Researchers Challenge NAND Flash with Phase Change Memory

By Michael Feldman

June 6, 2011

A team of researchers at the University of California, San Diego (UCSD) has built a solid state storage system that they claim outperforms state-of-the-art flash memory products. The new system, know as Moneta, uses phase change memory, a technology that some predict will replace the NAND flash memory used in nearly every solid state drive (SSD) today. The UCSD prototype will be on display this week at the Design Automation Conference in San Diego.

Moneta, which is the Latin name for the goddess of memory, uses phase change memory (PCM) module from Micron Technology as the storage medium and employs Xilinx FPGAs as memory controllers. The system DIMM, referred to as Onyx, is a 640 MB device made up of 40 of Micron’s 16 MB PCM (P8P) chips.* The entire system houses 16 of these DIMMs, yielding 10 GB of raw storage: 8 GB usable and 2 GB for error correction and metadata. The card connects to a Linux host via a PCIe x8 link. The prototype and its performance characteristics are described in detail here.

Like NAND flash memory, phase change memory is non-volatile, but it’s a total different digital animal. PCM is based on a metal alloy called chalcogenide and works, as its name suggests, by changing the state of the alloy. To write data, electrical current is applied to generate heat, which switches the alloy between its crystalline and its amorphous state. To read the data back, a smaller current is applied to determine which state the alloy is in.

PCM has several significant advantages over NAND. To begin with, it is inherently more robust than NAND, having an average endurance of 100,000,000 write cycles as compared to 100,000 for enterprise-class SLC NAND. That means the wear leveling algorithm can be much simpler, resulting in less software and memory overhead. PCM is also byte addressable, which means it is much more efficient than NAND at accessing smaller chunks of data, again reducing system complexity. Finally, PCM is just faster. The bits can be switched quickly, and can be flipped in place, avoiding the laborious block erase-and-write cycle required by NAND.

A longer term advantage cited for PCM is that it will scale better than NAND (or even DRAM) as process geometries shrink below 20nm. This has yet to be proven, but it has certainly encouraged companies like Micron and Intel to invest heavily in the technology.

The PCM hardware is only half of the Moneta story though. The system’s real value is contained in the software, which has been tuned for the higher performing PCM. Specifically, they developed a low-level block driver for Moneta that bypasses the Linux I/O scheduler, such that throughput and latency are optimized.

“We’ve found that you can build a much faster storage device, but in order to really make use of it, you have to change the software that manages it as well. Storage systems have evolved over the last 40 years to cater to disks, and disks are very, very slow,” said Steven Swanson, professor of Computer Science and Engineering at UCSD and director of its Non-Volatile Systems Lab (NVSL). “Designing storage systems that can fully leverage technologies like PCM requires rethinking almost every aspect of how a computer system’s software manages and accesses storage. Moneta gives us a window into the future of what computer storage systems are going to look like, and gives us the opportunity now to rethink how we design computer systems in response.”

To prove the technology’s worth, the UCSD team compared Moneta to a well known NAND flash product, in this case, an 80GB Fusion-io ioDrive. However, being a first-generation prototype, the results are somewhat of a mixed bag.

For 512-byte block accesses, Moneta can read data at 327 MB/sec and write at 91 MB/sec; which is between two and seven times faster than the Fusion-io ioDrive. For large data blocks (32K and up), Moneta can read data at 1.1 GB/sec and write data at 371 MB/sec. Here the PCM system is still outrunning the ioDrive for reads, but is actually somewhat slower in writes. The researchers attribute this to the ioDrive’s better aggregate bandwidth associated with its larger memory capacity, all of which is optimized for large block sizes.

Keep in mind that Fusion-io also offers PCIe flash devices with much more capacity and associated bandwidth than the version the UCSD researchers benchmarked. For example, Fusion-io claims 1.5 GB/sec in both reads and writes for its 320 GB ioDrive Duo with 64K block sizes, and 6.0 GB/sec in reads and 4.4 GB/sec in writes for its giant 5TB ioDrive Octal. Obviously size matters.

