OWC Shares Mid-2014 Retina MacBook Pro Unboxing, SSD Tests

[jpm224]

So happy they didn't make 16 gb of RAM standard for the 13 inch. Now I don't feel ripped off for the pro I bought a month ago (2.6-8-512 13 inch model).

The .2 ghz upgrade probably wouldn't have been worth waiting (they were $75 off at best buy when I bought), and now I'm happy knowing there won't be a significant upgrade to the whole line until at least mid-late 2015.

 

[Orlandoech]

So happy they didn't make 16 gb of RAM standard for the 13 inch. Now I don't feel ripped off for the macbook I bought a month ago (2.6-8-512 13 inch model)

Same model I have. Love it.

 

[StintheBeast]

Is it me or does the 13" pciSSD perform much less than a budget Samsung Evo?

It's not you. Those benchmarks are awful. Even the 120gb 840 evo has up to 540MB/s reads and 410MB/s writes. and you can pick up one of those drives for $80. Yet Apple can charge $200 to upgrade to a 256gb from a 128gb. With those specs? No thanks.

 

[Irock619]

My god these computers are getting to be so expensive. I hope my 2011 MBP doesn't crap the bed because I could never afford one of these new computers

 

[QCassidy352]

I don't think the two tests are comparable. From what I understand, Blackmagic measures the read/write speeds of large chunks of data while this QuickBench test is measuring the read/write speeds of smaller chunks of data.

The two tests MUST be different. Black magic shows mid 600s for both read and write on my late 2013 rMBP, which is very much standard for that model. Even accounting for the lower capacity of these, no way there's that difference.

 

[henzpwnapple]

Using a mid-2012 MBPR.. I feel that macbook's life cycle is dropping faster than iphone's relatively..

 

[sprezz]

I guess i will have to wait until 2015 to replace my 17 incher from 2009... Still going strong, although the battery has aged to 2 hours max... Oh well...

 

[Hellhammer]

The analysis is extremely poor. You should not be looking at the average data rate when testing multiple transfer sizes because the difference between 4KB and 1024KB transfer speeds is close to tenfold. OWC's testing is also fairly bad because they only ran the standard suite, which tops out at the IO size of 1024KB and thus cannot take full advantage of the drive.

The SSDs are the same as in the original Haswell rMBP, so to be honest there is nothing to see here. Bad tests and even worse analysis are just confusing everyone now.

 

[Mashurrab]

Still no Flash Storage upgrades from OWC!!!

I was thinking the same thing, plus why would OWC do a tear down and benchmark tests. I thought they will replace the SSDs with their own and show us the difference but that did not happen because they do not have a compatible offering,

 

[544263]

Away from my MacBook now, but I could have sworn my 2012 rMBP had reads around 400MB/s and writes around 500MB/s.

MacBook 2011 + 2012, with optibay replacement (SATA III Adapter) -> Two SSD's in raid 0 -> Almost 1000 MB/sec.

See http://hardwrk.com/macbook-ssd-hdd-adapter-kit.html

 

[Mashurrab]

Why does my 2011 MacBook Air get better SSD performance than a 2014 MacBook "Pro"? That's just not right.

If true then you should be happy#

 

[yjchua95]

It's not you. Those benchmarks are awful. Even the 120gb 840 evo has up to 540MB/s reads and 410MB/s writes. and you can pick up one of those drives for $80. Yet Apple can charge $200 to upgrade to a 256gb from a 128gb. With those specs? No thanks.

These benchmarks are random read and writes, which is normal.

Sequential read and writes will be far higher.

 

[potatis]

Didn't this thing have two fans when first released?

 

[Skika]

Didn't this thing have two fans when first released?

I imagine the one with the dedicated graphics has 2 fans while the integrated has one.

 

[MacGeek1987]

Put a Samsung SSD in my MBP a few months ago that puts my 4 year old computer on par with the new ones. Granted the i5 is a few generations behind and the graphics card blows, but it still performs basic tasks just as well as my brand new, fully optioned out iMac. Maybe I'll upgrade next year.

 

[milo]

Even the 120gb 840 evo has up to 540MB/s reads and 410MB/s writes.

You're comparing the highest numbers ("up to") with average numbers, did you even look at the full chart of numbers in this article? Top read/write numbers for 1028k were 584/566 on the 256 and 593/438 on the 128. All of which are higher than the EVO numbers you listed. Not to mention that 1024 is a low number to top out at, with a benchmark that goes up to bigger chunks you're likely to see even higher results.

Honestly I've never seen an "average" number for SSD benchmark before, it's flat out meaningless unless you directly compare it to other drives running the same benchmark and also generating that same average.

