Slashdot Mirror


Samsung SSD On a Tiny M.2 Stick Is Capable of Read Speeds Over 2GB/sec

MojoKid writes: Samsung has just announced its new SM951-NVMe SSD, the industry's first NVMe SSD to employ an M.2 form-factor. Samsung says the new gumstick style drive is capable of sequential read and write speeds of 2,260 MB/sec and 1,600 MB/sec respectively. Comparable SATA-based M.2 SSDs typically can only push read/write speeds of 540 MB/sec and 500 MB/sec, while most standard PCIe versions muster just north of 1GB/sec. The Samsung SM951-NVMe's performance is actually very comparable to the Intel SSD 750 Series PCIe x4 card but should help kick notebook performance up a notch in this common platform configuration.

72 comments

  1. Faster than DDR by Anonymous Coward · · Score: 0

    This is faster by most measures than the top speed of the original DDR RAM standard.

    1. Re:Faster than DDR by gweihir · · Score: 4, Informative

      Actually, it is not in any sane measurement. This is linear-speed only. For random access, even the original DDR RAM will trash even these disks by a large margin.

      --
      Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
    2. Re:Faster than DDR by Anonymous Coward · · Score: 1

      For random access, these still transfer 1.2 GB/s, and some of the competitors will push that to 1.6 GB/s. While DDR might have originally about 33% faster than that, I wouldn't call that "trashing", especially if in the bigger picture other disks are on par for random access.

    3. Re:Faster than DDR by Anonymous Coward · · Score: 1

      And yet, in terms of latency, DDR is still orders of magnitude faster.

    4. Re:Faster than DDR by gweihir · · Score: 1

      Seriously, find out the facts before posting BS. DDR has single word access times (and that is what counts) around 100ns. These drives have around 10...100us latency. A factor of 100...1000 is nowhere near comparable. What you are apparently unaware of is that DDR does not use any command queuing and has really small blocks, while the SSD uses, for memory, very large blocks and it only gets the 1.2GB/s because it knows about a whole set of future requests before it has to deliver them. That does not work for RAM.

      If you limit the SSD down to RAM-sized blocks and force it to deliver immediately, it gets completely trashed by DDR RAM, and that is the only sensible benchmark for random accesses to RAM.

      Some people really have no clue and only can compare numbers. Pathetic.

      --
      Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
  2. Has anyone waited 60 days? by PRMan · · Score: 0

    Samsung's benchmarks apparently can't be trusted until you wait 60 days. 840 EVO performance tanks on data written a long time ago. And Samsung has been very slow to respond to these complaints. What's to say this one won't do the same? I would choose the Intel model to be safe.

    --
    Peter predicted that you would "deliberately forget" creation 2000 years ago...
    1. Re:Has anyone waited 60 days? by Anonymous Coward · · Score: 1

      I've got an 840 EVO. There was a firmware bug that caused the performance to get really wonky after some time. I only realized it after running some HD diagnostics and getting really strange numbers that jumped all over the place. Things like 512kb chunks reading at 11mb/s and 1024kb chunks reading 520mb/s, while normally the speeds were fairly comparable.

      Anyways, you need to get the tool from Samsung to correct this. First it has to update the drive's firmware and then it has to rewrite all the data on the drive (don't worry, you shouldn't lose any data in this process) to get around the write leveling bug. Unfortunately, I don't think the tool is available for Linux still.

    2. Re:Has anyone waited 60 days? by Anonymous Coward · · Score: 0

      > Unfortunately, I don't think the tool is available for Linux still.

      Then the drive itself is completely useless to me. This one too I suppose.

    3. Re: Has anyone waited 60 days? by Dumass · · Score: 1

      Turns out the initial fix didn't solve the issue completely. There's another fix coming soon: http://www.anandtech.com/show/...

    4. Re:Has anyone waited 60 days? by ledow · · Score: 1

      Don't Intel models just fail without warning and die on you?

      Your turn.

      It's fun this "let's pick a random complaint against a manufacturer based on one affected product over many years out of dozens that work just fine".

    5. Re:Has anyone waited 60 days? by Anonymous Coward · · Score: 0

      Samsung has since released a flawless 850 Evo, so your argument is invalid until such a time that bugs crop up in newer, fixed, hardware.

    6. Re:Has anyone waited 60 days? by Dumass · · Score: 2

      The (original) fix is available as an ISO and USB stick image for non-Windows machines. I used it on a Macbook Pro last month. http://www.samsung.com/global/...

