Tag: dimm

Posted on

New family of Intel Xeon Scalable Processors enable software defined data infrastructures (SDDI) and SDDC

Intel Xeon Scalable Processors SDDI and SDDC

server storage I/O data infrastructure trends

Today Intel announced a new family of Xeon Scalable Processors (aka Purely) that for some workloads Intel claims to be on average of 1.65x faster than their predecessors. Note your real improvement will vary based on workload, configuration, benchmark testing, type of processor, memory, and many other server storage I/O performance considerations.

Intel Scalable Xeon Processors
Image via Intel.com

In general the new Intel Xeon Scalable Processors enable legacy and software defined data infrastructures (SDDI), along with software defined data centers (SDDC), cloud and other environments to support expanding workloads more efficiently as well as effectively (e.g. boosting productivity).

Data Infrastructures and workloads

Some target application and environment workloads Intel is positioning these new processors for includes among others:

  • Machine Learning (ML), Artificial Intelligence (AI), advanced analytics, deep learning and big data
  • Networking including software defined network (SDN) and network function virtualization (NFV)
  • Cloud and Virtualization including Azure Stack, Docker and Kubernetes containers, Hyper-V, KVM, OpenStack VMware vSphere, KVM among others
  • High Performance Compute (HPC) and High Productivity Compute (e.g. the other HPC)
  • Storage including legacy and emerging software defined storage software deployed as appliances, systems or server less deployment modes.

Features of the new Intel Xeon Scalable Processors include:

  • New core micro architecture with interconnects and on die memory controllers
  • Sockets (processors) scalable up to 28 cores
  • Improved networking performance using Quick Assist and Data Plane Development Kit (DPDK)
  • Leverages Intel Quick Assist Technology for CPU offload of compute intensive functions including I/O networking, security, AI, ML, big data, analytics and storage functions. Functions that benefit from Quick Assist include cryptography, encryption, authentication, cipher operations, digital signatures, key exchange, loss less data compression and data footprint reduction along with data at rest encryption (DARE).
  • Optane Non-Volatile Dual Inline Memory Module (NVDIMM) for storage class memory (SCM) also referred to by some as Persistent Memory (PM), not to be confused with Physical Machine (PM).
  • Supports Advanced Vector Extensions 512  (AVX-512) for HPC and other workloads
  • Optional Omni-Path Fabrics in addition to 1/10Gb Ethernet among other I/O options
  • Six memory channels supporting up to 6TB of RDIMM with multi socket systems
  • From two to eight  sockets per node (system)
  • Systems support PCIe 3.x (some supporting x4 based M.2 interconnects)

Note that exact speeds, feeds, slots and watts will vary by specific server model and vendor options. Also note that some server system solutions have two or more nodes (e.g. two or more real servers) in a single package not to be confused with two or more sockets per node (system or motherboard). Refer to the where to learn more section below for links to Intel benchmarks and other resources.

Software Defined Data Infrastructures, SDDC, SDX and SDDI

What About Speeds and Feeds

Watch for and check out the various Intel partners who have or will be announcing their new server compute platforms based on Intel Xeon Scalable Processors. Each of the different vendors will have various speeds and feeds options that build on the fundamental Intel Xeon Scalable Processor capabilities.

For example Dell EMC announced their 14G server platforms at the May 2017 Dell EMC World event with details to follow (e.g. after the Intel announcements).

Some things to keep in mind include the amount of DDR4 DRAM (or Optane NVDIMM) will vary by vendors server platform configuration, motherboards, several sockets and DIMM slots. Also keep in mind the differences between registered (e.g. buffered RDIMM) that give good capacity and great performance, and load reduced DIMM (LRDIMM) that have great capacity and ok performance.

Various nvme options

What about NVMe

It’s there as these systems like previous Intel models support NVMe devices via PCIe 3.x slots, and some vendor solutions also supporting M.2 x4 physical interconnects as well.

server storageIO flash and SSD
Image via Software Defined Data Infrastructure Essentials (CRC)

Note that Broadcom formerly known as Avago and LSI recently announced PCIe based RAID and adapter cards that support NVMe attached devices in addition to SAS and SATA.

server storage data infrastructure sddi

What About Intel and Storage

In case you have not connected the dots yet, the Intel Xeon Scalable Processor based server (aka compute) systems are also a fundamental platform for storage systems, services, solutions, appliances along with tin-wrapped software.

What this means is that the Intel Xeon Scalable Processors based systems can be used for deploying legacy as well as new and emerging software-defined storage software solutions. This also means that the Intel platforms can be used to support SDDC, SDDI, SDX, SDI as well as other forms of legacy and software-defined data infrastructures along with cloud, virtual, container, server less among other modes of deployment.

Image Via Intel.com

Moving beyond server and compute platforms, there is another tie to storage as part of this recent as well as other Intel announcements. Just a few weeks ago Intel announced 64 layer triple level cell (TLC) 3D NAND solutions positioned for the client market (laptop, workstations, tablets, thin clients). Intel with that announcement increased the traditional aerial density (e.g. bits per square inch or cm) as well as boosting the number of layers (stacking more bits as well).

