@scotth - This issue is within the data flow. The SATA II interface within the system (Host) can only accept the data so fast from the drive (Slave), A slow platter drive which offers a SATA III interface will rarely push the data flow to the upper limits of 6.0 Gb/s and the buffers have a chance to finish before the next stream of data would come.
With SSD's there is no break in the stream like a fragmented HDD would have. Thus, the SATA interface buffers overflow and the data blocks CRC fails so then that data block then needs to be retransmitted.
While on the surface this doesn't look like such a big deal having a few blocks needing to be resent. The truth is often times in 1000 of blocks! And then some of these blocks then need to be re-sent a third, forth and onward which can take a bit of time. If the person causes the system to exit out of the transaction the file/s then become corrupt.
In addition, the system is working harder than it needs to so both the system and drives thermals go up and the in the case of a laptop the systems battery runs down sooner.
The way the standard was written years ago (BTW: I was at that standards meeting here in Boston) the expectation was the more costly drives would be transplanted into the new system as the systems where cheaper! Who knew the storage sizes of drives would grow from 5 GB, 10 GB then onto 500 GB, 1TB, 2TB and now the massive 16 TB! And the price per GB of HDD’s would drop! from around $5 a GB or so in the 80’s, to around 5 cents a GB today! Even SSD’s are getting cheaper year to year.
So the standard was written to make sure older SATA I (1.5Gb/s) HDD’s would work in newer spec’ed systems SATA II (3.0 Gb/s) and later either of these drives would work in a SATA III (6.0 Gb/s) system (drives are upwardly transferable). The standard didn’t address newer spec’ed drives being used in older SATA 2.0 spec’d systems! As that was not expected. There was an addendum which was never ratified for the 2.0 standard which spelled things out. It later was put in the 3.0 standard, it is in section 13.2.1.16 - Word 76. Thats why there is a problem! As the older host interface chips didn’t know how to interface correctly to the newer drives. Here’s a bit more on the signaling aspect Design of an Open-Source SATA Core and here’s the SATA standard for reference Serial ATA Revision 3.0
So when people with older SATA II systems started installing newer SATA III drives they often thought the drive was working correctly as the data flow was not exceeding the 3.0 Gb/s ability of the system, to handle. To add to this the quality of the SATA cable also needs to be considered! The older SATA I & II cables where not shielded! Unlike the SATA III cables so you often encountered cross-talk across the pair sets. This in turn often created both a FEXT and NEXT problems as the standard allows for bi-directional communication concurrently.
7 Comments
Hi Dan,
Thanks for letting us know! I'll pass this along.
by Kristen Gismondi
@kristen - I would still offer the kit alone or look at getting a different makers drive that is auto sense. There are a few cheaper ones besides Samsung!
by Dan
Hi Dan, what sort of performance or reliability issues would these Crucial SSDs in older computer cause? We looked into this a year ago but it fell off our radar. We have had no negative feedback from customers since then and the product is still highly reviewed.
by Scott Head
@scotth The problem is they aren't listed to be SATA II compatible, so it can cause a reliability issue. Sure it may work unofficially but that doesn't mean it's going to be reliable - it could be there in firmware but not rated due to an issue not worth chasing down for example.
by Nick
Sure, I get that, but we haven't heard of any reliability issues. I agree it's not good to market something as compatible when the manufacturer says it's not. I'm just trying to understand what the issue is more specifically since we haven't had any issues.
by Scott Head
Show 2 more comments