    7. Re: Has anyone waited 60 days? by Anonymous Coward · · Score: 0

      The initial fix didn't solve anything, it just rewrote all the data on the drive so it would be "new" data and burn up a little more of your drive's lifespan.

    8. Re:Has anyone waited 60 days? by justthinkit · · Score: 2

      They don't die randomly. They die on an exact schedule, and about three times sooner than more durable brands.

      --
      I come here for the love
    9. Re:Has anyone waited 60 days? by mlts · · Score: 1

      Does the tool need to be run on MS just once (like a firmware flash), or is it a driver in the OS? If the former, I can probably slap Windows on briefly just to run the fix. If it has to be loaded and run... heck with that. Intel may not be perfect, but they are a good baseline of what SSD should be measured by.

    10. Re:Has anyone waited 60 days? by Anonymous Coward · · Score: 0

      It is a 1-time thing. You can probably use something like Hiren's Boot CD (I'm sure many of you already have a copy laying around) to boot up MiniXP, and run the tool from there. It'll probably take an hour or so, but then you should be able to reboot your machine and be done with it.

      By the way, you guys running it on Windows might get some cryptic nonsense errors about non-standard firmware when trying to run the tool. It is most likely because you're using an AMD drive instead of the default Microsoft driver. You'll have to uninstall those drivers (make sure you mark them to be deleted and not just uninstalled), reboot your machine and then try the tool again. Once complete, revert to your AMD drivers.

    11. Re:Has anyone waited 60 days? by Anonymous Coward · · Score: 0

      It hasn't been out long enough to know if the issue has been resolved with TLC in general.

  3. Re:Apple already done did it by Tumbleweed · · Score: 4, Insightful

    Who do you think MAKES those M.2 SSDs in Apples? Hint: Samsung

  4. Can we get systems with M.2 ports on the front? by Gondola · · Score: 1

    So I can actually use M.2 drives like a flash drive? That would be awesome.

    1. Re:Can we get systems with M.2 ports on the front? by pushing-robot · · Score: 2

      I'm don't think M.2 can handle ESD and hotplugging, but Thunderbolt is essentially the external version of M.2.

      --
      How can I believe you when you tell me what I don't want to hear?
    2. Re:Can we get systems with M.2 ports on the front? by wonkey_monkey · · Score: 1

      You could, but the connector is only rated for 60 matings.

      --
      systemd is Roko's Basilisk.
    3. Re:Can we get systems with M.2 ports on the front? by adolf · · Score: 2

      Thunderbolt is essentially the external version of M.2.

      Thunderbolt and M.2 are alike in that they both have an implementation of PCI Express. They're otherwise rather dissimilar.

      So, they're essentially PCI Express.

    4. Re:Can we get systems with M.2 ports on the front? by hjf · · Score: 2

      Thunderbolt is dead. Apple was the only one to adopt it, and it was for only one generation. It's been replaced, for good or bad, for USB 3.

    5. Re:Can we get systems with M.2 ports on the front? by TwoEyedJack · · Score: 1

      ESD maybe not, but hot plugging is OK. Some of the fingers on the connector are longer to make sure things get powered in the correct order.

    6. Re:Can we get systems with M.2 ports on the front? by Em+Adespoton · · Score: 1

      Thunderbolt is dead. Apple was the only one to adopt it, and it was for only one generation. It's been replaced, for good or bad, for USB 3.

      My 2015 Retina MacBook Pro says otherwise. Unless you mean that Thunderbolt was replaced with Thunderbolt 2, which is true.

    7. Re:Can we get systems with M.2 ports on the front? by pushing-robot · · Score: 1

      Well, they both combine PCIe x4 and other common interfaces over a single electrical connector... they differ from an engineering standpoint, but for practical use TB is as close to an external M.2 as you'll find.

      --
      How can I believe you when you tell me what I don't want to hear?
    8. Re:Can we get systems with M.2 ports on the front? by jittles · · Score: 4, Funny

      You could, but the connector is only rated for 60 matings.

      60 matings is much better than most of the readers on slashdot will ever hope to achieve!

    9. Re:Can we get systems with M.2 ports on the front? by adolf · · Score: 2

      No, not the same at all.

      Thunderbolt has PCI Express and DisplayPort. It is used as an expansion bus for external peripherals.

      M.2 has PCI Express and USB and SATA. It is used an an expansion bus for internal peripherals.

      They're practically very dissimilar. Of the four electrical interfaces supported amongst them, they share just one in common. These aren't crazy words that only an engineer would understand.

