Your best source of advice (first order approximation), is
to read the customer reviews on Newegg or Amazon. If customers
note "mine died after three weeks", then you know there
are potentially firmware problems with the unit in question.
If the one you're interested in, passes the "smell" test,
then, you're off to the review sites, for more info. For
example, a recent Samsung, died in the hands of reviewers,
and the reviewers now claim, only the review units had
a bugged version of firmware. The release ones (the ones
end-users buy) had an OK version. The review sites can
help you a bit in that regard.
And looking at the reviews, you can see how many IOPs
an SSD achieved in testing, and decide whether that
really makes a difference to you. The zero seek time,
is responsible for a lot of the improvement you see
in an SSD. Whether it's 10,000 or 40,000 IOPs, maybe
that is more apparent if all you do is benchmark them
all day long.
If a unit has bad firmware, stutters or slows down
after you've done a 4KB random write test, those are
actually corner conditions that can arise in the real
world. If you're doing something that does a lot of
small writes, and at high speed, that can uncover
garbage collection or TRIM issues. But if this
SSD is being used for more ordinary things, like
reading an email every ten minutes, browsing with
your web browser, you might not notice anything like
that (drive performance stays consistent). The drive
has time to wear-level and clean up behind the scenes.
But if you pound the thing with synthetic benchmarks,
some drives need all night to clean up (or, you run
"secure erase" and give them a chance to clean up
the "fast way").
Every controller has "had some problems". It's the
ability to fix them via firmware, and preferably
before product is in the hands of users, that
makes the difference. And it's true, that some
architectures are better than others, have a
faster control processor or whatever. And that
should show up in the IOP rate, which you can
get from actual, detailed reviews.
*******
The SSD is a storage device. If you clone from
one hard drive, to another hard drive which is smaller,
the same issue arises with respect to size. So there
must be cloning tools which can handle a size difference
in a partition. The SSD doesn't change that.
What does change with an SSD, is the "preferred alignment".
You could use an ignorant cloning tool, one only
designed for hard drives and CHS alignment. That
gives an SSD with a non-preferred alignment (slight
speed penalty). And then, you use a separate
"alignment tool", to move things to one megabyte
(1MB) boundaries. So there's certainly more than
one path to getting the job done.
An SSD could report a non-standard CHS value, one
intended to encourage an older tool, to use a better
alignment. But I don't think that is guaranteed,
and deserves verification by the user.
If you install an SSD in a computer, and install Win7
or Win8 to it, the installer already "prefers" 1MB alignment,
and can handle that detail for you. (Even Linux does
that now.) But when it comes to mixing and matching
OSes and tools, you could end up with either a
CHS (63 sector) or 1MB alignment, and in the 63 sector case,
use an alignment tool to fix it up.
If time is more important than money, you can find
a tool that handles things like that. Or, waste
more time (like I would), looking for a free
solution
I bet there's a way to do it. I
might even boot a Linux Live CD and try GParted,
or use one of the dedicated distributions, and
get it done that way.
http://www.paragon-software.com/home/partition-alignment/index.html
If you put a blank SSD in a computer, and installed
Win7 or Win8 to it, then you wouldn't need a tool
like that.
Maybe the tool to look for, is a reporting tool,
which can document the state the thing is in ?
You might waste less money buying tools that way.
Paul
