMark 2.0 paint a different picture? Let's find out.

ScienceMark v2.0

Although last updated almost 3 years ago, and despite its rudimentary interface, ScienceMark v2.0 remains a favorite for accurately calculating bandwidth on even the newest chipsets.


In a fluke for the ages, our Everest and ScienceMark bandwidth results were exactly the same for the i5-661, which never happens with these two benchmarks. The results still weren't good though, with the Clarkdale chip seriously lagging behind the Lynnfield and Bloomfield processors.


While the i5-661's ScienceMark memory latency figure is not quite as bleak as the one we achieved in Everest, it is still approximately 70% higher than what we are used to seeing from a Nehalem-based chip. Obviously having the IMC off-die was a costly decision when it comes to the memory sub-system's performance. But as previously discussed, it's a necessary evil for the good of the integrated graphics processor.

Although the i5-661 hobbled its way through the memory benchmarks, these are still just synthetic numbers. The real-life applications and games are what count. Let's check those out next.