I went to the Engadget article in question to see what you mean, still not seeing your objection:
"The 128GB Samsung SSD in his 11-inch Air was able to achieve 246 MB/s write and 264 MB/s read speeds. When he switched to the 13-inch model, however, speeds dropped to 156 MB/s and 208 MB/s, respectively, using that notebook's 128GB Toshiba SSD"
Same size, very different speeds.
Quote:
During our tests, the 256GB Samsung drive in our older [MacBook Air] model achieved 214 MB/s write and 251 MB/s read speeds, while the 128GB Toshiba drive in the new MacBook Air scored 184 MB/s and 203 MB/s during write and read tests, respectively,? the publication said.
This might have something to do with the Apple lawsuit with Samsung so they needed a second supplier Toshiba as a stop gap measure?
I bought the new Mac Mini i7 dual core w/ a 256 GB SSD that should arrive this Wednesday. If I get the slower Toshiba SSD drive then I am just going to upgrade it with a fast one from OWC.
If I were buying a new MBA I would make sure i received one with a Samsung drive. That performance difference is too much for me to ignore. If that meant opening the MBA and returning it immediately so be it.
Quote:
Originally Posted by MacSuperiority
Its right there in plain sight.
Speaking of being "right there in plain sight"...
What you're quoting is the additional testing to see how the new MBAs compared to the old MBAs they've recently deprecated. It's not the focus of the article, which is the Samsung and Toshiba SSD cards in the new MBAs have a very wide performance margin for the same capacity drive.
If I were buying a new MBA I would make sure i received one with a Samsung drive. That performance difference is too much for me to ignore. If that meant opening the MBA and returning it immediately so be it.
I ordered my MBA and hope to have it by the end of this week. That's the first thing I'm going to check when I receive it. I'm not going to sell my late-2010 MBA until I know the new one fits the ticket!
It was always luck of the draw with hard drives, whether you got a noisy one or a quiet one. The more things change, the more they stay the same.
That's generally been the case with Apple, and most every manufacturer of computers. For the most part hard drives were fairly similar in performance across the lines, and few people noticed the speed. Here, synthetic testing makes it quite obvious.
The only exception to the rule is with Time Capsule. In those, you always get "server grade" hard drives .
There are faster SSDs in the MBA. There are also slower SSDs in the MBA.
That exactly the concern, though it's still an open question as to how much it impacts user time. It would be more interesting to know the speed from cold boot, and maybe open a handful of stock apps on boot.
That exactly the concern, though it's still an open question as to how much it impacts user time. It would be more interesting to know the speed from cold boot, and maybe open a handful of stock apps on boot.
1) As a user, I wouldn't settle for the Toshiba SSD because, as a user, if I were buying a the machine secondhand I'd specifically look for one with the Samsung SSD and, as a seller, I'd specifically note that it has the faster, Samsung SSD. I don't expect most buyers to care, though those that read and post on this forum would certainly skew those figures if we were to do a poll.
2) Even though those are synthetic tests they are much more accurate to overall performance than the synthetic tests of CPU and GPU performance since it's a much simpler reading and writing to disk. I may be wrong, but experience tells me that the Samsung drives are considerably faster in the real world. That said, seeing the real world impact would be nice.
Yes, but I don't think that was ascii's point. Just saying that you get a luck of the draw in some form or another with drives with platter drives or SSDs. Now, it looks like performance might be a luck of the draw, because you don't know what brand and model drive you'll get, mainly because the spec you order is just size and drive type.
I couldn't bear reading people's attempt to explain the speed differences without necessary SSD knowledge. I had to create an account just to post my 2 cents on the topic.
This is what I know about SSDs:
1) Larger capacity SSDs typically have shown faster read/write speeds. This is because larger SSDs use larger GB per chip in its constructions. This gap used to be something like 20-50% difference per storage size level (64GB vs 128GB vs 256GB). Anything larger than 256GB used to double stack smaller RAM chipsets to achieve size, but have similar performance as the 128GB or 256GB counterparts. Newer generation SSDs have shown a smaller performance gap between capacity sizes.
2) SSD speed is greatly influenced by the SSD controller and chipset construction. Samsung has 22nm chips and are more efficient and faster than the industry's normal 32nm construction --Toshiba?). However, SSD chipset controllers can make a significant difference for the same chipset construction---a difference large enough to explain 280MB/s vs. 180MB/s (see OCZ's Vertex 2's 280MB/s vs. Vertex 3's 550MB/s using the SanForce SF2200!). Variation in performance among manufacturers are evident in the SSD world.
3) Use ATTO or HD Tune Pro (both are Windows programs, but it's the standard benchmark tools for PCs to test SSD speeds) to see if the speed difference is a random issue or a fundamental one.
4) I own a Vertex 3 SSD, had had SSDs with 60MB/s, 150MB/s, 180MB/s, 285MB/s, 550MB/s and 700MB/s (Revodrive x2) experience and the OS experience is marginal after 200MB/s speeds. The sweet spot for the industry for price to speed right now is the 128GB doing mid to upper 2XXMB/s. Anything above this is like buying a Ferrari instead of a normal sports car (BMW M3, let's say).
I would not be happy to know that the speed difference is a 50% difference. I would be interested to see how fast the difference is between the 64GB and 128GB SSDs on the MBA 11" MBA as well as the 128GB vs. 256GB on the MBA 13".
edit: Don't use these benchmarks to determine that Samsung is faster than Toshiba SSD cards as Toshiba SSD cards are all from the 2010 MBAs while the Samung is from the 2011 MBA. Use it only to see how the newer MBAs are significantly better than the 2010 MBAs, at least with the Samsung SSD card.