Other NAND SSD vendors are pushing the capacity/bandwidth envelop as well. Micron’s 700GB P320h flash drive delivers 3 GB/sec for reading and 2 GB/sec for writing (128K blocks); the latest 900GB RamSan-70 card for Texas Memory Systems manages 2 GB/sec reads and 1.4 GB/sec writes; and Virident Systems’ 800GB tachIOn drive tops out at 1.44 GB/sec and 1.2 GB/sec for reads and writes, respectively.

Of course, I/O bandwidth is just a single criteria. Performance under various loads and conditions (like when a drive becomes full) is also important. Cost, latency, longevity, power consumption are other factors that need to be considered.

In the case of power consumption, the UCSD researchers point out that because of the less idiosyncratic behavior of PCM compared to NAND, the software driver is simpler, thus there is a reduction in CPU usage. The researchers say that their Moneta prototype spends between 20 and 50 percent less CPU time performing I/O operations for small requests. That frees up the CPU to do application work and reduces the overall power consumption of the storage system.

The bottom line is that if Moneta can be scaled up to larger memory capacities and aggregate bandwidth on a single PCIe device, its natural advantage in read performance, small (i.e., random) write performance, and endurance will give the NAND-based SSD makers something to think about. The UCSD researchers contend that a PCM-based storage system will be of particular value in applications like high-performance caching systems and key-value stores, that require high performance reads and small writes. For example, applications that do a lot of random I/O, like large graph computations, are especially suited to Moneta-type architectures.

The UCSD team plans to upgrade Moneta in the next six to nine months, using denser and higher performing PCM devices. No PCM roadmap from Micron or anyone else has been made public, but the researchers say that the Moneta technology could be commercially available in just a few years.

*The original version of this story erroneously reported the DIMM and P8P storage specifications in GB rather than MB. –Editor

Subscribe to HPCwire's Weekly Update!

Be the most informed person in the room! Stay ahead of the tech trends with industy updates delivered to you every week!

Messina Update: The U.S. Path to Exascale in 16 Slides

April 26, 2017

Paul Messina, director of the U.S. Exascale Computing Project, provided a wide-ranging review of ECP’s evolving plans last week at the HPC User Forum. Read more…

By John Russell

IBM, Nvidia, Stone Ridge Claim Gas & Oil Simulation Record

April 25, 2017

IBM, Nvidia, and Stone Ridge Technology today reported setting the performance record for a “billion cell” oil and gas reservoir simulation. Read more…

By John Russell

ASC17 Makes Splash at Wuxi Supercomputing Center

April 24, 2017

A record-breaking twenty student teams plus scores of company representatives, media professionals, staff and student volunteers transformed a formerly empty hall inside the Wuxi Supercomputing Center into a bustling hub of HPC activity, kicking off day one of 2017 Asia Student Supercomputer Challenge (ASC17). Read more…

By Tiffany Trader

Groq This: New AI Chips to Give GPUs a Run for Deep Learning Money

April 24, 2017

CPUs and GPUs, move over. Thanks to recent revelations surrounding Google’s new Tensor Processing Unit (TPU), the computing world appears to be on the cusp of a new generation of chips designed specifically for deep learning workloads. Read more…

By Alex Woodie

HPE Extreme Performance Solutions

Remote Visualization Optimizing Life Sciences Operations and Care Delivery

As patients continually demand a better quality of care and increasingly complex workloads challenge healthcare organizations to innovate, investing in the right technologies is key to ensuring growth and success. Read more…

Musk’s Latest Startup Eyes Brain-Computer Links

April 21, 2017

Elon Musk, the auto and space entrepreneur and severe critic of artificial intelligence, is forming a new venture that reportedly will seek to develop an interface between the human brain and computers. Read more…

By George Leopold

MIT Mathematician Spins Up 220,000-Core Google Compute Cluster

April 21, 2017

On Thursday, Google announced that MIT math professor and computational number theorist Andrew V. Sutherland had set a record for the largest Google Compute Engine (GCE) job. Sutherland ran the massive mathematics workload on 220,000 GCE cores using preemptible virtual machine instances. Read more…

By Tiffany Trader

NERSC Cori Shows the World How Many-Cores for the Masses Works

April 21, 2017

As its mission, the high performance computing center for the U.S. Department of Energy Office of Science, NERSC (the National Energy Research Supercomputer Center), supports a broad spectrum of forefront scientific research across diverse areas that includes climate, material science, chemistry, fusion energy, high-energy physics and many others. Read more…

