Prototype SATA 6Gbps SSD gets benchmarked: yessir, it's hasty
We've already seen Seagate tout the speed advantages of its SATA 6Gbps hard drive, but it's the 6Gbps SSD that we're really curious about. PC Perspective managed to snag itself a Marvell prototype drive, and even though they could only test the read capabilities of it, the results are nothing short of titillating. Reviewers pitted the Marvell drive against Intel's well-respected X25-M G2, and their (admittedly limited) testing led them to discover a 33 percent increase in burst performance over one of the quickest SSDs on the market today. In case you're still not impressed, you should know that they also saw a 27 percent uptick in sustained read performance (compared to the X25-M G2) and a 175 percent increase over the aforementioned SATA 6Gbps Seagate HDD. Obviously it's still too early to tell whether the 6Gbps SSD really is the best thing since the vacuum tube, but if these ultra-early results are any indication of what's to come, we suggest you start packing those pennies away right now to finance your next storage upgrade.























Yum.
@Schmitty
Too bad my 5 drive RAID array demolishes its sustained read and write speeds by 200% at 1/10th the cost and 80 times the capacity.
Makes the faster Windows boot due to seek times a moot point.
@Kurian Sustained, yes. Wonderful. What about random?
@Kurian Too bad you're an asshat trying to show off.
@Kurian Douche-nozzle comment.
How about temperature?
How about energy consumption?
How about size?
Go-go-gadget troll. Think server-farm and YOU LOSE every way conceivable.
I find it amusing that 3gbps SATA lasted so long, because of mechanical drive limitations. Now that 6gbps SATA is barely out of the gate, we'll probably have SSDs which saturate it within a year. Two decently fast SSDs in raid0 would do the trick today. Even the pci-e based FusionIO cards hit a bandwidth wall.
What I'm trying to say here is, its about freaking time we ditched the copper and focused on optical solutions. I'm not sure how LightPeak will play out, but if Apple gets an early exclusive on the tech it might actually make their underlying hardware relevant.
@Akhen
But not at a relevant price.
@Akhen
You go copper at some point. The problem is how to get all of those signals on the bus and through the necessary chips.
Also, Apple will at most be a month or two early (like they are with Intel CPUs) if LightsPeak hits, otherwise there'd be no traction among the wider world of computing.
@Schmitty ..badum tish? :p
@Akhen
Uhh.... You realize we can do Hundreds of Gigabytes over copper right? Even 6Gbps Sata doesn't compare to the bandwith in a 16x PCIe 3.0 slot, which is 16GBps or 128Gbps. Copper is fine. Optical makes sense for long distance transmission. But for short interconnects, Copper is more then good enough, for the next decade or two.
@Schmitty Speed is more important than $/GB value when you pass some minimum requirements.
SSDs are the way of the future in local storage!
I'll probably only be able to afford the 4GB version of these drives...shame that won't even hold a Windows installation.
@Evan
LOL, 6 Gbps is the speed. The speed, not the capacity of the drive. :)
@Leo
:facepalm: I know that. I am saying that this will retail for so much money that I would only be able to afford a really, really small version of it.
@Evan
I understood you and totally agree, it's gonna cost more than an arm and a leg.
@chansthename Throw in a kidney and you might come close
Anyone know why I'm showing up as (Unverified)?
@(Unverified) You need to log in and choose a user name. Its since the site got upgraded..
@(Unverified) You have to log out then log back in. Took me a while to figure that out. I was looking all over every relevant page for a button to get verified.
Whats the speed ratio comparison of the theoretical max on SATA 6Gbps vs. DDR3 1600 (PC3 12800) desktop ram. I'm assuming the SSD's are not outrunning the RAM still, right?
@Slick
No, but when you put them on the SATA bus then the limiter is the bus and not the storage medium.
