Has anyone tested with ECC memory? For non-ECC memory, it partitions off 1/8 of the installed RAM to do "virtual ECC". I was wondering if the parity calculation on every byte of memory was slowing the drive down. The benchmarks I've seen show it runs fast, but not as fast as it should for having 2 sata channels on the drive.
Normally ECC memory is a bit slower than regular desktop memory, but it shouldn't make a difference here. However, relieving the IO processor of the parity check could since it seems to be a bottleneck.
Is ICH10 good enough for RAID 0 on two of these (4 sata channels)? Intel RAID drivers use to allocate a chunk of memory for cache. Do they still do that? Hardware RAID seems like somewhat of a waste for 0, but a big 2GB cache like the higher end Areca's support might be nice. System RAM would be much faster and cheaper, but Vista is too stupid to use it on its own.
Normally ECC memory is a bit slower than regular desktop memory, but it shouldn't make a difference here. However, relieving the IO processor of the parity check could since it seems to be a bottleneck.
Is ICH10 good enough for RAID 0 on two of these (4 sata channels)? Intel RAID drivers use to allocate a chunk of memory for cache. Do they still do that? Hardware RAID seems like somewhat of a waste for 0, but a big 2GB cache like the higher end Areca's support might be nice. System RAM would be much faster and cheaper, but Vista is too stupid to use it on its own.