The net result is not only more bits per square inch, also more per cubic inch or cm. This is all part of a continued evolution of NAND flash including from 2D to 3D, MCL to TLC, 32 to 64 layer.  In other words, NAND flash-based Solid State Devices (SSDs) are very much still a relevant and continue to be enhanced technology even with the emerging 3D XPoint and Optane (also available via Amazon in M.2) in the wings.

server memory evolution
Via Intel and Micron (3D XPoint launch)

Keep in mind that NAND flash-based technologies were announced almost 20 years ago (1999), and are still evolving. 3D XPoint announced two years ago, along with other emerging storage class memories (SCM), non-volatile memory (NVM) and persistent memory (PM) devices are part of the future as is 3D NAND (among others). Speaking of 3D XPoint and Optane, Intel had announcements about that in the past as well.

Where To Learn More

Learn more about Intel Xeon Scalable Processors along with related technology, trends, tools, techniques and tips with the following links.

What This All Means

Some say the PC is dead and IMHO that depends on what you mean or define a PC as. For example if you refer to a PC generically to also include servers besides workstations or other devices, then they are alive. If however your view is that PCs are only workstations and client devices, then they are on the decline.

However if your view is that a PC is defined by the underlying processor such as Intel general purpose 64 bit x86 derivative (or descendent) then they are very much alive. Just as older generations of PCs leveraging general purpose Intel based x86 (and its predecessors) processors were deployed for many uses, so to are today’s line of Xeon (among others) processors.

Even with the increase of ARM, GPU and other specialized processors, as well as ASIC and FPGAs for offloads, the role of general purpose processors continues to increase, as does the technology evolution around. Even with so called server less architectures, they still need underlying compute server platforms for running software, which also includes software defined storage, software defined networks, SDDC, SDDI, SDX, IoT among others.

Overall this is a good set of announcements by Intel and what we can also expect to be a flood of enhancements from their partners who will use the new family of Intel Xeon Scalable Processors in their products to enable software defined data infrastructures (SDDI) and SDDC.

Ok, nuff said (for now…).

Cheers
Gs

Greg Schulz – Multi-year Microsoft MVP Cloud and Data Center Management, VMware vExpert (and vSAN). Author Cloud and Virtual Data Storage Networking (CRC Press), The Green and Virtual Data Center (CRC Press), Resilient Storage Networks (Elsevier) and twitter @storageio. Watch for the spring 2017 release of his new book "Software-Defined Data Infrastructure Essentials" (CRC Press).

Courteous comments are welcome for consideration. First published on https://storageioblog.com any reproduction in whole, in part, with changes to content, without source attribution under title or without permission is forbidden.

All Comments, (C) and (TM) belong to their owners/posters, Other content (C) Copyright 2006-2023 Server StorageIO(R) and UnlimitedIO. All Rights Reserved.

Posted on

NVMe Place NVM Non Volatile Memory Express Resources

Updated 8/31/19
NVMe place server Storage I/O data infrastructure trends

Welcome to NVMe place NVM Non Volatile Memory Express Resources. NVMe place is about Non Volatile Memory (NVM) Express (NVMe) with Industry Trends Perspectives, Tips, Tools, Techniques, Technologies, News and other information.

Disclaimer

Please note that this NVMe place resources site is independent of the industry trade and promoters group NVM Express, Inc. (e.g. www.nvmexpress.org). NVM Express, Inc. is the sole owner of the NVM Express specifications and trademarks.

NVM Express Organization
Image used with permission of NVM Express, Inc.

Visit the NVM Express industry promoters site here to learn more about their members, news, events, product information, software driver downloads, and other useful NVMe resources content.

 

The NVMe Place resources and NVM including SCM, PMEM, Flash

NVMe place includes Non Volatile Memory (NVM) including nand flash, storage class memories (SCM), persistent memories (PM) are storage memory mediums while NVM Express (NVMe) is an interface for accessing NVM. This NVMe resources page is a companion to The SSD Place which has a broader Non Volatile Memory (NVM) focus including flash among other SSD topics. NVMe is a new server storage I/O access method and protocol for fast access to NVM based storage and memory technologies. NVMe is an alternative to existing block based server storage I/O access protocols such as AHCI/SATA and SCSI/SAS devices commonly used for access Hard Disk Drives (HDD) along with SSD among other things.

Server Storage I/O NVMe PCIe SAS SATA AHCI
Comparing AHCI/SATA, SCSI/SAS and NVMe all of which can coexist to address different needs.

Leveraging the standard PCIe hardware interface, NVMe based devices (that have an NVMe controller) can be accessed via various operating systems (and hypervisors such as VMware ESXi) with both in the box drivers or optional third-party device drivers. Devices that support NVMe can be 2.5″ drive format packaged that use a converged 8637/8639 connector (e.g. PCIe x4) coexisting with SAS and SATA devices as well as being add-in card (AIC) PCIe cards supporting x4, x8 and other implementations. Initially, NVMe is being positioned as a back-end to servers (or storage systems) interface for accessing fast flash and other NVM based devices.