On the PCI-e bus you get better results, but then you're held up by the fact that your DDR is volatile and thus only good as a temporary cache, whereas most SSD solutions are for long term (1min or more) storage.
And if you think 1 minute is short, consider who would use a RAM-based high speed disk.
@Slick
Take a look at this: http://en.wikipedia.org/wiki/List_of_device_bandwidths
It lists bandwidths for all kinds of devices, including SATA and DDR3.
@daniel142005
Having looked that over (That chart just got bookmarked!) I did notice that USB 3.0 has a higher theoretical throughput than the new SATA 3.
Now I've never seen a USB device reach its theoretical max yet, but then I really don't know a great way of checking is my HDD is maximizing the throughput os SATA 2 either. And I know what USB is heavily reliant on bus speeds to get data from point A to B, but just pulling numbers out of thin air, it would almost seem that SATA 3 is almost potentially limiting the speed of SSD's versus them using USB 3.0. Any of that right, or am I just off in left field?
I'm excited for the time when these SSD's get cheaper and larger. My next laptop will have an SSD for sure, just have to wait till the tipping point.
"...and a 175 percent increase over the aforementioned SATA 6Gbps Seagate HDD."
You forgot to mention that it is about 175X more expensive than the Seagate HDD. :)
Isn't there enough talk about SSD speed? We know that they're a lot faster but when is there going to be more talk about capacity??
@Reapered I have an external eSATA/USB 1TB mechanical hard drive. What do I need a big SSD for?
@ethana2 a big SSD to totally replace relatively slow HDD's. Why have a fast boot time just to get to slow executions of programs and personal files?
@Reapered I doubt you have more than 120GB of programs you usually use. You do what you can to speed up your computer, and a SSD in addition to your spinning storage is way better than none. At this point they're not "useless because they're expensive / GB" as some people seem to claim, but they're also nowhere close to replacing HDDs.
Screw capacity, I want price (and not garbage SSD's that are worse than magnetic drives). Gimme $1/GB and I'll start to consider it.
My 2g X-25M is bottlenecked by the SATA-I in my laptop. Ubuntu still has all the data from the SSD it needs to boot in a fraction of one second though, so I'd say the main /experience/ bottleneck of my laptop is nVidia proprietary drivers. 3rd party binary drivers can and will make any operating system suck.
That or x86.. As I like to say, it's high time 1978 died.
@ethana2
Wait... You bought an X-25 M and use it with a SATA 1.5 connection? Point?
The question people should be asking when buying an SSD shouldn't be - "can I replace my HDD with this?" because that's doubtful for most users. The question people should pose is "is this the best upgrade for my computer I can get with $300?"
Obviously everyone has different spending limits and purposes. Gamers, for example, might opt for a more powerful graphics card or bigger monitor for a more immersive experience. Ultraportable users, on the other hand, might desire the durability and boot times. But to say that're not useful yet because $/GB HDDs are far superior is missing the point - it's an additional component at this point in time, not a replacement. Like Turbo for start-up times and most of your programs, not a new engine.
@YpoCaramel: It's not necessarily an additional component. Many people don't need large media storage, so a reasonable-sized SSD would be fine as a replacement drive. That's why $/GB is somewhat important; we don't need terabytes, but somewhere over 100GB and for reasonable money.
Or in other words, yes small SSDs can be great value now as an additional component (40GB or 64GB for a boot drive), but a lot of people are waiting until they can use it as a replacement for a basic main drive.
We've pretty much hit the upper transfer speeds of SATA-II with current SSD's. Kinda upsetting to know they wont get any faster for those of us on SATA-II (aka almost all of us) but there are still the improvements in capacity, reliability, and price to look forward to.
@Nitesh And I think capacity, reliability, and price are more important, since the speed benefits are already large and very obvious.
Give me a sub-$200 160GB SSD and I will be there like shareware.
I don't need a speed race in SSDs. Most are very good already. I would like a price race at a capacity large enough to hold my OS and a few apps/games.