In the database workload, our freshly SE'ed drives actually outperform the tweaked drives.
Intel's X25-M is hit harder than OCZ's Vertex 2; the X25-M's performance is nearly cut in half across all tested queue depths. Meanwhile, the Vertex 2 at least stays pretty consistent throughout the different queue depths.
Intel's X25-M gets slammed by the same performance drop in our file server benchmark. OCZ's Vertex 2 is largely unaffected after the various tweaks.
There are minor differences between drives in the Web server workload. Intel's X25-M turns things around a bit, showing a slight performance increase, while the OCZ Vertex 2 takes a minor hit after all of the tweaks are applied.
The workstation workload demonstrates similar results as the file server test. Mainly, Intel's performance is almost cut in half. OCZ's Vertex 2 isn't affected much at all.
- Optimizing Precious Solid-State Storage
- Start By Enabling AHCI (With TRIM)
- Disable System Restore And Drive Indexing
- Disable The Page File And Hibernation
- Disable Prefetch, SuperFetch, And Windows Write-Cache Buffer Flushing
- Disable SuperFetch (Again), Windows Search, ClearPageFileAtShutdown, And LargeSystemCache
- Power Settings Adjustment
- Test Setup And Method
- Benchmark Results: CrystalDiskMark
- Reclaiming Storage Capacity
- Benchmark Results: Iometer
- Benchmark Results: Iometer Streaming
- Benchmark Results: 4 KB Random Reads And Writes
- Benchmark Results: PCMark Vantage Storage Test
- What's The Deal With Write-Cache Buffer Flushing?