By Rob Farber

Nvidia P100 Shows 1.3-2.3x Speedup Over K80 GPU on Financial Apps

April 20, 2017

When it comes to the true performance of the latest silicon, every end user knows that the best processor is the one that works best for their application. Read more…

By Tiffany Trader

Messina Update: The U.S. Path to Exascale in 16 Slides

April 26, 2017

Paul Messina, director of the U.S. Exascale Computing Project, provided a wide-ranging review of ECP’s evolving plans last week at the HPC User Forum. Read more…

By John Russell

ASC17 Makes Splash at Wuxi Supercomputing Center

April 24, 2017

A record-breaking twenty student teams plus scores of company representatives, media professionals, staff and student volunteers transformed a formerly empty hall inside the Wuxi Supercomputing Center into a bustling hub of HPC activity, kicking off day one of 2017 Asia Student Supercomputer Challenge (ASC17). Read more…

By Tiffany Trader

Groq This: New AI Chips to Give GPUs a Run for Deep Learning Money

April 24, 2017

CPUs and GPUs, move over. Thanks to recent revelations surrounding Google’s new Tensor Processing Unit (TPU), the computing world appears to be on the cusp of a new generation of chips designed specifically for deep learning workloads. Read more…

By Alex Woodie

NERSC Cori Shows the World How Many-Cores for the Masses Works

April 21, 2017

As its mission, the high performance computing center for the U.S. Department of Energy Office of Science, NERSC (the National Energy Research Supercomputer Center), supports a broad spectrum of forefront scientific research across diverse areas that includes climate, material science, chemistry, fusion energy, high-energy physics and many others. Read more…

By Rob Farber

Hyperion (IDC) Paints a Bullish Picture of HPC Future

April 20, 2017

Hyperion Research – formerly IDC’s HPC group – yesterday painted a fascinating and complicated portrait of the HPC community’s health and prospects at the HPC User Forum held in Albuquerque, NM. HPC sales are up and growing ($22 billion, all HPC segments, 2016). Read more…

By John Russell

Knights Landing Processor with Omni-Path Makes Cloud Debut

April 18, 2017

HPC cloud specialist Rescale is partnering with Intel and HPC resource provider R Systems to offer first-ever cloud access to Xeon Phi "Knights Landing" processors. The infrastructure is based on the 68-core Intel Knights Landing processor with integrated Omni-Path fabric (the 7250F Xeon Phi). Read more…

By Tiffany Trader

CERN openlab Explores New CPU/FPGA Processing Solutions

April 14, 2017

Through a CERN openlab project known as the ‘High-Throughput Computing Collaboration,’ researchers are investigating the use of various Intel technologies in data filtering and data acquisition systems. Read more…

By Linda Barney

DOE Supercomputer Achieves Record 45-Qubit Quantum Simulation

April 13, 2017

In order to simulate larger and larger quantum systems and usher in an age of “quantum supremacy,” researchers are stretching the limits of today’s most advanced supercomputers. Read more…

By Tiffany Trader

Google Pulls Back the Covers on Its First Machine Learning Chip

April 6, 2017

This week Google released a report detailing the design and performance characteristics of the Tensor Processing Unit (TPU), its custom ASIC for the inference phase of neural networks (NN). Read more…

By Tiffany Trader

Quantum Bits: D-Wave and VW; Google Quantum Lab; IBM Expands Access

March 21, 2017

For a technology that’s usually characterized as far off and in a distant galaxy, quantum computing has been steadily picking up steam. Read more…

By John Russell

Trump Budget Targets NIH, DOE, and EPA; No Mention of NSF

March 16, 2017

President Trump’s proposed U.S. fiscal 2018 budget issued today sharply cuts science spending while bolstering military spending as he promised during the campaign. Read more…

By John Russell

HPC Compiler Company PathScale Seeks Life Raft

March 23, 2017

HPCwire has learned that HPC compiler company PathScale has fallen on difficult times and is asking the community for help or actively seeking a buyer for its assets. Read more…

By Tiffany Trader

Nvidia Responds to Google TPU Benchmarking

April 10, 2017

Nvidia highlights strengths of its newest GPU silicon in response to Google's report on the performance and energy advantages of its custom tensor processor. Read more…