      I can't drive a DisplayPort monitor with M.2, and I can't connect a SATA drive to Thunderbolt.

      SATA and eSATA are practically the same thing. M.2 and Thunderbolt are not.

      In other words, Thunderbolt is NOT to M.2 as eSATA is to SATA.

      In other words, both apples and oranges have a few things in common, but a lot more things that are not. At the end of the day, it's still apples and oranges.

      (Are we done yet?)

    10. Re:Can we get systems with M.2 ports on the front? by Khyber · · Score: 1

      "I can't connect a SATA drive to Thunderbolt"

      https://www.youtube.com/watch?...

      1s and 0s are fucking 1s and 0s. All that matters is that the data gets where it needs to go and has adequate bandwidth with which to do so.

      You could run a GPU off the M.2 slot. It's just a PCI-E 2.0 X2. You may not get the best performance obviously, but it would work. All it takes is the electrical contacts and data path and drivers.

      --
      Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
    11. Re:Can we get systems with M.2 ports on the front? by Hamsterdan · · Score: 1

      And YOU just made my day :)

      --
      I've got better things to do tonight than die.
    12. Re:Can we get systems with M.2 ports on the front? by adolf · · Score: 1

      You mean PCI-E 3.0 x4, SATA 3.0, and/or USB 3.0. Native, with pins dedicated to those purposes. (There's currently 5 different M.2 card keyings standardized.)

      Also, from your own link:

      I found stumbled across the Delock SATA to Thunderbolt adapter through a Google search, and I was hopeful that it might perform as well as it looks on paper. Unfortunately that's not the case.

      The biggest issue is how difficult it is to get a drive working with it. After plugging it in and attaching a drive, it's hit or miss whether that drive will actually mount on my Macbook Pro Retina.

      A secondary issue involves hot swapping disks. Once a drive is ejected from the Mac's interface and another is inserted, the Delock will not mount the second drive. I had to disconnect everything, re-attach, and go through the same hurdles I had to go through to get the drive working in the first place.

      When it does finally mount the performance is far below the stated 6 gigabits per second potential, with benchmarks on a high-end SSD capping out at 3 gigabits per second even though the drive can go faster than that.

      The bottom line? This product needs work. I can't recommend it.

      The difficulty stems from the fact that Thunderbolt does not include SATA, requiring funky PCI-E to SATA chips that barely work, whereas M.2 supports SATA natively.

      I can plug a SATA SSD into a USB adapter on my router and connect that with a serial cable to an ancient laptop with TCP/IP and NFS using SLIP or PPP or some other thing, and transfer fucking 1s and 0s. That doesn't mean that the serial port on the ancient laptop somehow groks SATA.

      (Now are we done yet?)

  5. heat? by sshir · · Score: 1

    Does it throttle under heavy use? My concern is that it's rather small and without heat sinks...

    1. Re:heat? by Khyber · · Score: 0

      As the M.2 drive sits on the PCI-E bus, it takes up PCI-E lanes. If you had 32 PCI-E lanes, you COULD have two GPUs on full 16x slots. Throw that M.2 in, well, now you've got 30 lanes, so at best, you're getting a 16X and 16x @ 8x lanes option. Remove the M.2 card, those lanes are free and you can run dual GPUs max throttle (assuming you've got CPUs that can keep up.)

      Does that clarify things for you, some?

      --
      Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
    2. Re:heat? by Anonymous Coward · · Score: 0

      No, it doesn't, it's irrelevant nonsense. GP asked made no mention of GPUs, and asked if the SSD itself would throttle. Then mentioned heat sinks.... on a fucking SSD. Your comment has literally no connection to GPs question whatsoever. Not to mention you fucking suck at math. If you have 32 PCI-E lanes, and you plug in an M.2 at 4x, you probably don't have 30 lanes left, though you may need to consult a mathematician to confirm that.

    3. Re:heat? by Khyber · · Score: 1

      If you put the SSD in, you've just throttled the system from being able to fully utilize two GPUs. Yes, the system will throttle. How you couldn't pull that away from my explanation is beyond me. M.2 PROVIDES 4x lanes PLUS SATA Express (which is another 2x lanes)

      Multiple functions are supported for add-in cards, including the following device classes: Wi-Fi, Bluetooth, satellite navigation, near field communication (NFC), digital radio, Wireless Gigabit Alliance (WiGig), wireless WAN (WWAN), and solid-state drives (SSDs). Exposed buses are PCI Express 3.0, Serial ATA (SATA) 3.0 and USB 3.0, which is backward compatible with USB 2.0. The SATA revision 3.2 specification, in its gold revision as of August 2013, standardizes the SATA M.2 as a new format for storage devices and specifies its hardware layout.

