I think you may have it mixed up. He is copying from an SSD to the TB drive (which has 2 HDs in RAID 0), and not to SSDs within the TB drive. The specs he is getting are in line with what LaCie reports on their website.
I checked back, and you are right that it is the SSD. That is pretty disappointing given the cost of the box. Could it be something with the controller? Certainly Thunderbolt ought to be capable of letting the SSDs operate at full speed.
It could be the controller. Some SATA controllers are slower than others but it doesn't make much sense that a two-port controller would be limited to 250MB/s, considering that this isn't the cheapest device either. Other option would be a firmware limitation in case LaCie wants to prevent user upgrades to sell the the TBA SSD model at a premium.
I hope these aren't the final results, would be fairly disappointing. Lets cross our fingers and wait for Anand's review ;-)
How do these numbers compare with current USB3 SSDs?
Yeah, yeah USB3 [sux compared to TB | is the greatest interface in the history of the universe]. I don't care about the theoretical rate of what advanced USB3 chips someday might deliver --- my question has to do with what do commonplace USB3 chips deliver TODAY --- to HDs and to SSDs.
What a day, I open my laptop and there is a review on Anandtech comparing SSDs to HDDs based on sequential read and write speeds! >>>>
What is surprising is that the SSDs in RAID 0 deliver so poor performance. We are looking at only 20-30MB/s increase (40-50MB/s with RAM disk) in performance over two 7200rpm HDs in RAID 0.
>>>>>
Just like a review from NewEgg. Wow. Anand spending years and years educating the population by inserting a separate page for each SSD review just for saying "the four corners of SSD performance are..." and now, here it is, comparing a RAID SSD based on sequential speeds.
If nothing else, at least run a Light test 2011 and a couple of 4K benchmarks.
First of all, this is NOT a review. If you actually read the whole article, you should have noticed that all the tests were run by a forum member, we didn't have any way of affecting the tests he ran. Hence ALL we got is sequential reads and writes, and that's all we can analyze.
That doesn't change the fact that the speeds are lower than you should expect. Besides, this is an external HD, so sequential speeds matter more than with internal HDs (more likely to be used as storage of media, not as boot drive).
There's 10 micro-controllers on that PCB! Why on earth is there that much silicon in this thing? My quad interface drive only uses 3... I mean aside from a SATA controller, Thunderbolt controller, and a fan controller, what else does this thing need? It's only using software RAID for crying out loud. Too bad that pic is too blurry to read what any of them are.
Also, just to guess, it looks like the drives may be hooked up to a single SATA port via a port multiplier system rather than to 2 dedicated SATA ports.
Hello everyone, i made the same type of tests on Lacie Little Big Disk Thunderbolt but with Sandforce SATA3 SSDs (Corsair Force GT 120GBytes). The results are far better !
The AHCI Controller of the disk is showing a 3Gbps negociated link and a maximum link of 6Gbps. It seems the controller is limiting the bandwidth for some reason ...
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
8 Comments
Back to Article
KPOM - Monday, September 26, 2011 - link
I think you may have it mixed up. He is copying from an SSD to the TB drive (which has 2 HDs in RAID 0), and not to SSDs within the TB drive. The specs he is getting are in line with what LaCie reports on their website.KPOM - Monday, September 26, 2011 - link
I checked back, and you are right that it is the SSD. That is pretty disappointing given the cost of the box. Could it be something with the controller? Certainly Thunderbolt ought to be capable of letting the SSDs operate at full speed.Kristian Vättö - Monday, September 26, 2011 - link
It could be the controller. Some SATA controllers are slower than others but it doesn't make much sense that a two-port controller would be limited to 250MB/s, considering that this isn't the cheapest device either. Other option would be a firmware limitation in case LaCie wants to prevent user upgrades to sell the the TBA SSD model at a premium.I hope these aren't the final results, would be fairly disappointing. Lets cross our fingers and wait for Anand's review ;-)
name99 - Monday, September 26, 2011 - link
How do these numbers compare with current USB3 SSDs?Yeah, yeah USB3 [sux compared to TB | is the greatest interface in the history of the universe]. I don't care about the theoretical rate of what advanced USB3 chips someday might deliver --- my question has to do with what do commonplace USB3 chips deliver TODAY --- to HDs and to SSDs.
zsero - Monday, September 26, 2011 - link
What a day, I open my laptop and there is a review on Anandtech comparing SSDs to HDDs based on sequential read and write speeds!>>>>
What is surprising is that the SSDs in RAID 0 deliver so poor performance. We are looking at only 20-30MB/s increase (40-50MB/s with RAM disk) in performance over two 7200rpm HDs in RAID 0.
>>>>>
Just like a review from NewEgg. Wow. Anand spending years and years educating the population by inserting a separate page for each SSD review just for saying "the four corners of SSD performance are..." and now, here it is, comparing a RAID SSD based on sequential speeds.
If nothing else, at least run a Light test 2011 and a couple of 4K benchmarks.
Kristian Vättö - Tuesday, September 27, 2011 - link
First of all, this is NOT a review. If you actually read the whole article, you should have noticed that all the tests were run by a forum member, we didn't have any way of affecting the tests he ran. Hence ALL we got is sequential reads and writes, and that's all we can analyze.That doesn't change the fact that the speeds are lower than you should expect. Besides, this is an external HD, so sequential speeds matter more than with internal HDs (more likely to be used as storage of media, not as boot drive).
repoman27 - Tuesday, September 27, 2011 - link
There's 10 micro-controllers on that PCB! Why on earth is there that much silicon in this thing? My quad interface drive only uses 3... I mean aside from a SATA controller, Thunderbolt controller, and a fan controller, what else does this thing need? It's only using software RAID for crying out loud. Too bad that pic is too blurry to read what any of them are.Also, just to guess, it looks like the drives may be hooked up to a single SATA port via a port multiplier system rather than to 2 dedicated SATA ports.
jacktronics - Thursday, September 29, 2011 - link
Hello everyone, i made the same type of tests on Lacie Little Big Disk Thunderbolt but with Sandforce SATA3 SSDs (Corsair Force GT 120GBytes). The results are far better !http://img189.imageshack.us/img189/2212/diskspeedt...
http://img41.imageshack.us/img41/9081/dsc00918zu.j...
The RAID configuration is 128Kbytes.
Read : 480MBytes/s
Write : 356MBytes/s
The AHCI Controller of the disk is showing a 3Gbps negociated link and a maximum link of 6Gbps. It seems the controller is limiting the bandwidth for some reason ...