Storage with the speed of memory? XPoint, XPoint, that's our plan

And if it can't do it, (other) ReRAM can


Analysis Since the virtual dawn of computing, storage – where data puts its feet up when it's at home – has been massively slower than memory, where data puts on its trainers and goes for a quick run.

That massive access speed gap has been getting narrower and narrower with each storage technology advance: paper tape, magnetic tape, disk drives, and flash memory. Now it is set to shrink again with even faster non-volatile solid-state memory, led by 3D XPoint.

This will have far-reaching effects on servers and on storage as system hardware and software change to take advantage of this much faster – near-memory speed in fact – storage.

Memory contents are volatile or ephemeral, in that they are lost when power is turned off. Storage contents, be they held in NAND, on disk or tape, are persistent or non-volatile because they stay unharmed by power being switched off to the device.

Paradoxically, storage media that doesn’t move is faster than storage media that does, with solid state, non-volatile memory or NAND being much faster to access than rotating disk and streaming tape.

How much faster?

Nanoseconds and microseconds

A memory access can take 200 nanoseconds*. This is slow by CPU cache-access speed standards, with a level 1 processor cache access taking half a nanosecond, 400 times faster, and a level 2 cache access needing seven nanoseconds.

A disk-drive seek, the time for the read-write head to move across the platter and reach the right track area, can take 10 million nanoseconds – that’s 10 milliseconds or 10,000 microseconds. Let’s say it’s slightly more than two orders of magnitude slower than a memory access.

A solid state drive (SSD) is vastly faster than this. Micron’s latest 9100, implemented as an NVMe protocol PCIe bus card, has a 120 microsecond read-access latency and a 30 microsecond write-access latency. That’s at least an order of magnitude faster than disk but 150 times slower than a DRAM access for write access and 600 times slower for a read access.

Why the asymmetry between read and write access? We understand it is because writes are cached in an SSD’s DRAM buffer whereas reads have to access, of course, the slower NAND.

Although NAND is much faster than disk it is still a great deal slower than memory and the essential coding rule of thumb that says IO is slower than a memory access still stands. It means coders, keen to shorten application run times, will try to restrict the amount of IO that’s carried out, and server/storage system engineers will try to use NAND caching to get over disk data delays, and DRAM caching to get over NAND access delays.

Servers with multi-socket, multi-core CPUs are inherently liable to be in IO wait states, as the storage subsystems, hundreds of times slower than memory at best, struggle to read and write data fast enough, meaning the processors wait, doing nothing, while turgidly slow storage does its thing.

The gaping access time gulf

If we treat an L1 cache reference as taking one second, then, on that measurement scale, a disk access can take six years, four months, 19 hours, 33 minutes, and 20 seconds. Reading from Micron’s 9100 SSD is much faster though, taking two days, 18 hours and 40 minutes. That’s good by disk standards, but it’s still tortuously slow by DRAM access speed standards – which take six mins and 40 secs.

This great gaping gulf in access speed between volatile memory and persistent storage promises to be narrowed by Intel and Micron’s 3D XPoint memory, a new form of non-volatile, solid state storage - if it lives up to expectations.

3d_XPOint_structure

XPoint is made from Phase Change Memory material, a chalcogenide glass, and it has electricity applied to it, causing a bulk change in some aspect of its state that causes a resistance-level change. This can be reversed, leading the resistance levels signalling a binary 1 or zero. The XPoint die has two layers and a crossbar design, hence its full 3D CrossPoint name.

The bulk change is said to be different from the amorphous-to-crystalline state change used in Phase Change Memory (PCM) technology., which also modifies the material’s resistance level

The early marketing messages from Intel and Micron talked about it being 1,000 times faster than flash, as well as having 1,000 times more endurance, and a cost expected to be more than flash but less than DRAM.

Jim Handy of Objective Analysis charted XPoint’s position on a log-scale graph;

Handy_XPOint_log_scale_chart

The chart plots media in a 2D space defined by cost/GB and bandwidth, and we can see 3D XPoint sitting between NAND and DRAM.

XPoint details

Unlike NAND, it is byte-addressable, not block-addressable. Intel says it will introduce Optane brand XPoint SSDs (meaning NVMe-access drives) and XPoint DIMMs, meaning XPoint chips mounted on DIMMs to access the memory bus directly, instead of the relatively slower PCIe bus used by the NVMe XPoint drives.

An Intel presenter, Frank Hady, Intel Fellow and Chief 3D XPoint Storage Architect, provided some Optane numbers at the Seventh Annual Non-Volatile Memories Workshop 2016 at UC San Diego recently.

We learnt that Optane had this kind of profile:

  • 20nm process
  • SLC (1 bit/cell)
  • 7 microsec latency, or 7,000 nanoseconds
  • 78,500 (70:30 random) read/write IOPS
  • NVMe interface

XPoint’s bandwidth is not clear at this point.