      That would also include physics accelerators and GPUs in those device classes. You might need a micro power connector for something with any reasonable power, but that's about it.

      Been playing with this since it was known as NGFF. When have you been using the stuff?

      --
      Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
  6. M.2 Specification by Archangel+Michael · · Score: 1

    Looking at the Wikipedia Article and the images for the different pinouts for the M.2 Specification, I have serious concerns about the ability to inadvertently flipping the cards, and inserting them upside down. Take a look at the B vs M configuration, which is exactly a mirror of each other.

    UNLESS there is part of the spec that I am not seeing about another notching somewhere, the ability to flip these over and inserting them wrong is going to be a huge issue. And looking at all the examples on the page, I don't see anything to mitigate against inserting these upside down.

    --
    Agent K: A *person* is smart. People are dumb, stupid, panicky animals, and you know it.
    1. Re:M.2 Specification by wonkey_monkey · · Score: 1

      Take a look at the B vs M configuration, which is exactly a mirror of each other.

      One is 6 pins wide, the other is 5 pins wide, so not exactly a mirror.

      --
      systemd is Roko's Basilisk.
    2. Re:M.2 Specification by Anonymous Coward · · Score: 0

      Instead of just looking at the wikipedia article, it might be worthwhile to actually read the words on it. First off, this is an internal board, if you're incapable of figuring out which type of card you need, reading installation instructions, or simply looking at the card and figuring out which side is obviously the top, you really should not be opening your computer case. Ever. Second, the B key is 6 pins, the M key is 5 pins. It's not physically possible to insert a card upside down.

    3. Re:M.2 Specification by Archangel+Michael · · Score: 1

      So, you can't stick a 5 pin connector into a six pin hole?

      And you have more faith in users than I do.

      --
      Agent K: A *person* is smart. People are dumb, stupid, panicky animals, and you know it.
    4. Re:M.2 Specification by mlts · · Score: 1

      I have worked with people who could stick any object in any connector... they just had to get a big enough of a hammer. The most common I've encountered are VGA plugs into serial ports (which bend the pins in all directions.)

      I am guessing that the people who designed this connector's configuration assumes it is not going to be user accessible for the post part, so they didn't really worry about it being 100% foolproof.

    5. Re:M.2 Specification by wonkey_monkey · · Score: 1

      So, you can't stick a 5 pin connector into a six pin hole?

      No, because both also have a longer set of pins next to the 5 or 6 pins.

      [n pins] [gap] [5 pins]
      [6 pins] [gap] [n-1 pins]

      --
      systemd is Roko's Basilisk.
    6. Re:M.2 Specification by Archangel+Michael · · Score: 1

      so they didn't really worry about it being 100% foolproof.

      Filed under "what could possibly go wrong"

      --
      Agent K: A *person* is smart. People are dumb, stupid, panicky animals, and you know it.
    7. Re:M.2 Specification by j-turkey · · Score: 1

      Having just purchased an M.2 drive (a non-NVMe Samsumg SM951 drive), installing the drive up-side down is about as likely as managing to install a DIMM backwards.

      --

      -Turkey

    8. Re:M.2 Specification by Anonymous Coward · · Score: 0

      If you create a connector that any idiot can use, then only idiots will find it useful.

      I've heard that some of Apples connectors can be connected any way you want.

  7. much needed!! by magarity · · Score: 3

    my work computer could really go for one of these; this kind of performance is needed to be able to grind through all the corporate security software.

  8. "Just annouced" eh? by Krazy+Kanuck · · Score: 2

    I suppose one could interpret a press release from January 7th as "Samsung has just announced its new", it was announced during CES on January 7th, here's the press release: http://www.samsung.com/global/... Linked article says model numbers haven't been released either.. here you go: 128GB NVME SSD SM951 - MZVPV128HDGL-00000 256GB NVME SSD SM951 - MZHPV256HDGL-00000 512GB NVME SSD SM951 - MZHPV512HDGL-00000 They've been shipping these in Lenovo and Apple laptops, they are scarce, but available (Amazon, RamCity, Ebay, Etc) at a little more than $1 per GB.