Comments
I went to the Engadget article in question to see what you mean, still not seeing your objection:
"The 128GB Samsung SSD in his 11-inch Air was able to achieve 246 MB/s write and 264 MB/s read speeds. When he switched to the 13-inch model, however, speeds dropped to 156 MB/s and 208 MB/s, respectively, using that notebook's 128GB Toshiba SSD"
Same size, very different speeds.
During our tests, the 256GB Samsung drive in our older [MacBook Air] model achieved 214 MB/s write and 251 MB/s read speeds, while the 128GB Toshiba drive in the new MacBook Air scored 184 MB/s and 203 MB/s during write and read tests, respectively,? the publication said.
Its right there in plain sight.
I bought the new Mac Mini i7 dual core w/ a 256 GB SSD that should arrive this Wednesday. If I get the slower Toshiba SSD drive then I am just going to upgrade it with a fast one from OWC.
http://blog.macsales.com/11248-2011-...th-owc-6g-ssds
Marcus
Its right there in plain sight.
Speaking of being "right there in plain sight"...What you're quoting is the additional testing to see how the new MBAs compared to the old MBAs they've recently deprecated. It's not the focus of the article, which is the Samsung and Toshiba SSD cards in the new MBAs have a very wide performance margin for the same capacity drive.
If I were buying a new MBA I would make sure i received one with a Samsung drive. That performance difference is too much for me to ignore. If that meant opening the MBA and returning it immediately so be it.
I ordered my MBA and hope to have it by the end of this week. That's the first thing I'm going to check when I receive it. I'm not going to sell my late-2010 MBA until I know the new one fits the ticket!
I ordered my MBA and hope to have it by the end of this week. That's the first thing I'm going to check when I receive it.
Likewise, me.
For those that can't read the Samsung SSD outperforms the Toshiba
It was always luck of the draw with hard drives, whether you got a noisy one or a quiet one. The more things change, the more they stay the same.
That's generally been the case with Apple, and most every manufacturer of computers. For the most part hard drives were fairly similar in performance across the lines, and few people noticed the speed. Here, synthetic testing makes it quite obvious.
The only exception to the rule is with Time Capsule. In those, you always get "server grade" hard drives
Phil
Well, I guess that kills that rumor of faster SSD's in the new Air.
There are faster SSDs in the MBA. There are also slower SSDs in the MBA.
There are faster SSDs in the MBA. There are also slower SSDs in the MBA.
That exactly the concern, though it's still an open question as to how much it impacts user time. It would be more interesting to know the speed from cold boot, and maybe open a handful of stock apps on boot.
That exactly the concern, though it's still an open question as to how much it impacts user time. It would be more interesting to know the speed from cold boot, and maybe open a handful of stock apps on boot.
1) As a user, I wouldn't settle for the Toshiba SSD because, as a user, if I were buying a the machine secondhand I'd specifically look for one with the Samsung SSD and, as a seller, I'd specifically note that it has the faster, Samsung SSD. I don't expect most buyers to care, though those that read and post on this forum would certainly skew those figures if we were to do a poll.
2) Even though those are synthetic tests they are much more accurate to overall performance than the synthetic tests of CPU and GPU performance since it's a much simpler reading and writing to disk. I may be wrong, but experience tells me that the Samsung drives are considerably faster in the real world. That said, seeing the real world impact would be nice.
It was always luck of the draw with hard drives, whether you got a noisy one or a quiet one. The more things change, the more they stay the same.
If you buy an SSD, it's ALWAYS quiet, guaranteed.
If you buy an SSD, it's ALWAYS quiet, guaranteed.
Yes, but I don't think that was ascii's point. Just saying that you get a luck of the draw in some form or another with drives with platter drives or SSDs. Now, it looks like performance might be a luck of the draw, because you don't know what brand and model drive you'll get, mainly because the spec you order is just size and drive type.
I couldn't bear reading people's attempt to explain the speed differences without necessary SSD knowledge. I had to create an account just to post my 2 cents on the topic.
This is what I know about SSDs:
1) Larger capacity SSDs typically have shown faster read/write speeds. This is because larger SSDs use larger GB per chip in its constructions. This gap used to be something like 20-50% difference per storage size level (64GB vs 128GB vs 256GB). Anything larger than 256GB used to double stack smaller RAM chipsets to achieve size, but have similar performance as the 128GB or 256GB counterparts. Newer generation SSDs have shown a smaller performance gap between capacity sizes.
2) SSD speed is greatly influenced by the SSD controller and chipset construction. Samsung has 22nm chips and are more efficient and faster than the industry's normal 32nm construction --Toshiba?). However, SSD chipset controllers can make a significant difference for the same chipset construction---a difference large enough to explain 280MB/s vs. 180MB/s (see OCZ's Vertex 2's 280MB/s vs. Vertex 3's 550MB/s using the SanForce SF2200!). Variation in performance among manufacturers are evident in the SSD world.
3) Use ATTO or HD Tune Pro (both are Windows programs, but it's the standard benchmark tools for PCs to test SSD speeds) to see if the speed difference is a random issue or a fundamental one.
4) I own a Vertex 3 SSD, had had SSDs with 60MB/s, 150MB/s, 180MB/s, 285MB/s, 550MB/s and 700MB/s (Revodrive x2) experience and the OS experience is marginal after 200MB/s speeds. The sweet spot for the industry for price to speed right now is the 128GB doing mid to upper 2XXMB/s. Anything above this is like buying a Ferrari instead of a normal sports car (BMW M3, let's say).
I would not be happy to know that the speed difference is a 50% difference. I would be interested to see how fast the difference is between the 64GB and 128GB SSDs on the MBA 11" MBA as well as the 128GB vs. 256GB on the MBA 13".
There are faster SSDs in the MBA. There are also slower SSDs in the MBA.
Heh, true, I meant the rumored 400MB/s drive though.