 

[milo]

I didn't realize how much MR dropped the ball on this one until I clicked over to the OWC article. Besides MR listing the average speed instead of the top speed (OWC just posted the screenshot and didn't single out the pointless "average" number), OWC has two more tests for each drive with larger file sizes and much higher numbers.

http://blog.macsales.com/25770-owc-unboxes-tests-ssd-speeds-of-new-13-and-15-retina-macbook-pros

Top numbers were actually 791 read 744 write for the 256G and 762 read 454 write for the 128G. Those are really good numbers, faster than any SATA drive can handle and I suspect comparable the previous generation of these machines if not a bit faster.

Really misleading spin by MR, you guys really should correct your article as you give the opposite impression of what the actual speeds of these drives are.

 

[The Mercurian]

My god these computers are getting to be so expensive. I hope my 2011 MBP doesn't crap the bed because I could never afford one of these new computers

But.......they just got cheaper by a couple of hundred !?! I think you are confused.

 

[QCassidy352]

The analysis is extremely poor. You should not be looking at the average data rate when testing multiple transfer sizes because the difference between 4KB and 1024KB transfer speeds is close to tenfold. OWC's testing is also fairly bad because they only ran the standard suite, which tops out at the IO size of 1024KB and thus cannot take full advantage of the drive.

The SSDs are the same as in the original Haswell rMBP, so to be honest there is nothing to see here. Bad tests and even worse analysis are just confusing everyone now.

Thank you. That makes much more sense now. Shame on MR for the very misleading spin.

 

[Weaselboy]

As another data point, here is my BlackMagic test of the new version 128GB 13" Retina I picked up yesterday. This is with Filevault on.

 

[jclo]

I didn't realize how much MR dropped the ball on this one until I clicked over to the OWC article. Besides MR listing the average speed instead of the top speed (OWC just posted the screenshot and didn't single out the pointless "average" number), OWC has two more tests for each drive with larger file sizes and much higher numbers.

http://blog.macsales.com/25770-owc-unboxes-tests-ssd-speeds-of-new-13-and-15-retina-macbook-pros

Top numbers were actually 791 read 744 write for the 256G and 762 read 454 write for the 128G. Those are really good numbers, faster than any SATA drive can handle and I suspect comparable the previous generation of these machines if not a bit faster.

Really misleading spin by MR, you guys really should correct your article as you give the opposite impression of what the actual speeds of these drives are.

The analysis is extremely poor. You should not be looking at the average data rate when testing multiple transfer sizes because the difference between 4KB and 1024KB transfer speeds is close to tenfold. OWC's testing is also fairly bad because they only ran the standard suite, which tops out at the IO size of 1024KB and thus cannot take full advantage of the drive.

The SSDs are the same as in the original Haswell rMBP, so to be honest there is nothing to see here. Bad tests and even worse analysis are just confusing everyone now.

FYI, OWC updated its original post several hours after it posted the first two tests, which were the only available at the time. I've updated the post based on your feedback and the new tests.

 

[koban4max]

So happy they didn't make 16 gb of RAM standard for the 13 inch. Now I don't feel ripped off for the pro I bought a month ago (2.6-8-512 13 inch model).

The .2 ghz upgrade probably wouldn't have been worth waiting (they were $75 off at best buy when I bought), and now I'm happy knowing there won't be a significant upgrade to the whole line until at least mid-late 2015.

you sure have a negative attitude.

 

[Dayadan]

As another data point, here is my BlackMagic test of the new version 128GB 13" Retina I picked up yesterday. This is with Filevault on.

So I'm guessing that is the new Mid 2014 Retina with a 2.6ghz processor right?
If that's the cause I wonder how much of a performance hit filevault takes.

This is a BlackMagic Test of the late 2013 13" Retina I picked up a few weeks ago.

 

[Weaselboy]

So I'm guessing that is the new Mid 2014 Retina with a 2.6ghz processor right?
If that's the cause I wonder how much of a performance hit filevault takes.

This is a BlackMagic Test of the late 2013 13" Retina I picked up a few weeks ago.

Yes... mine is the new 2014 model that came out yesterday. FV makes some difference, but very little in my experience and tests I have read.

I have the 128GB drive which is quite a but slower than the 256/512 drive like I suspect you have?

 

[Dayadan]

Yes... mine is the new 2014 model that came out yesterday. FV makes some difference, but very little in my experience and tests I have read.

I have the 128GB drive which is quite a but slower than the 256/512 drive like I suspect you have?

Yeah I have the 256gb drive, now we should wait for someone with a new 2014 w/ a 256 ssd to compare with.

Also I wonder if the new 2014 13" retinas only come with Sandisk SSD's or if they can come with samsung drives.