NVMe as back-end storage
NVMe as a “back-end” I/O interface for NVM storage media

NVMe as front-end server storage I/O interface
NVMe as a “front-end” interface for servers or storage systems/appliances

NVMe has also been shown to work over low latency, high-speed RDMA based network interfaces including RoCE (RDMA over Converged Ethernet) and InfiniBand (read more here, here and here involving Mangstor, Mellanox and PMC among others). What this means is that like SCSI based SAS which can be both a back-end drive (HDD, SSD, etc) access protocol and interface, NVMe can also being used for back-end can also be used as a front-end of server to storage interface like how Fibre Channel SCSI_Protocol (aka FCP), SCSI based iSCSI, SCSI RDMA Protocol via InfiniBand (among others) are used.

NVMe features

Main features of NVMe include among others:

  • Lower latency due to improve drivers and increased queues (and queue sizes)
  • Lower CPU used to handle larger number of I/Os (more CPU available for useful work)
  • Higher I/O activity rates (IOPs) to boost productivity unlock value of fast flash and NVM
  • Bandwidth improvements leveraging various fast PCIe interface and available lanes
  • Dual-pathing of devices like what is available with dual-path SAS devices
  • Unlock the value of more cores per processor socket and software threads (productivity)
  • Various packaging options, deployment scenarios and configuration options
  • Appears as a standard storage device on most operating systems
  • Plug-play with in-box drivers on many popular operating systems and hypervisors

Shared external PCIe using NVMe
NVMe and shared PCIe (e.g. shared PCIe flash DAS)

NVMe related content and links

The following are some of my tips, articles, blog posts, presentations and other content, along with material from others pertaining to NVMe. Keep in mind that the question should not be if NVMe is in your future, rather when, where, with what, from whom and how much of it will be used as well as how it will be used.

  • How to Prepare for the NVMe Server Storage I/O Wave (Via Micron.com)
  • Why NVMe Should Be in Your Data Center (Via Micron.com)
  • NVMe U2 (8639) vs. M2 interfaces (Via Gamersnexus)
  • Enmotus FuzeDrive MicroTiering (StorageIO Lab Report)
  • EMC DSSD D5 Rack Scale Direct Attached Shared SSD All Flash Array Part I (Via StorageIOBlog)
  • Part II – EMC DSSD D5 Direct Attached Shared AFA (Via StorageIOBlog)
  • NAND, DRAM, SAS/SCSI & SATA/AHCI: Not Dead, Yet! (Via EnterpriseStorageForum)
  • Non Volatile Memory (NVM), NVMe, Flash Memory Summit and SSD updates (Via StorageIOblog)
  • Microsoft and Intel showcase Storage Spaces Direct with NVM Express at IDF ’15 (Via TechNet)
  • MNVM Express solutions (Via SuperMicro)
  • Gaining Server Storage I/O Insight into Microsoft Windows Server 2016 (Via StorageIOblog)
  • PMC-Sierra Scales Storage with PCIe, NVMe (Via EEtimes)
  • RoCE updates among other items (Via InfiniBand Trade Association (IBTA) December Newsletter)
  • NVMe: The Golden Ticket for Faster Flash Storage? (Via EnterpriseStorageForum)
  • What should I consider when using SSD cloud? (Via SearchCloudStorage)
  • MSP CMG, Sept. 2014 Presentation (Flash back to reality – Myths and Realities – Flash and SSD Industry trends perspectives plus benchmarking tips)– PDF
  • Selecting Storage: Start With Requirements (Via NetworkComputing)
  • PMC Announces Flashtec NVMe SSD NVMe2106, NVMe2032 Controllers With LDPC (Via TomsITpro)
  • Exclusive: If Intel and Micron’s “Xpoint” is 3D Phase Change Memory, Boy Did They Patent It (Via Dailytech)
  • Intel & Micron 3D XPoint memory — is it just CBRAM hyped up? Curation of various posts (Via Computerworld)
  • How many IOPS can a HDD, HHDD or SSD do (Part I)?
  • How many IOPS can a HDD, HHDD or SSD do with VMware? (Part II)
  • I/O Performance Issues and Impacts on Time-Sensitive Applications (Via CMG)
  • Via EnterpriseStorageForum: 5 Hot Storage Technologies to Watch
  • Via EnterpriseStorageForum: 10-Year Review of Data Storage

Non-Volatile Memory (NVM) Express (NVMe) continues to evolve as a technology for enabling and improving server storage I/O for NVM including nand flash SSD storage. NVMe streamline performance enabling more work to be done (e.g. IOPs), data to be moved (bandwidth) at a lower response time using less CPU.

NVMe and SATA flash SSD performance

The above figure is a quick look comparing nand flash SSD being accessed via SATA III (6Gbps) on the left and NVMe (x4) on the right. As with any server storage I/O performance comparisons there are many variables and take them with a grain of salt. While IOPs and bandwidth are often discussed, keep in mind that with the new protocol, drivers and device controllers with NVMe that streamline I/O less CPU is needed.

