Hybrid Memory Cube Angles for Exascale

By Michael Feldman

July 10, 2012

Computer memory is currently undergoing something of an identity crisis. For the past 8 years, multicore microprocessors have been creating a performance discontinuity, the so-called memory wall. It’s now fairly clear that this widening gap between compute and memory performance will not be solved with conventional dynamic random-access memory (DRAM) products. But there is one technology under development that aims to close that gap, and its first use case will likely be in the ethereal realm of supercomputing.

About a year and half ago, memory-maker Micron Technology came up with the Hybrid Memory Cube (HMC), a multi-chip module (MCM) device specifically designed to scale the memory wall. The goal was to offer a memory technology that matches the needs of core-happy CPUs and GPUs and do so in a way that is attractive to computer makers.

In a nutshell, HMC glues a logic control chip to a 3D memory stack, all of which are connected with Through Silicon Vias (TSVs). The technology promises not only to deliver an order of magnitude performance increase, but also to keep pace with future microprocessors as those designs continue to add cores. Micron claims a single HMC device can deliver 15 times the performance of today’s DDR3 modules and can do so with 70 percent less energy and in 90 percent less space. Latency is expected to decrease as well, although no specific claims are being made in that regard.

According to Dean Klein, VP of Micron’s Memory System Development, the problem with conventional DRAM technology is that they’ve pushed CMOS technology about as far as it’s going to go under the DDR model. Although DDR4 products are slated to ship before this end year, there is currently no DDR5 on the drawing board. That’s a problem, especially considering that DDR5 would probably be coming out toward the end of the decade, just when the first exascale supercomputers are expected to appear.

But even if DDR evolution is maintained through 2020, it would almost certainly fall short of the needs of exascale computing. Such machines are expected to require per-node memory bandwidth in excess of 500 terabytes/second. Klein says they just can’t boost the signal rates much more on the DDR design, and if they tried, power consumption would go in the wrong direction.

The HMC design gets around those limitations by going vertical and using the TSV technology to parallelize communication to the stack of memory chips, which enables much higher transfer rates. Bandwidth between the logic and the DRAM chips are projected to top a terabit per second (128 GB/second), which is much more in line with exascale needs.

Another important aspect of the design is that the interface abstracts the notion of reads and writes. That means a microprocessor’s memory controller doesn’t need to know about the underlying technology that stores the bits. So one could build an HMC device that was made up of DRAM or NAND flash, or even some combination of these technologies. That frees up the microprocessor and other peripheral devices from being locked into a particular memory type and, in general, should make system designs more flexible.

To move HMC beyond a science project, Micron put together a consortium and attracted key players, including competitors, to back the technology. Today the Hybrid Memory Cube Consortium consists of some of the industry’s heaviest hitters: Samsung, Microsoft, IBM, ARM, HP, IBM, Altera, Xilinx, Open-Silicon, and SK hynix. The group’s immediate goal is to develop a standard interface for the technology so that multiple manufacturers can build compliant HMC devices. The formal standard is due out later this year.

A key partner with Micron has been Intel, a vendor with a particular interest in high-performance memory. The chipmaker’s immediate motivation to support HMC is its Xeon line (including, soon, the manycore Xeon Phi), which is especially dependent on performant memory. In fact, without such memory, the value of high-end server chips is greatly diminished, since additional cores doesn’t translate into more performance for the end user. The relative success of future multicore and manycore processors will depend, to a large extent, on memory wall-busting technology.

Further out, Intel is looking at HMC as a technology to support its own aspirations to develop components for exascale supercomputers. Last year Intel helped Micron build an HMC prototype, which CTO Justin Rattner talked up at last September’s Intel Developer Forum. Although the chipmaker will presumably assist Micron if and when it starts churning out commercial silicon, neither company has offered a timeline for an HMC product launch. Klein did say that its prototype has been in the hands of select customers (HPC users and others) for several months, and their intent is to commercialize the technology.

And not just for high performance computing market. Although supercomputing has the greatest immediate need for such technology, other application areas, like networking, could also benefit greatly from HMC’s high bandwidth characteristics. And because of the promised power savings, even the high-volume mobile computing market is a potential target.

The biggest challenge for HMC is likely to be price. In particular, the use of TSV and 3D chip-stacking is in its infancy and by all accounts, will not come cheaply — at least not initially. And when you’re talking about 10PB of memory for an exascale machine or 1MB for a mobile phone, cost is a big consideration.

