Disk Speed Bench, Macbook Pro mid 2017 15 '' TB 256 SSD

Discussion in 'MacBook Pro' started by jsangil, Jul 1, 2017.

  1. jsangil macrumors newbie


    Jul 1, 2017
    Hi. First of all, I am new in this forum, so...greetings from Spain. Of course, I apologize for my bad English.

    I'm doing disk speed tests on my Macbook Pro mid 2017 15 '' TB with 256 SSD (KL Processor) and I'm a bit disappointed. I am looking for possible becnhmaks with this same disk capacity but I can not find it.

    Anyone with this same model and disk capacity, could tell me the speed of your disk?

    Thanks and best regards.

    Attached Files:

  2. keysofanxiety macrumors 604


    Nov 23, 2011
    Speeds are fine. For quicker write speeds you'll need a larger SSD as the Flash chips work in parallel.

    Read speeds are more important. At 2.2GB/s that's insanely fast.
  3. Waski macrumors newbie

    Jun 22, 2013
    I have exactly the same results and the answer for this can be in this video:
    This guy also got lower results on blackmagic hard drive test and then use a different benchmark to get correct results.

    According to Apple 2017 macbook pro should achieve something around 3100 Read and 2200 Write.

    Unfortunately I don't have this second benchmark test to check this out, so could anyone confirm this?
  4. HaddockW macrumors member

    Aug 4, 2017
    San Francisco
    I think you need larger capacitu. My 2016 Loving.. It.

    Attached Files:

  5. New_Mac_Smell macrumors 68000


    Oct 17, 2016
    It's actually 3200R/2200W, but if you look, according to Apple it is "Up to 3200 Read / 2200 Write*", clearly stating that these were achieved under lab conditions for a top end 15" with 2TB drive (Always read in full).

    Don't be getting a base model and expecting it to perform as equally well as the top end, as others have pointed out it's to do with SSDs, the bigger the faster really. The SSD speeds on the whole are lightning fast however, you're still 10x faster than consumer grade ones.
  6. ZapNZs macrumors 68020


    Jan 23, 2017
    Don't be sad! Be glad!

    The only thing we can say for certain is that your SSD is really fast, and that the write speeds you are getting in the (specific) Blackmagic App for an SSD of your size is NOT unusual.

    Note that my knowledge here is limited - someone smarter than me could correct me if I make a mistake or further elaborate if they care to.

    Different benchmarks of the same SSD can show very different results (this example shows the difference between the benchmarks of the 256 GB MBP nTB using Blackmagic vs. QuickBench.) Simple things like the size of the SSD (larger sizes = higher write speeds, usually), to the size of the transfer files, to the amount of free space on the SSD, to the programs running in the background, to the (recovery) pauses in test intervals, to whether or not indexing is enabled or not (as the OS may try to index the test file), to the compressibility of the data can have a significant impact on outcome - then there's the technical variables such as the workload's IO intensity and the testing's queue depth, to which my own understanding is limited-at-best.

    To my understanding, Blackmagic uses incompressible data. I do not believe it tests across multiple transfer sizes - even though this can hugely affect measured performance:

    960 Pro

    This carries significant real-world implications given how much one's personal workload could be oriented in one direction (while at the same time all Users [and their Apps/OS] use heterogeneous file sizes, to some extent.) To me, this limits most of the generalizability of the test, and makes it more of a fun novelty than anything else.

    You could try running your own benchmarks based on your own usage. For example, I have my own benchmark folder, which contains four different types of file sizes totaling about 30 GB of many very small Excel/Word/Visio documents, many small AACs and JPEGs, some medium-sized datasets/visualization outputs, and several large video files and datasets. Essentially, these mimic some of the files I work with daily, and I can time how long it takes the very small/small/medium/large file sets individually, or as a whole, and compare the results of one SSD to another to determine how this affects my real-world performance. The transfer speeds of the many small files is dramatically slower than the large files (think like 1/100th the performance!)
  7. stylinexpat macrumors 6502a


    Mar 6, 2009
    Over 2000 and over 3000 sound like good numbers to me.
  8. jerryk macrumors 68040

    Nov 3, 2011
    SF Bay Area
    You are replying to a thread from more than 7 months ago.
  9. stylinexpat macrumors 6502a


    Mar 6, 2009
    Shopping for a laptop now is why..

Share This Page