MacMini i5 2.5Ghz
16 GB Ram
Windows 7 Ultimate (Using Bootcamp)
330 240 GB SSD
Intel Solid State Drive Toolbox 3.0.3
Mac OS X 10.7.4
Background:
- Bootcamp doesn't natively support AHCI so using IDE mode.
- The Toolbox lists this drive model as "Unknown"
- Fresh Install of all Operating Systems
I setup various tests using IO meter, XBENCH, and Crystal Disk Mark and do not get anywhere close to the specs listed by Intel. For IO meter, I performed three tests using 1GB, 8GB, and 24GB test files with Queue Depths of 32. I used the same benchmark parameters as listed in the specs: 4KB Sequential Reads/Writes.
Below are screenshots of the results of the tests that I performed. I am assuming that the IDE mode is one issue however, if that were the case, it seems that I should have gotten outstanding results with XBENCH on the Mac OS X partition.
1GB Sequential Read: 20,685 IO and 80.80 MB/s
1GB Sequential Write: 29,686 IO and 116 MB/s
8GB Sequential Read: 18,929 IO and 74 MB/s
8GB Sequential Write: 29,644 IO and 116 MB/s
24GB Sequential Read: 14,464 IO and 57 MB/s
24GB Sequential Write: 29,688 IO and 116 MB/s
XBENCH Result in Mac OS X 10.7.4: Sequential READ = 32 MB/s and Sequential WRITE = 297 MB/s
Crystal Disk Mark in Win 7 Partition
AHCI mode makes a huge difference, you're fortunate to get what you have now without it. The 4K scores are very dependent on AHCI mode, particularly in benchmarks. I don't understand why one benchmark would be better in IDE mode.
It looks like you are barely in SATA III performance territory, if at all. Does that PC support SATA III?
parsec - thanks for your reply.
Yes, it is unfortunate - I guess I should have done my homework a bit more. The 2011 MacMini does support SATA III but mine shipped with a SATA II 5400 Drive.
Even in IDE mode, the performance is without a doubt much better but I had hoped to run VMware Workstation on the Windows side with a VSA (Virtual Storage Appliance) and nested ESXi's for testing and studying.
I'll test around on the Mac OS X side and see if I can my hands on a more reliable disk bench marking tool and post my results.
