Test Setup:

As always, we did our best to deliver a clean set of benchmarks, with each test repeated three times and an average of those results is what we’re reporting here. In the rare case where performance was inconsistent, we continued repeating the test until we got three results that were consistent.

Test Setup:

• Shuttle X27-D with an Intel Atom 330 (2x1.6GHz with HyperThreading) CPU
  
• Gigabyte GA-GC230D with an Intel Atom 230 (1x1.6GHz with HyperThreading) CPU
  
• Jetway JNC62K with an AMD Sempron LE-1200 (1x2.1GHz) CPU
  
• 2GB DDR2 533MHz at 3-3-3-9-2T timings (SPD)
  
• Enermax 285W mATX PSU / Seasonic 60W DC-DC PSU
  
• Seagate 7200.10 200GB SATA hard drive
  
• Windows Vista 32-bit SP1

Lavalys Everest Memory Performance

Website: Lavalys



Clearly even the extremely inexpensive single core AMD Sempron CPU has oodles of memory bandwidth compared to the Atom. We would have preferred to test the 8W dual-core Atoms against the latest AMD 15W 2650e, but unfortunately there's more chance of finding hens teeth than one of these in retail.

The Shuttle X27-D drops a bit of performance to the already limited Gigabyte GA-DC230D - we can attribute this to the greater stress being put on the limited front side bus and single channel memory from both cores, as well as the Gigabyte BIOS offering more performance options for the memory.

SiSoftware Sandra Lite XII.2008.SP2c (14.24)

Website: Sisoftware



In Sisoft Sandra the AMD board dwarfs the Atoms, with the single-core Atom still faster than the dual-core too buy a few hundred MB/s. This continues when we look at latency too, but strangely Sisoft Sandra claims the AMD setup is now the slowest of the lot, despite the faster memory clock and direct connect architecture, which should be more efficient than the slower front side bus-MCH-memory bus combination from Intel.