Didn't take long for my web-searches, and I could've discovered the unpleasantness about this by "zooming" photographs of the drive as shown within an ISRT review for my ASUS Z68 motherboard a week or so ago.
THESE ARE SATA-II DRIVES!! It seemed to me that the benchies I'd seen for ISRT had been done with an SATA_III SSD. And I found other indications -- other benchies -- showing that while an SATA-II SSD would show some significant (but not stunning) improvement in sustained reads, the write speed actually degraded the overall rating compared to any SATA-II HDD alone. I think it would especially degrade the rating at least to same degree for an SATA-III HDD.
So waiting while holding my breath, sitting on the edge of my chair, beginning every morning with a search for those Larsen Creek SSD's at places like NewEgg (and I thought I'd searched there before, so AkumaX's contributory link is "news" . . . ) -- it's been a waste. UNLESS -- I misunderstood. UNLESS -- there's an SATA-III Larsen Creek that we're waiting for.
So I apparently made a "good" though not "optimal" decision -- not optimal from the price and capacity standpoint. I used the bottom-end Elm Crest 510 SSD for my ISRT caching.
I get noticeable performance improvements over my 3Ware-AMCC RAID5 with four SATA-II HDD's and a 256 MB controller-cache. Using less than a fourth of the electrical power. Much much faster than even that.
Also by the way. You can only get marginal improvements in ISRT performance by adding a hi-performance HDD like the latest SATA-III VelociRaptor. And it still wouldn't matter much if you cached a three-HDD RAID0 through the SSD. The performance of all such options asymptotically approaches the performance of the SSD configured as a drive operating alone. There might be SOME performance improvement -- noticeable -- by hinging the SSD to those different HDD configurations, but only when you need to use new or different software. And then -- only the first time when it's read into the cache. For the rest of it, a lot of the improvement will come from OS access and OS programs, which would be read into the cache in aspects and ways determined by your "desktop computing" habits with Windows.
What I'd hope for in the near future is this. I'd hope that ISRT allows for even bigger cache sizes, unless there's some inapparent reason why that wouldn't help. You could have a system just jam-packed and chock-full of (legitimately licensed) software if you use a lot of it and different kinds of it. Then, caching the OS and that software from an even bigger HDD -- for instance a TB drive -- would be fantastic. Even with a last-gen Samsung F3.
One thing I'm attempting to avoid. Caching HUGE files that are neither software nor used daily: to whit -- if you capture digital or analog TV to MPG or the Microsoft wrapper file for MPG with Media Center, you want those things written initially to another hard disk for two seemingly conflicting reasons:
1) As per setups with Media Center would recommend, use a separate disk for captures -- to avoid gumming things up when reads and writes occur simultaneously. That would improve performance, even for a slower choice of the second disk.
2) Don't cache your reads of MPG recorded video through the ISRT-SSD. Your speed from any old SATA-II drive will be good enough. I even think it would be a shame and a waste to get a second SSD and cache that second data-drive. You gonna watch the same movie every day? Some HD recordings can be 10GB or larger. You want one such film-file crammed into your SSD-Cache drive?