    1. Re:"Just annouced" eh? by Anonymous Coward · · Score: 0

      Yeah... people might be forgiven for thinking they had seen it before as well....

    2. Re:"Just annouced" eh? by Holi · · Score: 1

      No they haven't, the AHCI-Sm951 was launched in January, its the similar model names causing your confusion. OS X does not support NVME, so no, there are no NVME drives being supplied to Apple. And finally these NVME drives are just entering production and are not available in either the OEM or Retail channels yet.

      --
      Sorry, teleporters just kill you and then make a copy. A perfect, soul-less copy.
    3. Re:"Just annouced" eh? by Holi · · Score: 1

      "With its SM951 and other Samsung PCIe SSDs using a 10-nanometer class** MLC NAND flash production platform, Samsung believes it is in an excellent position to rapidly expand the global PCIe SSD market. Samsung will also work on timely introduction of next-generation PCIe SSDs that support the NVMe*** interface, which will provide even further performance increases." If you had read your own link you would have seen this gem. (emphasis mine).

      --
      Sorry, teleporters just kill you and then make a copy. A perfect, soul-less copy.
    4. Re:"Just annouced" eh? by tlhIngan · · Score: 1

      OS X does not support NVME, so no, there are no NVME drives being supplied to Apple. And finally these NVME drives are just entering production and are not available in either the OEM or Retail channels yet.

      The new MacBook supports NVMe on OS 10.10.3, so support is rapidly coming...

  9. Consumers are not going to notice much difference. by m.dillon · · Score: 2

    Well, nobody with a laptop is really going to notice much of a difference because frankly there isn't a whole lot of software that actually needs that kind of performance over the ~550 MBytes/sec that can already be obtained with SATA-III. Certainly not that would be run on a laptop anyway.

    It's just using the PCI-e lanes on the M.2 connector instead of the SATA-III lanes. This isn't a magical technology. There's a loss of robustness and portability that gets traded off. It does point to SATA needing another few speed bumps, though. The fundamental serial link technology used at the physical level by PCI-e and SATA is almost identical. The main difference is that SATA is designed for cabling while M.2 is not (at least not M.2's PCI-e lanes).

    -Matt

  10. Re:Consumers are not going to notice much differen by hjf · · Score: 1

    There are plenty of people working with 4K video nowadays. Even "just" HD video. A lot of folks move a LOT of data with "just their laptop". It's a trend. "Specialized workstations" we only know because we're here, but the truth is, most people just don't want a PC anymore.

    Want a shocker? A LOT of people are just not replacing their broken PC anymore. They're happy with what their phone or tablet can do. And if they do get a PC, it's almost always a laptop.

    Only gamers care for "big rigs" nowadays.

  11. Re:Consumers are not going to notice much differen by Em+Adespoton · · Score: 1

    Indeed... my raw editing platform for HD video is a MBP -- and even with the new drives, the R/W is still the bottleneck (but just barely).

    Any post production work goes onto beefier hardware, but for initial splicing and storyboarding of video, the MBP works quite well.

  12. Re:Spiderdick by Anonymous Coward · · Score: 0

    Gotta be Spiderdick.

    Does whatever a Spiderdick does...

  13. Re:Apple already done did it by Holi · · Score: 1

    No it's not. Apple does not support NVME so there are no NVME drives in any Apple products.

    --
    Sorry, teleporters just kill you and then make a copy. A perfect, soul-less copy.
  14. Re:Consumers are not going to notice much differen by mlts · · Score: 1

    The concept of a workstation has been pretty much marginalized due to things being "good enough". I might see one that is mainly to interact with a dedicated appliance (CNC mill), or perhaps a few workstations when working with definite tasks, but they tend to be bit players compared to desktops or laptops.

    The desktop is becoming a role, as opposed to a device. For example, the Surface Pro when plugged into a dock functions as a desktop role. Same with most laptops.

    As for laptops, they are nowhere near as expandable as a desktop... but they will do. A laptop with a decent SSD, 8-16 GB of RAM, and four cores can do OK at virtualization for small tasks.

  15. Re:Apple already done did it by Anonymous Coward · · Score: 1

    OS X 10.10.3 added NVME support for the new MacBook retina.

  16. Re:Consumers are not going to notice much differen by NJRoadfan · · Score: 1

    SATA Express is the cabled equivalent of M.2, but it is limited to 2 PCIe lanes vs. the 4 possible with M.2.