By Tiffany Trader

CPU-based Visualization Positions for Exascale Supercomputing

March 16, 2017

In this contributed perspective piece, Intel’s Jim Jeffers makes the case that CPU-based visualization is now widely adopted and as such is no longer a contrarian view, but is rather an exascale requirement. Read more…

By Jim Jeffers, Principal Engineer and Engineering Leader, Intel

For IBM/OpenPOWER: Success in 2017 = (Volume) Sales

January 11, 2017

To a large degree IBM and the OpenPOWER Foundation have done what they said they would – assembling a substantial and growing ecosystem and bringing Power-based products to market, all in about three years. Read more…

By John Russell

TSUBAME3.0 Points to Future HPE Pascal-NVLink-OPA Server

February 17, 2017

Since our initial coverage of the TSUBAME3.0 supercomputer yesterday, more details have come to light on this innovative project. Of particular interest is a new board design for NVLink-equipped Pascal P100 GPUs that will create another entrant to the space currently occupied by Nvidia's DGX-1 system, IBM's "Minsky" platform and the Supermicro SuperServer (1028GQ-TXR). Read more…

By Tiffany Trader

Leading Solution Providers

Tokyo Tech’s TSUBAME3.0 Will Be First HPE-SGI Super

February 16, 2017

In a press event Friday afternoon local time in Japan, Tokyo Institute of Technology (Tokyo Tech) announced its plans for the TSUBAME3.0 supercomputer, which will be Japan’s “fastest AI supercomputer,” Read more…

By Tiffany Trader

Is Liquid Cooling Ready to Go Mainstream?

February 13, 2017

Lost in the frenzy of SC16 was a substantial rise in the number of vendors showing server oriented liquid cooling technologies. Three decades ago liquid cooling was pretty much the exclusive realm of the Cray-2 and IBM mainframe class products. That’s changing. We are now seeing an emergence of x86 class server products with exotic plumbing technology ranging from Direct-to-Chip to servers and storage completely immersed in a dielectric fluid. Read more…

By Steve Campbell

IBM Wants to be “Red Hat” of Deep Learning

January 26, 2017

IBM today announced the addition of TensorFlow and Chainer deep learning frameworks to its PowerAI suite of deep learning tools, which already includes popular offerings such as Caffe, Theano, and Torch. Read more…

By John Russell

BioTeam’s Berman Charts 2017 HPC Trends in Life Sciences

January 4, 2017

Twenty years ago high performance computing was nearly absent from life sciences. Today it’s used throughout life sciences and biomedical research. Genomics and the data deluge from modern lab instruments are the main drivers, but so is the longer-term desire to perform predictive simulation in support of Precision Medicine (PM). There’s even a specialized life sciences supercomputer, ‘Anton’ from D.E. Shaw Research, and the Pittsburgh Supercomputing Center is standing up its second Anton 2 and actively soliciting project proposals. There’s a lot going on. Read more…

By John Russell

HPC Startup Advances Auto-Parallelization’s Promise

January 23, 2017

The shift from single core to multicore hardware has made finding parallelism in codes more important than ever, but that hasn’t made the task of parallel programming any easier. Read more…

By Tiffany Trader

Facebook Open Sources Caffe2; Nvidia, Intel Rush to Optimize

April 18, 2017

From its F8 developer conference in San Jose, Calif., today, Facebook announced Caffe2, a new open-source, cross-platform framework for deep learning. Caffe2 is the successor to Caffe, the deep learning framework developed by Berkeley AI Research and community contributors. Read more…

By Tiffany Trader

HPC Technique Propels Deep Learning at Scale

February 21, 2017

Researchers from Baidu’s Silicon Valley AI Lab (SVAIL) have adapted a well-known HPC communication technique to boost the speed and scale of their neural network training and now they are sharing their implementation with the larger deep learning community. Read more…

By Tiffany Trader

US Supercomputing Leaders Tackle the China Question

March 15, 2017

Joint DOE-NSA report responds to the increased global pressures impacting the competitiveness of U.S. supercomputing. Read more…

By Tiffany Trader

  • arrow
  • Click Here for More Headlines
  • arrow
Share This