back to article Everspin's new gig: a gig or two of non-volatile RAM on PCIe

Non-volatile memory outfit Everspin's popped some of its Spin Torque MRAM onto a PCIe card in the hope system builders get excited about a new tier of memory. Or is it a new tier of storage? The first of the new “nvNITRO E” range will be a half-height, half-length PCIe card that can operate as an NVMe solid state disk, or as …

  1. Anonymous Coward
    Anonymous Coward

    low latency trading

    ... does not have much use for a device with 6us latency. Any data stores with this much latency would be done asynchronously and away from the critical path, which means there is little difference between 20us and 6us. Also, few GB capacity is a joke, it is much simpler and cheaper just to mmap a chunk of memory from RAM.

  2. Aitor 1 Silver badge

    KISS

    Unless there is a strong reason to go this route, I would avoid it.

    It just makes things too complicated.. and with SSD speeds being what they are these days and enterprise SSDs having supercaps for them to finish writing in case of power failure, I just dont see the point of this.

    Yes, in some critical infraestructure it might be useful... but that infraestructure already has secure power AND I suspect everyone is using ACID databases properly.. so again, what is the point here?

    This looks like a lot like the XBox One special small memory thing.. that they just god rid of.

  3. Ogi

    > The first of the new “nvNITRO E” range will be a half-height, half-length PCIe card that can operate as an NVMe solid state disk, or as memory mapped IO (MMIO).

    How is this different to other NVMe setups? I have a PCIe NVM card (120GB) in my server. It cost me £70 all in, and is rated at 6Gb/s bulk transfer and some stupidly high IOPS that I can't remember right now.

    I can use it as swap (in which case it just becomes allocatable memory, and the OS handles all the paged MMAPing), or I can use it as a file store, and MMAP files directly on it for the same effect.

    I don't see what is special about this startup, except their NVMe offerings are really low capacity? The 120GB SSD is running as swap and it actually works pretty well. I have used ~105GB of swap on the 32GB RAM machine, and it was still usable and churned out data at an acceptable rate (this was for peak loading, most of the time 32GB is enough, I just didn't want the machine to die when the peaks come in, and couldn't justify the cost of 256GB of RAM for it ).

    > Everspin asserts its product is rather faster than Intel's 10 µs and, critically, that you can read and write to it all you like without the prospect of the medium degrading.

    That will depend on the cost. When my SSD eventually wears out, I will just buy another one (as they would most likely have gotten cheaper by then) and carry on.

    It might be cheaper to just treat the SSD as consumable, and replace when they wear out. Each time you replace you will get a newer/faster/cheaper/higher capacity version due to the march of technology.

    > The cards claim 1,500,000 IOPS with six microsecond end-to-end latency, making them rather useful in scenarios like high-frequency trading where the odd microsecond can be the difference between profitability and purgatory.

    HFT shops have long since moved away into FPGAs with local RAM, computers have been relegated to babysitting the FPGAs and monitoring/restarting them as and when needed. You don't need uber low latency memory for that.

    They do mention merging their memory with FPGAs, and that might prove an useful niche for the technology, but that hasn't been done yet, and no idea if it is a worthwhile and profitable niche (FPGAs do need to store some data, but not much, most of HFT is simple arb strategies just done stupidly fast based on data in/out of the network port)

    1. iTheHuman

      Your ssd would have a much broader latency curve that results in those occasional several ms pe operations.

      The small sizes aren't very useful for enterprises, but nvm with dram-like latency distributions over pcie is a big deal. For one thing, dram sucks up power even when it's not doing anything and has to refresh its leaky capacitors every ~50-100us.

  4. tomocean

    MRAM is not an SSD replacement

    These emerging NVMe storage class memory technologies (i.e. RERAM, MRAM, PCM, et al) are an entirely new class of storage (though some of the technologies have been in research for many years). They are generally 1000 times faster than SSD and have characteristics closer to DRAM than SSD. In the future, when coupled with SCM, it is likely that SSD will become the equivalent of a spinning magnetic hard drive.

  5. Fazal Majid

    DDRDrive

    DDRDrive introduced a similar product years ago. It held 4GB of DRAM backed by GB of SLC NAND and a supercapacitor, for $2000 list. If the card detects power loss, the supercap has enough juice to write the RAM contents to the flash. It was primarily marketed to ZFS users to accelerate the ZFS intent log (write cache), as DRAM does not suffer from the performance cliff of most SSDs.

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Biting the hand that feeds IT © 1998–2022