Additional NVMe Resources

Also check out the Server StorageIO companion micro sites landing pages including thessdplace.com (SSD focus), data protection diaries (backup, BC/DR/HA and related topics), cloud and object storage, and server storage I/O performance and benchmarking here.

If you are in to the real bits and bytes details such as at device driver level content check out the Linux NVMe reflector forum. The linux-nvme forum is a good source if you are developer to stay up on what is happening in and around device driver and associated topics.

Additional learning experiences along with common questions (and answers), as well as tips can be found in Software Defined Data Infrastructure Essentials book.

Software Defined Data Infrastructure Essentials Book SDDC

Disclaimer

Disclaimer: Please note that this site is independent of the industry trade and promoters group NVM Express, Inc. (e.g. www.nvmexpress.org). NVM Express, Inc. is the sole owner of the NVM Express specifications and trademarks. Check out the NVM Express industry promoters site here to learn more about their members, news, events, product information, software driver downloads, and other useful NVMe resources content.

NVM Express Organization
Image used with permission of NVM Express, Inc.

Wrap Up

Watch for updates with more content, links and NVMe resources to be added here soon.

Ok, nuff said (for now)

Cheers
Gs

Greg Schulz – Microsoft MVP Cloud and Data Center Management, VMware vExpert 2010-2017 (vSAN and vCloud). Author of Software Defined Data Infrastructure Essentials (CRC Press), as well as Cloud and Virtual Data Storage Networking (CRC Press), The Green and Virtual Data Center (CRC Press), Resilient Storage Networks (Elsevier) and twitter @storageio. Courteous comments are welcome for consideration. First published on https://storageioblog.com any reproduction in whole, in part, with changes to content, without source attribution under title or without permission is forbidden.

All Comments, (C) and (TM) belong to their owners/posters, Other content (C) Copyright 2006-2024 Server StorageIO and UnlimitedIO. All Rights Reserved. StorageIO is a registered Trade Mark (TM) of Server StorageIO.

Posted on

Intel Micron 3D XPoint server storage NVM SCM PM SSD

3D XPoint server storage class memory SCM


Storage I/O trends

Updated 1/31/2018

Intel Micron 3D XPoint server storage NVM SCM PM SSD.

This is the second of a three-part series on the recent Intel and Micron 3D XPoint server storage memory announcement. Read Part I here and Part III here.

Is this 3D XPoint marketing, manufacturing or material technology?

You can’t have a successful manufactured material technology without some marketing, likewise marketing without some manufactured material would be manufactured marketing. In the case of 3D XPoint and its announcement launch, their real technology shown, granted it was only wafer and dies as opposed to an actual DDR4 DIMM or PCIe Add In Card (AIC) or drive form factor Solid State Device (SSD) product. On the other hand, on a relative comparison basis, even though there is marketing collateral available to learn more from, this was far from a over the big-top made for TV or web circus event, which can be a good thing.


Wafer unveiled containing 3D XPoint 128 Gb dies

Who will get access to 3D XPoint?

Initially 3D XPoint production capacity supply will be for the two companies to offer early samples to their customers later this year with general production slated for 2016 meaning early real customer deployed products starting sometime in 2016.

Is it NAND or NOT?

3D XPoint is not NAND flash, it is also not NVRAM or DRAM, it’s a new class of NVM that can be used for server class main memory with persistency, or as persistent data storage among other uses (cell phones, automobiles, appliances and other electronics). In addition, 3D XPoint is more durable with a longer useful life for writing and storing data vs. NAND flash.

Why is 3D XPoint important?

As mentioned during the Intel and Micron announcement, there have only been seven major memory technologies introduced since the transistor back in 1947, granted there have been many variations along with generational enhancements of those. Thus 3D XPoint is being positioned by Intel and Micron as the eighth memory class joining its predecessors many of which continue to be used today in various roles.


Major memory classes or categories timeline

In addition to the above memory classes or categories timeline, the following shows in more detail various memory categories (click on the image below to get access to the Intel interactive infographic).

Intel History of Memory Infographic
Via: https://intelsalestraining.com/memory timeline/ (Click on image to view)

What capacity size is 3D XPoint?

Initially the 3D XPoint technology is available in a 2 layer 128 bit (cell) per die capacity. Keep in mind that there are usually 8 bits to a byte resulting in 16 GByte capacity per chip initially. With density improvements, as well as increased stacking of layers, the number of cells or bits per die (e.g. what makes up a chip) should improve, as well as most implementations will have multiple chips in some type of configuration.

What will 3D XPoint cost?

During the 3D XPoint launch webinar Intel and Micron hinted that first pricing will be between current DRAM and NAND flash on a per cell or bit basis, however real pricing and costs will vary depending on how packaged for use. For example if placed on a DDR4 or different type of DIMM or on a PCIe Add In Card (AIC) or as a drive form factor SSD among other options will vary the real price. Likewise as with other memories and storage mediums, as production yields and volumes increase, along with denser designs, the cost per usable cell or bit can be expected to further improve.