  17. I Want by sexconker · · Score: 1

    I want a bootable PCIe 4.0 (or 3.0) x16 (or x8) card that gives me 4 (or 2) m.2 slots and a RAID controller.
    Bonus points for passing through TRIM when possible (Intel does it with RAID 0 and RAID 1, I believe) and doing the OPAL/Bitlocker/whatever crypto passthrough so your OS can use the drive's built-in crypto instead of layering its own on top.

    Alternatively, give me comparable NVMe SSDs with a PCIe connector instead of an m.2 connector, and give me a motherboard with a RAID controller connected to the PCIe bus, supporting all of the above (I'll use a PCIe riser if I have to).

    Hell, I'd even settle for a bootable RAID controller behind 2 NVMe m.2 slots (PCIe 3.0 x4 or better for each) on the motherboard itself.

    But as it is, these types of drives are notoriously difficult to buy because Samsung sells them to OEMs only and when you do get them (from RamCity) you're stuck with what your mobo gives you for the m.2 slot or you're stuck with some dodgy m.2 -> PCIe card that gives you 2 slots at PCIe 3.0 2x, or PCIe 3.0 4x with no boot support, and never any way to properly RAID the damned things.

  18. Re:Consumers are not going to notice much differen by sexconker · · Score: 1

    That's a limitation of the Intel chipsets.
    SATA Express lets you throw x lanes of PCIe y. That's the whole point of SATA Express instead of doing SATA 12 Gbps. SATA Express is scalable.

  19. Re:Consumers are not going to notice much differen by Xoltri · · Score: 1

    640k ought to be enough for anybody!

    --
    -Xoltri
  20. Not sure that TFA is comparing apples to apples by j-turkey · · Score: 3, Informative

    ...is capable of sequential read and write speeds of 2,260 MB/sec and 1,600 MB/sec respectively. Comparable SATA-based M.2 SSDs typically can only push read/write speeds of 540 MB/sec and 500 MB/sec,

    Non-SATA M.2 drives are already on the market. Comparing the newest drive to SATA-based M.2 drives does not help much, I'd rather see it compared to what it supersedes. In this case, I'm more interested in a comparison with a PCIe 3.0 4-lane M.2 SSD drive that doesn't support NVMe. The drive specification for the earlier non-NVMe SM951 is not that far off of that of the new drive. The earlier drive is rated at sequential read and write speeds of 2,150 MB/sec 1,500 MB/sec respectively. Again, not all that far off.

    That being said...I'm curious to see the difference that NVMe makes in real-world benchmarks, and where the difference is...especially because I just built a new system with a non-NVMe SM951 SSD. :)

    --

    -Turkey

    1. Re:Not sure that TFA is comparing apples to apples by Anonymous Coward · · Score: 0

      NVMe doesn't do much in terms of bandwidth improvements for low queue depth workloads. It has significantly better latency though and supports much larger queue depths.

  21. Re:Apple already done did it by Anonymous Coward · · Score: 1

    No it's not. Apple does not support NVME so there are no NVME drives in any Apple products.

    Not true. The 2015 Macbook uses an NVMe drive.

  22. Re:Apple already done did it by Crypto+Gnome · · Score: 1

    No it's not. Apple does not support NVME so there are no NVME drives in any Apple products.

    Yet....

    --
    Visit CryptoGnome in his home.
  23. Re: Can we get systems with M.2 ports on the front by Anonymous Coward · · Score: 0

    Apple tried to monopolise FireWire in the same way as they've tried to monopolise thunderbolt. And the same thing has happened. Sure, there are and will be niche uses, but Thunderbolt's opportunity to be The Standard for next gen peripherals has passed.

  24. Re:Consumers are not going to notice much differen by Blaskowicz · · Score: 1

    There is a nice improvement going from AHCI to NVMe protocol, though. I/O gets lower latency, less CPU intensive, less "blocking".

    That may seem "philosophical" still. At a first approximation latency is halved. The tech will be a good thing to have once the drives get plentiful and cheap.

  25. Re:Consumers are not going to notice much differen by Blaskowicz · · Score: 1

    Weren't workstations of the 80s and 90s just powerful microcomputers?
    CPU with memory protection (e.g. 68010 plus MMU), SCSI disks, high resolution (about a megapixel), several megabytes of memory, advanced OS : Unix-like, Windows NT or something else.

    By that measure, any good low end desktop computer is a workstation. By 2001, that had Windows XP and Ultra DMA IDE modes ; a decade or less later we had SATA with NCQ (no need for SCSI), support for dual monitor and SMP as standard (dual and quad core).