If we construct a latency table looking at memory and storage media, from L1 cache to disk, and including XPoint, this is what we see:

Latency table

With a seven-microsecond latency XPoint is only 35 times slower to access than DRAM. This is a lot better than NVMe NAND, with Micron’s 9100 being 150 times slower than DRAM for writes and 600 times slower for reads.

If we chart five items from the table horizontally to show this, here is the result:

Latency chart

It’s obvious that XPoint is much closer to memory latency than NAND. XPoint can be used as a caching technology for NAND (or disk), in which case not much changes – a cache is a cache. Or it can be used as a form of memory, meaning that applications would execute on data fetched directly from XPoint instead of, or as well as DRAM.

XPoint chips mounted on DIMMs can be expected to have even shorter latencies than seven microseconds, and this is likely to be the XPoint storage-class memory (SCM) form factor.

System and software changes

With SCM applications won’t need to execute write IOs to get data into persistent storage. A memory-level (zero-copy) operation moving data into XPoint will do that. This is just one of the changes that systems and software will have to take on board when XPoint is treated as storage-class memory, meaning persistent memory.

Others include:

  • Development of file systems that are aware of persistent memory
  • Operating system support for storage-class memory
  • Processors designed to use hybrid DRAM and XPoint memory.

We’re bound to see Linux having its code extended to support XPoint storage-class memory.

Microsoft is working on adding XPoint storage-class memory support into Windows. It will provide zero-copy access and have Direct Access Storage volumes, known as DAX volumes; presumably the DAS acronym is so widely understood to mean disk drives and SSDS connected directly to a server by SATA or SAS links as to be unusable for XPoint Direct Access Storage. DAX volumes have memory-mapped files and applications can access them without going through the current IO stack.

That stack of code is based on disk drive era assumptions in which IO stack traversal time was such a small proportion of the overall IO time that it on its own did not matter. But, with XPoint access time at the single digit microsecond level, stack efficiency becomes much more important.

There will be specific SCM drivers to help ensure this, and there are more details in this video:

Youtube Video

Microsoft NVM presentation at the NVM Summit 2016 - Storage Class Memory Support in the Windows OS. Click image to go to video.

XPoint effects on servers and storage

XPoint-equipped servers will suffer much less from being IO-bound. Their caching will be more efficient. Ones using SCM will be able to have larger application working sets in memory and spend less time in writing data.

Other things being equal, XPoint-equipped servers should be able to support more VMs or containers when XPoint caching is used, and many more VMs or containers when XPoint NVMe SSDs are used, and more again when XPoint SCM is used.

Storage arrays will be able to respond faster when using XPoint caching, and faster again, overall, when using XPoint SSDs. They can be expected to use an NVMe fabric link with remote direct memory access, and function as shared XPoint direct-attached storage.

Fusion-io tried to introduce storage-class memory with its original PCIe flash cards but the company was too small to get an SCM-level changes into operating systems. The combined Intel and Micron pairing is certainly large enough for that, with both companies well aware of the need for XPoint-supporting ecosystems of hardware and software suppliers.

If, and it is most certainly an “if” still, XPoint delivers on its potential then we are going to see the virtual elimination of SAS and SATA-type storage IO waits and, with SCM, a substantial increase in overall server efficiency.

Should XPoint under-deliver, then other resistive RAM (ReRAM) technologies may step forward. We know SanDisk and Toshiba are working in that area but none of them are as ready as XPoint for prime time. Let’s hope it lives up to its extraordinary promise and delivers the XPoint goods. ®

* All the latency numbers used here are as accurate as I can get them but there can be a lot of variance so treat them as indicative rather than absolute.


Other stories you might like

  • Chip shortage forces temporary Raspberry Pi 4 price rise for the first time

    Ten-buck increase for 2GB model 'not here to stay' says Upton

    The price of a 2GB Raspberry Pi 4 single-board computer is going up $10, and its supply is expected to be capped at seven million devices this year due to the ongoing global chip shortage.

    Demand for components is outstripping manufacturing capacity at the moment; pre-pandemic, assembly lines were being red-lined as cloud giants and others snapped up parts fresh out of the fabs, and the COVID-19 coronavirus outbreak really threw a spanner in the works, so to speak, exacerbating the situation.

    Everything from cars to smartphones have been affected by semiconductor supply constraints, including Raspberry Pis, it appears. Stock is especially tight for the Raspberry Pi Zero and the 2GB Raspberry Pi 4 models, we're told. As the semiconductor crunch shows no signs of letting up, the Raspberry Pi project is going to bump up the price for one particular model.

    Continue reading
  • Uncle Sam to clip wings of Pegasus-like spyware – sorry, 'intrusion software' – with proposed export controls

    Surveillance tech faces trade limits as America syncs policy with treaty obligations

    More than six years after proposing export restrictions on "intrusion software," the US Commerce Department's Bureau of Industry and Security (BIS) has formulated a rule that it believes balances the latitude required to investigate cyber threats with the need to limit dangerous code.