Where to read, watch and learn more

Storage I/O trends

Additional learning experiences along with common questions (and answers), as well as tips can be found in Software Defined Data Infrastructure Essentials book.

Software Defined Data Infrastructure Essentials Book SDDC

What This All Means

DRAM which has been around for sometime has plenty of life left for many applications as does NAND flash including new 3D NAND, vNAND and other variations. For the next several years, there will be a co-existences between new and old NVM and DRAM among other memory technologies including 3D XPoint. Read more in this series including Part I here and Part III here.

Disclosure: Micron and Intel have been direct and/or indirect clients in the past via third-parties and partners, also I have bought and use some of their technologies direct and/or in-direct via their partners.

Ok, nuff said, for now.

Gs

Greg Schulz – Microsoft MVP Cloud and Data Center Management, VMware vExpert 2010-2017 (vSAN and vCloud). Author of Software Defined Data Infrastructure Essentials (CRC Press), as well as Cloud and Virtual Data Storage Networking (CRC Press), The Green and Virtual Data Center (CRC Press), Resilient Storage Networks (Elsevier) and twitter @storageio. Courteous comments are welcome for consideration. First published on https://storageioblog.com any reproduction in whole, in part, with changes to content, without source attribution under title or without permission is forbidden.

All Comments, (C) and (TM) belong to their owners/posters, Other content (C) Copyright 2006-2024 Server StorageIO and UnlimitedIO. All Rights Reserved. StorageIO is a registered Trade Mark (TM) of Server StorageIO.

Posted on

3D XPoint nvm pm scm storage class memory

Part III – 3D XPoint server storage class memory SCM


Storage I/O trends

Updated 1/31/2018

3D XPoint nvm pm scm storage class memory.

This is the third of a three-part series on the recent Intel and Micron 3D XPoint server storage memory announcement. Read Part I here and Part II here.

What is 3D XPoint and how does it work?

3D XPoint is a new class or class of memory (view other categories of memory here) that provides performance for reads and writes closer to that of DRAM with about 10x the capacity density. In addition to the speed closer to DRAM vs. the lower NAND flash, 3D XPoint is also non-volatile memory (NVM) like NAND flash, NVRAM and others. What this means is that 3D XPoint can be used as persistent higher density fast server memory (or main memory for other computers and electronics). Besides being fast persistent main memory, 3D XPoint will also be a faster medium for solid state devices (SSD’s) including PCIe Add In Cards (AIC), m2 cards and drive form factor 8637/8639 NVM Express (NVMe) accessed devices that also has better endurance or life span compared to NAND flash.


3D XPoint architecture and attributes

The initial die or basic chip building block 3D XPoint implementation is a layer 128 Gbit device which if using 8 bits would yield 16GB raw. Over time increased densities should become available as the bit density improves with more cells and further scaling of the technology, combined with packaging. For example while a current die could hold up to 16 GBytes of data, multiple dies could be packaged together to create a 32GB, 64GB, 128GB etc. or larger actual product. Think about not only where packaged flash based SSD capacities are today, also think in terms of where DDR3 and DDR4 DIMM are at such as 4GB, 8GB, 16GB, 32GB densities.

The 3D aspect comes from the memory being in a matrix initially being two layers high, with multiple rows and columns that intersect, where those intersections occur is a microscopic material based switch for accessing a particular memory cell. Unlike NAND flash where an individual cell or bit is accessed as part of a larger block or page comprising several thousand bytes at once, 3D XPoint cells or bits can be individually accessed to speed up reads and writes in a more granular fashion. It is this more granular access along with performance that will enable 3D XPoint to be used in lower latency scenarios where DRAM would normally be used.

Instead of trapping electrons in a cell to create a bit of capacity (e.g. on or off) like NAND flash, 3D XPoint leverages the underlying physical material propertied to store a bit as a phase change enabling use of all cells. In other words, instead of being electron based, it is material based. While Intel and Micron did not specify what the actual chemistry and physical materials that are used in 3D XPoint, they did discuss some of the characteristics. If you want to go deep, check out how the Dailytech makes an interesting educated speculation or thesis on the underlying technology.

Watch the following video to get a better idea and visually see how 3D XPoint works.



3D XPoint YouTube Video

What are these chips, cells, wafers and dies?

Left many dies on a wafer, right, a closer look at the dies cut from the wafer

Dies (here and here) are the basic building block of what goes into the chips that in turn are the components used for creating DDR DIMM for main computer memory, as well as for create SD and MicroSD cards, USB thumb drives, PCIe AIC and drive form factor SSD, as well as custom modules on motherboards, or consumption via bare die and wafer level consumption (e.g. where you are doing really custom things at volume, beyond using a soldering iron scale).

Storage I/O trends

Has Intel and Micron cornered the NVM and memory market?