Other technologies like HP’s memristor, Magneto-resistive Random-Access Memory (MRAM), or Phase Change Memory (PCM) could come to the fore in time for the exascale era, but each one has its own challenges. As Klein notes, there is no holy grail of memory that encapsulates every desired attribute — high performance, low-cost, non-volatile, low-power, and infinite endurance.

The nice thing about HMC is that it can encapsulate DRAM as well as other memory technologies as they prove themselves. For the time being though, dynamic random-access memory will remain as the foundation of computer memory in the datacenter. “DRAM is certainly going to with us, at least until the end of the decade,” admits Klein. “We really don’t have a replace technology that looks as attractive.”

Subscribe to HPCwire's Weekly Update!

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

MLPerf Inference 4.0 Results Showcase GenAI; Nvidia Still Dominates

March 28, 2024

There were no startling surprises in the latest MLPerf Inference benchmark (4.0) results released yesterday. Two new workloads — Llama 2 and Stable Diffusion XL — were added to the benchmark suite as MLPerf continues Read more…

Q&A with Nvidia’s Chief of DGX Systems on the DGX-GB200 Rack-scale System

March 27, 2024

Pictures of Nvidia's new flagship mega-server, the DGX GB200, on the GTC show floor got favorable reactions on social media for the sheer amount of computing power it brings to artificial intelligence.  Nvidia's DGX Read more…

Call for Participation in Workshop on Potential NSF CISE Quantum Initiative

March 26, 2024

Editor’s Note: Next month there will be a workshop to discuss what a quantum initiative led by NSF’s Computer, Information Science and Engineering (CISE) directorate could entail. The details are posted below in a Ca Read more…

Waseda U. Researchers Reports New Quantum Algorithm for Speeding Optimization

March 25, 2024

Optimization problems cover a wide range of applications and are often cited as good candidates for quantum computing. However, the execution time for constrained combinatorial optimization applications on quantum device Read more…

NVLink: Faster Interconnects and Switches to Help Relieve Data Bottlenecks

March 25, 2024

Nvidia’s new Blackwell architecture may have stolen the show this week at the GPU Technology Conference in San Jose, California. But an emerging bottleneck at the network layer threatens to make bigger and brawnier pro Read more…

Who is David Blackwell?

March 22, 2024

During GTC24, co-founder and president of NVIDIA Jensen Huang unveiled the Blackwell GPU. This GPU itself is heavily optimized for AI work, boasting 192GB of HBM3E memory as well as the the ability to train 1 trillion pa Read more…

MLPerf Inference 4.0 Results Showcase GenAI; Nvidia Still Dominates

March 28, 2024

There were no startling surprises in the latest MLPerf Inference benchmark (4.0) results released yesterday. Two new workloads — Llama 2 and Stable Diffusion Read more…

Q&A with Nvidia’s Chief of DGX Systems on the DGX-GB200 Rack-scale System

March 27, 2024

Pictures of Nvidia's new flagship mega-server, the DGX GB200, on the GTC show floor got favorable reactions on social media for the sheer amount of computing po Read more…

NVLink: Faster Interconnects and Switches to Help Relieve Data Bottlenecks

March 25, 2024

Nvidia’s new Blackwell architecture may have stolen the show this week at the GPU Technology Conference in San Jose, California. But an emerging bottleneck at Read more…

Who is David Blackwell?

March 22, 2024

During GTC24, co-founder and president of NVIDIA Jensen Huang unveiled the Blackwell GPU. This GPU itself is heavily optimized for AI work, boasting 192GB of HB Read more…

Nvidia Looks to Accelerate GenAI Adoption with NIM

March 19, 2024

Today at the GPU Technology Conference, Nvidia launched a new offering aimed at helping customers quickly deploy their generative AI applications in a secure, s Read more…

The Generative AI Future Is Now, Nvidia’s Huang Says

March 19, 2024

We are in the early days of a transformative shift in how business gets done thanks to the advent of generative AI, according to Nvidia CEO and cofounder Jensen Read more…

Nvidia’s New Blackwell GPU Can Train AI Models with Trillions of Parameters

March 18, 2024

Nvidia's latest and fastest GPU, codenamed Blackwell, is here and will underpin the company's AI plans this year. The chip offers performance improvements from Read more…

Nvidia Showcases Quantum Cloud, Expanding Quantum Portfolio at GTC24

March 18, 2024

Nvidia’s barrage of quantum news at GTC24 this week includes new products, signature collaborations, and a new Nvidia Quantum Cloud for quantum developers. Wh Read more…

Alibaba Shuts Down its Quantum Computing Effort

November 30, 2023

In case you missed it, China’s e-commerce giant Alibaba has shut down its quantum computing research effort. It’s not entirely clear what drove the change. Read more…

Nvidia H100: Are 550,000 GPUs Enough for This Year?