    The BIS on Wednesday announced an interim final rule that defines when an export license will be required to distribute what is basically commercial spyware, in order to align US policy with the 1996 Wassenaar Arrangement, an international arms control regime.

    The rule [PDF] – which spans 65 pages – aims to prevent the distribution of surveillance tools, like NSO Group's Pegasus, to countries subject to arms controls, like China and Russia, while allowing legitimate security research and transactions to continue. Made available for public comment over the next 45 days, the rule is scheduled to be finalized in 90 days.

    Continue reading
  • Global IT spending to hit $4.5 trillion in 2022, says Gartner

    The future's bright, and expensive

    Corporate technology soothsayer Gartner is forecasting worldwide IT spending will hit $4.5tr in 2022, up 5.5 per cent from 2021.

    The strongest growth is set to come from enterprise software, which the analyst firm expects to increase by 11.5 per cent in 2022 to reach a global spending level of £670bn. Growth has fallen slightly, though. In 2021 it was 13.6 per cent for this market segment. The increase was driven by infrastructure software spending, which outpaced application software spending.

    The largest chunk of IT spending is set to remain communication services, which will reach £1.48tr next year, after modest growth of 2.1 per cent. The next largest category is IT services, which is set to grow by 8.9 per cent to reach $1.29tr over the next year, according to the analysts.

    Continue reading
  • Memory maker Micron moots $150bn mega manufacturing moneybag

    AI and 5G to fuel demand for new plants and R&D

    Chip giant Micron has announced a $150bn global investment plan designed to support manufacturing and research over the next decade.

    The memory maker said it would include expansion of its fabrication facilities to help meet demand.

    As well as chip shortages due to COVID-19 disruption, the $21bn-revenue company said it wanted to take advantage of the fact memory and storage accounts for around 30 per cent of the global semiconductor industry today.

    Continue reading
  • China to allow overseas investment in VPNs but Beijing keeps control of the generally discouraged tech

    Foreign ownership capped at 50%

    After years of restricting the use and ownership of VPNs, Beijing has agreed to let foreign entities hold up to a 50 per cent stake in domestic VPN companies.

    China has simultaneously a huge market and strict rules for VPNs as the country's Great Firewall attempts to keep its residents out of what it deems undesirable content and influence, such as Facebook or international news outlets.

    And while VPN technology is not illegal per se (it's just not practical for multinationals and other entities), users need a licence to operate one.

    Continue reading
  • Microsoft unveils Android apps for Windows 11 (for US users only)

    Windows Insiders get their hands on the Windows Subsystem for Android

    Microsoft has further teased the arrival of the Windows Subsystem for Android by detailing how the platform will work via a newly published document for Windows Insiders.

    The document, spotted by inveterate Microsoft prodder "WalkingCat" makes for interesting reading for developers keen to make their applications work in the Windows Subsystem for Android (WSA).

    WSA itself comprises the Android OS based on the Android Open Source Project 1.1 and, like the Windows Subsystem for Linux, runs in a virtual machine.

    Continue reading
  • Software Freedom Conservancy sues TV maker Vizio for GPL infringement

    Companies using GPL software should meet their obligations, lawsuit says

    The Software Freedom Conservancy (SFC), a non-profit which supports and defends free software, has taken legal action against Californian TV manufacturer Vizio Inc, claiming "repeated failures to fulfill even the basic requirements of the General Public License (GPL)."

    Member projects of the SFC include the Debian Copyright Aggregation Project, BusyBox, Git, GPL Compliance Project for Linux Developers, Homebrew, Mercurial, OpenWrt, phpMyAdmin, QEMU, Samba, Selenium, Wine, and many more.

    The GPL Compliance Project is described as "comprised of copyright holders in the kernel, Linux, who have contributed to Linux under its license, the GPLv2. These copyright holders have formally asked Conservancy to engage in compliance efforts for their copyrights in the Linux kernel."

    Continue reading
  • DRAM, it stacks up: SK hynix rolls out 819GB/s HBM3 tech

    Kit using the chips to appear next year at the earliest

    Korean DRAM fabber SK hynix has developed an HBM3 DRAM chip operating at 819GB/sec.

    HBM3 (High Bandwidth Memory 3) is a third generation of the HBM architecture which stacks DRAM chips one above another, connects them by vertical current-carrying holes called Through Silicon Vias (TSVs) to a base interposer board, via connecting micro-bumps, upon which is fastened a processor that accesses the data in the DRAM chip faster than it would through the traditional CPU socket interface.

    Seon-yong Cha, SK hynix's senior vice president for DRAM development, said: "Since its launch of the world's first HBM DRAM, SK hynix has succeeded in developing the industry's first HBM3 after leading the HBM2E market. We will continue our efforts to solidify our leadership in the premium memory market."

    Continue reading

Biting the hand that feeds IT © 1998–2021