We have heard proclamations, speculation and statements of the demise of DRAM, NAND flash and other volatile and NVM memories for years, if not decades now. Each year there is the usual this will be the year of “x” where “x” can include among others. Resistive RAM aka ReRAM or RRAM aka the memristor that HP earlier announced they were going to bring to market and then earlier this year canceling those plans while Crossbar continues to pursue RRAM. MRAM or Magnetorestive RAM, Phase Change Memory aka CRAM or PCM and PRAM, FRAM aka FeRAM or Ferroelectric RAM among others.

flash SSD and NVM trends

Expanding persistent memory and SSD storage markets

Keep in mind that there are many steps taking time measured in years or decades to go from research and development lab idea to prototype that can then be produced at production volumes in economic yields. As a reference for, there is still plenty of life in both DRAM as well as NAND flash, the later having appeared around 1989.

Industry vs. Customer Adoption and deployment timeline

Technology industry adoption precedes customer adoption and deployment

There is a difference between industry adoption and deployment vs. customer adoption and deployment, they are related, yet separated by time as shown in the above figure. What this means is that there can be several years from the time a new technology is initially introduced and when it becomes generally available. Keep in mind that NAND flash has yet to reach its full market potential despite having made significant inroads the past few years since it was introduced in 1989.

This begs the question of if 3D XPoint is a variation of phase change, RRAM, MRAM or something else. Over at the Dailytech they lay out a line of thinking (or educated speculation) that 3D XPoint is some derivative or variation of phase change, time will tell about what it really is.

What’s the difference between 3D NAND flash and 3D XPoint?

3D NAND is a form of NAND flash NVM, while 3D XPoint is a completely new and different type of NVM (e.g. its not NAND).

3D NAND is a variation of traditional flash with the difference between vertical stacking vs. horizontal to improve density, also known as vertical NAND or V-NAND. Vertical stacking is like building up to house more tenants or occupants in a dense environment or scaling up, vs scaling-out by using up more space where density is not an issue. Note that magnetic HDD’s shifted to perpendicular (e.g. vertical) recording about ten years ago to break through the super parametric barrier and more recently, magnetic tape has also adopted perpendicular recording. Also keep in mind that 3D XPoint and the earlier announced Intel and Micron 3D NAND flash are two separate classes of memory that both just happen to have 3D in their marketing names.

Where to read, watch and learn more

Storage I/O trends

Additional learning experiences along with common questions (and answers), as well as tips can be found in Software Defined Data Infrastructure Essentials book.

Software Defined Data Infrastructure Essentials Book SDDC

What This All Means

First, keep in mind that this is very early in the 3D XPoint technology evolution life-cycle and both DRAM and NAND flash will not be dead at least near term. Keep in mind that NAND flash appeared back in 1989 and only over the past several years has finally hit its mainstream adoption stride with plenty of market upside left. Same with DRAM which has been around for sometime, it too still has plenty of life left for many applications. However other applications that have the need for improved speed over NAND flash, or persistency and density vs. DRAM will be some of the first to leverage new NVM technologies such as 3D XPoint. Thus at least for the next several years, there will be a co-existences between new and old NVM and DRAM among other memory technologies. Bottom line, 3D XPoint is a new class of NVM memory, can be used for persistent main server memory or for persistent fast storage memory. If you have not done so, check out Part I here and Part II here of this three-part series on Intel and Micron 3D XPoint.

Disclosure: Micron and Intel have been direct and/or indirect clients in the past via third-parties and partners, also I have bought and use some of their technologies direct and/or in-direct via their partners.

Ok, nuff said, for now.

Gs

Greg Schulz – Microsoft MVP Cloud and Data Center Management, VMware vExpert 2010-2017 (vSAN and vCloud). Author of Software Defined Data Infrastructure Essentials (CRC Press), as well as Cloud and Virtual Data Storage Networking (CRC Press), The Green and Virtual Data Center (CRC Press), Resilient Storage Networks (Elsevier) and twitter @storageio. Courteous comments are welcome for consideration. First published on https://storageioblog.com any reproduction in whole, in part, with changes to content, without source attribution under title or without permission is forbidden.

All Comments, (C) and (TM) belong to their owners/posters, Other content (C) Copyright 2006-2024 Server StorageIO and UnlimitedIO. All Rights Reserved. StorageIO is a registered Trade Mark (TM) of Server StorageIO.

Posted on

Intel Micron unveil new 3D XPoint Non Volatie Memory NVM for servers storage

3D XPoint NVM persistent memory PM storage class memory SCM


Storage I/O trends

Updated 1/31/2018

This is the first of a three-part series on Intel Micron unveil new 3D XPoint Non Volatie Memory NVM for servers storage announcement. Read Part II here and Part III here.

In a webcast the other day, Intel and Micron announced new 3D XPoint non-volatile memory (NVM) that can be used for both primary main memory (e.g. what’s in computers, serves, laptops, tablets and many other things) in place of Dynamic Random Access Memory (DRAM), for persistent storage faster than today’s NAND flash-based solid state devices (SSD), not to mention future hybrid usage scenarios. Note that this announcement while having the common term 3D in it is different from the earlier Intel and Micron announcement about 3D NAND flash (read more about that here).