August 17, 2023

The GPU Squeeze continues to place a premium on Nvidia H100 GPUs. In a recent Financial Times article, Nvidia reports that it expects to ship 550,000 of its lat Read more…

Shutterstock 1285747942

AMD’s Horsepower-packed MI300X GPU Beats Nvidia’s Upcoming H200

December 7, 2023

AMD and Nvidia are locked in an AI performance battle – much like the gaming GPU performance clash the companies have waged for decades. AMD has claimed it Read more…

DoD Takes a Long View of Quantum Computing

December 19, 2023

Given the large sums tied to expensive weapon systems – think $100-million-plus per F-35 fighter – it’s easy to forget the U.S. Department of Defense is a Read more…

Synopsys Eats Ansys: Does HPC Get Indigestion?

February 8, 2024

Recently, it was announced that Synopsys is buying HPC tool developer Ansys. Started in Pittsburgh, Pa., in 1970 as Swanson Analysis Systems, Inc. (SASI) by John Swanson (and eventually renamed), Ansys serves the CAE (Computer Aided Engineering)/multiphysics engineering simulation market. Read more…

Choosing the Right GPU for LLM Inference and Training

December 11, 2023

Accelerating the training and inference processes of deep learning models is crucial for unleashing their true potential and NVIDIA GPUs have emerged as a game- Read more…

Intel’s Server and PC Chip Development Will Blur After 2025

January 15, 2024

Intel's dealing with much more than chip rivals breathing down its neck; it is simultaneously integrating a bevy of new technologies such as chiplets, artificia Read more…

Baidu Exits Quantum, Closely Following Alibaba’s Earlier Move

January 5, 2024

Reuters reported this week that Baidu, China’s giant e-commerce and services provider, is exiting the quantum computing development arena. Reuters reported � Read more…

Leading Solution Providers

Contributors

Comparing NVIDIA A100 and NVIDIA L40S: Which GPU is Ideal for AI and Graphics-Intensive Workloads?

October 30, 2023

With long lead times for the NVIDIA H100 and A100 GPUs, many organizations are looking at the new NVIDIA L40S GPU, which it’s a new GPU optimized for AI and g Read more…

Shutterstock 1179408610

Google Addresses the Mysteries of Its Hypercomputer 

December 28, 2023

When Google launched its Hypercomputer earlier this month (December 2023), the first reaction was, "Say what?" It turns out that the Hypercomputer is Google's t Read more…

AMD MI3000A

How AMD May Get Across the CUDA Moat

October 5, 2023

When discussing GenAI, the term "GPU" almost always enters the conversation and the topic often moves toward performance and access. Interestingly, the word "GPU" is assumed to mean "Nvidia" products. (As an aside, the popular Nvidia hardware used in GenAI are not technically... Read more…

Shutterstock 1606064203

Meta’s Zuckerberg Puts Its AI Future in the Hands of 600,000 GPUs

January 25, 2024

In under two minutes, Meta's CEO, Mark Zuckerberg, laid out the company's AI plans, which included a plan to build an artificial intelligence system with the eq Read more…

Google Introduces ‘Hypercomputer’ to Its AI Infrastructure

December 11, 2023

Google ran out of monikers to describe its new AI system released on December 7. Supercomputer perhaps wasn't an apt description, so it settled on Hypercomputer Read more…

China Is All In on a RISC-V Future

January 8, 2024

The state of RISC-V in China was discussed in a recent report released by the Jamestown Foundation, a Washington, D.C.-based think tank. The report, entitled "E Read more…

Intel Won’t Have a Xeon Max Chip with New Emerald Rapids CPU

December 14, 2023

As expected, Intel officially announced its 5th generation Xeon server chips codenamed Emerald Rapids at an event in New York City, where the focus was really o Read more…

IBM Quantum Summit: Two New QPUs, Upgraded Qiskit, 10-year Roadmap and More

December 4, 2023

IBM kicks off its annual Quantum Summit today and will announce a broad range of advances including its much-anticipated 1121-qubit Condor QPU, a smaller 133-qu Read more…

  • arrow
  • Click Here for More Headlines
  • arrow
HPCwire