Toshiba SSD vs Samsung SSD and inconsistent read/write speeds

Can someone test with TWO same inch size, same processors, same storage of SSDs, but different model of SSD, one from Samsung , one from Toshiba, n see the result!

1 min 30 sec clip video clip is like nothing!!!

I believe this DOES make a difference if you move around a lot of files or big files.

Can someone test with TWO same inch size, same processors, same storage of SSDs, but different model of SSD, one from Samsung , one from Toshiba, n see the result!

1 min 30 sec clip video clip is like nothing!!!

I believe this DOES make a difference if you move around a lot of files or big files.

I dont have the Toshiba one anymore so I cant do anymore testing.

But even a 1 min 30 sec clip with 2 transition and 2 titles added should have had at least a few seconds difference if it mattered?

I am no expert but thats what I would assume.

I dropped by my local Apple Store this evening to do some testing in light of all the conversations taking place. I fired up trusty old Xbench and I've enclosed my results. The test was run 3 times on each machine and both machines were identical in make & model, minus the hard drive manufacturer. I've supplied the aggregrate results, the variances between each of the three tests were minimal (0.01-0.05 MB/s).

You can see by the results that the Samsung is faster in all categories that were benched except for the Uncached Write [256K blocks]. The actual Xbench scoring was 342.99 to 303.48 in favor of the Samsung.

As I'm a firm believer in "real world" experiences I also perform some other tests and I've included those results:

Test #1 - A .dmg file (3.74GB in size) copied from a USB 2.0 HDD to the Desktop

Test #2 - A .dmg file (3.74GB in size) copied from the Desktop to the Documents folder, same user

Test #3 - A series of files (2,553 items, totaling 1.55GB) copied from a USB 2.0 HDD to the Desktop

Test #4 - A series of files (2,553 items, totaling 1.55GB) copied from the Desktop to the Documents folder, same user

Toshiba:

#1 - 1 minute, 43 seconds

#2 - 42 seconds

#3 - 48.8 seconds

#4 - 23.5 seconds

Samsung:

#1 - 1 minute, 43 seconds

#2 - 27.9 seconds

#3 - 48.5 seconds

#4 - 24 seconds

As I tried my best to time the file placement with my stop watch, there's a technical margin of error in the .1 to .5 second range for those results. You can see, however that the performance is very similar. The one area of concern was the Samsung's ability to copy the large .dmg file from one location on the HD to the another. It was noticeably faster, well over 10 seconds.

So if you charge by the seconds and the hour, the Toshiba would be for you. 🙂

Test #2 - A .dmg file (3.74GB in size) copied from the Desktop to the Documents folder, same user

Toshiba:

#2 - 42 seconds

Samsung:

#2 - 27.9 seconds

That is so wierd that this was the only difference in the tests?

RockStarInLife wrote:

That is so wierd that this was the only difference in the tests?

Only difference. I even ran the that test three times to ensure it wasn't a Finder reporting fluke and the results were +/- 0.2 each time (to account for my latency hitting the button). No surprise, honestly, this same thing happened in October of 2010 when the previous generation MacBook Airs came out. The Samsung benchmarked faster than the Toshiba (about the same discrenpencies) and people took to the forums posting about the differences. But yet in "real world" testing the difference between the drives were barely noticeable, if any.

Really not so weird when you think about it...

Cases 1 & 3 are from a USB 2.0 HDD which will be the limiting factor.

Case 4 is 2553 files. I suspect the overhead of creating, opening,initiating the i/o, closing each file hides the actual data transfer time differences.

Case 2 is a large single transfer SSD to SSD and thus shows the difference in hte speed of the chips.

Overall I suspect that the difference will only show markedly in a limited number of real world situations - never the less, I'm glad mine arrived Friday with the Samsung (13" i7 256GB ).

Steve

Is it possible to order these over the phone and try to specify that you want to have the samsung?

Nuke

Hi Nuke,

It's not possible. There's no way to identify which hard drive is in the machine from the outside. The part/model number are identical. Now, if you had the SN# of the machine I'm sure there is a way but the problem is you won't have the information at the time the machine is ordered and the sales rep has no way to see what SN# Apple will assign to you. That information isn't available until the item is sold out of inventory to you.

Once you have that SN# you'd then need access to Apple's internal systems that they manage their product inventory and verify against the shipping manifest.

So yeah... not possible.

Not only in Random Uncached Write [256K blocks] is Toshiba faster, but also in Random Uncached Read [4K blocks]. Doesn't it mean that is faster at launching apps and booting?. That would make it seem that for everyday user it will appear faster at common tasks, and people that use it for large file transfer in the SSD (like video, or photo editing) will notice Samsung faster?

It's totally unfair for some customers!

http://www.anandtech.com/show/4275/apple-now-using-samsung-ssds-in-macbook-air

Really not so weird when you think about it...

Cases 1 & 3 are from a USB 2.0 HDD which will be the limiting factor.

Case 4 is 2553 files. I suspect the overhead of creating, opening,initiating the i/o, closing each file hides the actual data transfer time differences.

Case 2 is a large single transfer SSD to SSD and thus shows the difference in the speed of the chips.

Ok... That makes sense now.

I guess we just need to see more tests where the chips would be the main difference.

What kinds of real workd tests would be good to see?

I orderend my MBA 11'' i7 128SSD online and I just recieved it today, went and checked which SSD I had and to my luck I got the Toshiba one. My only concern is that eventhough I bought it online can I return it to the store for an exchange but I thought these i7s were only built to order for the 11'' so is it still possible they can still swap me another one on the spot?

Karthic,

I guess its just a 50/50 chance on what you get. I will be ordering the exact same spec model on Friday, so maybe ill get lucky. Good luck on returning yours for the right one.

Please consider that both have their strong points, while Samsung is faster at sequential transfers Toshiba is faster at random read/writes, that will make the Toshiba faster at boot up and launching apps and almost every simple common task, it's for large file transfers as in photo or video editing that the Samsung speed will come in handy. Depending on which use you're going to give the MacBook Air maybe the Toshiba or the Samsung will be fine. Also, please remember the difference will be really small in everyday use...sometimes even unnoticeable. Another thing to consider is how reliable they turn out to be, usually Samsung SSDs decay faster, but I'm not sure about this model.

I would like to share you all my exp. I have the i7 MBA with 256 samsung ssd and my gf has the i5 MBA with 256 toshiba ssd. YES the blackmagic disk test showed that my ssd (samsung) has faster reading and writing speed. However when it comes to day to day use this is not the case. I tested the boot up speed on both machines, my one boots up in 21secs and my gf's one boots up in 16 secs. I also tested the shut down speed. Mine shut down in 4 secs and my gf's one shut down in 2 secs. What the f$:) i paid more and i thought i have a better machine but ....