Twitter hash tag #3DXpoint

The big picture, why this type of NVM technology is needed

Server and Storage I/O trends

  • Memory is storage and storage is persistent memory
  • No such thing as a data or information recession, more data being create, processed and stored
  • Increased demand is also driving density along with convergence across server storage I/O resources
  • Larger amounts of data needing to be processed faster (large amounts of little data and big fast data)
  • Fast applications need more and faster processors, memory along with I/O interfaces
  • The best server or storage I/O is the one you do not need to do
  • The second best I/O is one with least impact or overhead
  • Data needs to be close to processing, processing needs to be close to the data (locality of reference)


Server Storage I/O memory hardware and software hierarchy along with technology tiers

What did Intel and Micron announce?

Intel SVP and General Manager Non-Volatile Memory solutions group Robert Crooke (Left) and Micron CEO D. Mark Durcan did the joint announcement presentation of 3D XPoint (webinar here). What was announced is the 3D XPoint technology jointly developed and manufactured by Intel and Micron which is a new form or category of NVM that can be used for both primary memory in servers, laptops, other computers among other uses, as well as for persistent data storage.


Robert Crooke (Left) and Mark Durcan (Right)

Summary of 3D XPoint announcement

  • New category of NVM memory for servers and storage
  • Joint development and manufacturing by Intel and Micron in Utah
  • Non volatile so can be used for storage or persistent server main memory
  • Allows NVM to scale with data, storage and processors performance
  • Leverages capabilities of both Intel and Micron who have collaborated in the past
  • Performance Intel and Micron claim up to 1000x faster vs. NAND flash
  • Availability persistent NVM compared to DRAM with better durability (life span) vs. NAND flash
  • Capacity densities about 10x better vs. traditional DRAM
  • Economics cost per bit between dram and nand (depending on packaging of resulting products)

What applications and products is 3D XPoint suited for?

In general, 3D XPoint should be able to be used for many of the same applications and associated products that current DRAM and NAND flash-based storage memories are used for. These range from IT and cloud or managed service provider data centers based applications and services, as well as consumer focused among many others.


3D XPoint enabling various applications

In general, applications or usage scenarios along with supporting products that can benefit from 3D XPoint include among others’. Applications that need larger amounts of main memory in a denser footprint such as in-memory databases, little and big data analytics, gaming, wave form analysis for security, copyright or other detection analysis, life sciences, high performance compute and high-productivity compute, energy, video and content severing among many others.

In addition, applications that need persistent main memory for resiliency, or to cut delays and impacts for planned or un-planned maintenance or having to wait for memories and caches to be warmed or re-populated after a server boot (or re-boot). 3D XPoint will also be useful for those applications that need faster read and write performance compared to current generations NAND flash for data storage. This means both existing and emerging applications as well as some that do not yet exist will benefit from 3D XPoint over time, like how today’s applications and others have benefited from DRAM used in Dual Inline Memory Module (DIMM) and NAND flash advances over the past several decades.

Where to read, watch and learn more

Storage I/O trends

Additional learning experiences along with common questions (and answers), as well as tips can be found in Software Defined Data Infrastructure Essentials book.

Software Defined Data Infrastructure Essentials Book SDDC

What This All Means

First, keep in mind that this is very early in the 3D XPoint technology evolution life-cycle and both DRAM and NAND flash will not be dead at least near term. Keep in mind that NAND flash appeared back in 1989 and only over the past several years has finally hit its mainstream adoption stride with plenty of market upside left. Continue reading Part II here and Part III here of this three-part series on Intel and Micron 3D XPoint along with more analysis and commentary.

Disclosure: Micron and Intel have been direct and/or indirect clients in the past via third-parties and partners, also I have bought and use some of their technologies direct and/or in-direct via their partners.

Ok, nuff said, for now.

Gs

Greg Schulz – Microsoft MVP Cloud and Data Center Management, VMware vExpert 2010-2017 (vSAN and vCloud). Author of Software Defined Data Infrastructure Essentials (CRC Press), as well as Cloud and Virtual Data Storage Networking (CRC Press), The Green and Virtual Data Center (CRC Press), Resilient Storage Networks (Elsevier) and twitter @storageio. Courteous comments are welcome for consideration. First published on https://storageioblog.com any reproduction in whole, in part, with changes to content, without source attribution under title or without permission is forbidden.

All Comments, (C) and (TM) belong to their owners/posters, Other content (C) Copyright 2006-2024 Server StorageIO and UnlimitedIO. All Rights Reserved. StorageIO is a registered Trade Mark (TM) of Server StorageIO.

Posted on

Server and Storage IO Memory: DRAM and nand flash

Storage I/O trends

DRAM, DIMM, DDR3, nand flash memory, SSD, stating what’s often assumed

Often what’s assumed is not always the case. For example in along with around server, storage and IO networking circles including virtual as well as cloud environments terms such as nand (Negated AND or NOT And) flash memory aka (Solid State Device or SSD), DRAM (Dynamic Random Access Memory), DDR3 (Double Data Rate 3) not to mention DIMM (Dual Inline Memory Module) get tossed around with the assumption everybody must know what they mean.

