Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

yjchua95

macrumors 604
Original poster
Apr 23, 2011
6,725
233
GVA, KUL, MEL (current), ZQN
Hi chaps,

Out of pure curiosity, I'd love to know the performance of varying drive capacities in PCIe-equipped Macs.

Since the rMBPs have all sizes and capacities (the iMac lacks the 128GB variant and Air lacks the 1TB variant), I opted to post this thread here.

So here's basically the results that I'd love to know for these drives:
SD0128F (SanDisk 128GB)
SM0128F (Samsung 128GB)
SD0256F (SanDisk 256GB)
SD0512F (I always thought that Samsung was the sole supplier of the 512GB, but it turns out that some owners have the SD version of the 512GB, so I'm interested in this drive)

So far, I've these results from my own Macs, posted in read/write speeds (all MB/s)

SM0256F: 720/670
SM0512F: 750/720
SM1024F: 950/1035

I know there's going to be a handful who're going to laugh me off as a lunatic, but hey, we all have quirks of our own, don't we? :D
 
SD0128F (PCI)

Write: 310.5
Read: 667.1

That’s using the blackmagic speed test app. If I keep the app running the speeds are different every time basically. Almost gotten to 700 on reads, for example.

----------

I'm a little curious about the SM0128F... maybe someone will post. :)
 
SD0256F, PCIE

720/570 read/write (highest reading, Blackmagic)

This is measured from a running system, so there is always some activity slowing it down. But I don't think this SSD goes any faster.
 
SD0256F, PCIE

720/570 read/write (highest reading, Blackmagic)

This is measured from a running system, so there is always some activity slowing it down. But I don't think this SSD goes any faster.

SD256
575 Write
730 Read

Barney

Looks like the general trend is that SanDisk makes inferior SSDs compared to Samsung :p

I've got a couple of friends with late-2013 rMBPs (256GB SSDs), only that one of them had Samsung and the other had SanDisk. The difference is slightly noticeable during boot up (the spinning circle doesn't appear at all in the rMBP with SM0256F, but the spinning circle spun twice on the rMBP with SD0256F before entering the login screen). Although the delay was perhaps about 2-3 seconds, it was enough to irk the guy with the SD0256F.

Meanwhile, in tasks with heavy writes, the SM performed noticeably faster than the SD.

This comparison was pretty fair because both had a configuration of 2.4/16/256.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.