On the other hand, I find plenty of people who are not sure what those among other terms or things are, sometimes they are even embarrassed to ask, particular if they are a self-proclaimed expert.

So for those who need a refresh or primer, here you go, an excerpt from Chapter 7 (Servers – Physical, Virtual and Software) from my book "The Green and Virtual Data Center" (CRC Press) available at Amazon.com and other global venues in print and ebook formats.

7.2.2 Memory

Computers rely on some form of memory ranging from internal registers, local on-board processor Level 1 (L1) and Level 2 (L2) caches, random accessible memory (RAM), non-volatile RAM (NVRAM) or nand Flash (SSD) along with external disk storage. Memory, which includes external disk storage, is used for storing operating system software along with associated tools or utilities, application programs and data. Main memory or RAM, also known as dynamic RAM (DRAM) chips, is packaged in different ways with a common form being dual inline memory modules (DIMMs) for notebook or laptop, desktop PC and servers.

RAM main memory on a server is the fastest form of memory, second only to internal processor or chip based registers, L1, L2 or local memory. RAM and processor based memories are volatile and non-persistent in that when power is removed, the contents of memory are lost. As a result, some form of persistent memory is needed to keep programs and data when power is removed. Read only memory (ROM) and NVRAM are both persistent forms of memory in that their contents are not lost when power is removed. The amount of RAM that can be installed into a server will vary with specific architecture implementation and operating software being used. In addition to memory capacity and packaging format, the speed of memory is also important to be able to move data and programs quickly to avoid internal bottlenecks. Memory bandwidth performance increases with the width of the memory bus in bits and frequency in MHz. For example, moving 8 bytes on a 64 bit buss in parallel at the same time at 100MHz provides a theoretical 800MByte/sec speed.

To improve availability and increase the level of persistence, some servers include battery backed up RAM or cache to protect data in the event of a power loss. Another technique to protect memory data on some servers is memory mirroring where twice the amount of memory is installed and divided into two groups. Each group of memory has a copy of data being stored so that in the event of a memory failure beyond those correctable with standard parity and error correction code (ECC) no data is lost. In addition to being fast, RAM based memories are also more expensive and used in smaller quantities compared to external persistent memories such as magnetic hard disk drives, magnetic tape or optical based memory medias.

Memory diagram
Memory and Storage Pyramid

The above shows a tiered memory model that may look familiar as the bottom part is often expanded to show tiered storage. At the top of the memory pyramid is high-speed processor memory followed by RAM, ROM, NVRAM and FLASH along with many forms of external memory commonly called storage. More detail about tiered storage is covered in chapter 8 (Data Storage – Data Storage – Disk, Tape, Optical, and Memory). In addition to being slower and lower cost than RAM based memories, disk storage along with NVRAM and FLASH based memory devices are also persistent.

By being persistent, when power is removed, data is retained on the storage or memory device. Also shown in the above figure is that on a relative basis, less energy is used for power storage or memory at the bottom of the pyramid than for upper levels where performance increases. From a PCFE (Power, Cooling, Floor space, Economic) perspective, balancing memory and storage performance, availability, capacity and energy to a given function, quality of service and service level objective for a given cost needs to be kept in perspective and not considering simply the lowest cost for the most amount of memory or storage. In addition to gauging memory on capacity, other metrics include percent used, operating system page faults and page read/write operations along with memory swap activity as well memory errors.

Base 2 versus base 10 numbering systems can account for some storage capacity that appears to “missing” when real storage is compared to what is expected to be seen. Disk drive manufacturers use base 10 (decimal) to count bytes of data while memory chip, server and operating system vendors typically use base 2 (binary) to count bytes of data. This has led to confusion when comparing a disk drive base 10 GB with a chip memory base 2 GB of memory capacity, such as 1,000,000,000 (10^9) bytes versus 1,073,741,824 (2^30) bytes. Nomenclature based on the International System of Units uses MiB, GiB and TiB to denote million, billion and trillion bytes for base 2 numbering with base 10 using MB, TB and GB . Most vendors do document how many bytes, sometimes in both base 2 and base 10, as well as the number of 512 byte sectors supported on their storage devices and storage systems, though it might be in the small print.

Related more reading:
How much storage performance do you want vs. need?
Can RAID extend the life of nand flash SSD?
Can we get a side of context with them IOPS and other storage metrics?
SSD & Real Estate: Location, Location, Location
What is the best kind of IO? The one you do not have to do
SSD, flash and DRAM, DejaVu or something new?

Ok, nuff said (for now).

Cheers
Gs

Greg Schulz – Author Cloud and Virtual Data Storage Networking (CRC Press), The Green and Virtual Data Center (CRC Press) and Resilient Storage Networks (Elsevier).

All Comments, (C) and (TM) belong to their owners/posters, Other content (C) Copyright 2006-2024 Server StorageIO and UnlimitedIO LLC All Rights Reserved