Startup Takes Aim at Performance-Killing Vibration in Datacenter

By Michael Feldman

January 19, 2010

At SC09 last November in Portland, Oregon, I met a fellow named Larry Gordon who related this story:

“I’m riding a train to San Francisco, and I’m reading a newspaper. The train’s vibrating and my arms are vibrating and the newspaper’s vibrating. I can still read the newspaper, but it takes more concentration, so I don’t read as fast as I normally do. It’s just more work to focus on a moving target.”

Makes sense. But why is he telling me this story at a conference about supercomputing? As it turns out, Larry Gordon is the vice president of marketing for a startup called Green Platform Corporation, and the scenario he’s describing on the train is analogous to the problem that hard disk drives encounter when reading and writing data in a datacenter environment — a problem, he believes, the company has solved.

According to Gus Malek-Madani, Green Platform’s founder and CEO, vibration in the datacenter forces a disk drive to work extra hard to perform reads and writes, degrading performance. And not just a little. Independent tests indicate as much as two-thirds of I/O throughput is being lost due to vibration. Worse yet, as performance suffers, the disk has to do additional work to access data, slowing down the entire compute system and raising power consumption proportionally. Tests indicate that as vibration increases, energy use can more than double for a given job.

The source of all this vibration includes the hard drives themselves (the mechanical motion of the spinning platter and actuator arm movement), cooling fans, and chiller pumps, as well as the cumulative vibration reverberating back and forth between racks. While most solid state components are relatively immune to vibration, spinning disks are not. “The performance of these drives is incredibly damaged by vibration,” Malek-Madani told me.

Malek-Madani has made a 25-year career of using carbon fiber to battle the forces of vibration. In fact, Green Platform is an offshoot of another company of his called Composite Products, which offers anti-vibration racks and shelves for the high-end audio/video and scientific markets. Malek-Madani got the idea in his head that computer disks might also benefit from vibrational dampening, and when some initial tests (on a desktop PC) proved out, he decided to pursue the solution that would make commercial sense: disk storage in datacenters.

The solution he and his new company came up with is rather simple: build whole racks out of carbon fiber material and other composites to dampen datacenter vibration. (Carbon fiber is the same material used in aircraft and golf clubs for its superior strength, weight advantages and vibration damping characteristics.) The Green Platform rack, called the AVP-1000 (prototype pictured below) is designed to replace the traditional steel rack with a carbon fiber one. The AVP, which stands for Anti-Vibration Platform, is built as a standard 19-inch enclosure and is compatible with typical storage server form factors.

Up until now, hard drive manufacturers were concerned with vibration, but almost exclusively from a reliability point of view. It’s well understood that too much internal vibration inside the drive mechanism risks a disk failure. When vibration is detected, the actuator arm can freeze or slow down in order to prevent a head crash. The idea here is to protect the data media at the expense of a temporary performance drop. But the design treats the symptom, not the cause, and does nothing to address the cumulative effects of vibration in the environment. Green Platform treats the effects of normal and ongoing levels of ambient vibration in datacenters as well as shocks from seismic causes, rolling dollies, elevators, subways, construction, etc.

All of this is part of the more general problem of hard disk technology, which is stuck between the digital and mechanical worlds. As the density of disk media has become greater, more bits per square inch can be stored on the media, making it increasingly difficult for the actuator arm and read/write head to find a specific piece of information. Thus, while storage capacities have increased, access times have not kept up.

Likewise, the effect of mechanical vibration also works against the trend of denser media. Returning to Gordon’s original analogy of reading a newspaper on a train, it’s as if the read/write head at the end of the actuator must adjust to ever-smaller fonts. The idea that datacenter vibration is causing significant I/O performance degradation is supported by the poor performance hard disks often exhibit in the field compared to the manufacturers’ specs.

Recent tests seem to back up these claims. Lab testing by Sun Microsystems’ Systems Dynamic Characterization and Control team in San Diego looked at the effect of sequential I/O on SATA drives. They found that as the vibration level increased, disk throughput dropped from 40 to 15 MB/second. At the same time, the amount of energy the system consumed to write 10 terabytes of data increased from about 1.7 to 4.5 kilowatt-hours. The Sun testing also found that Green Platform’s AVP rack effectively dissipates nearly all of the vibration through the critical range of frequencies, restoring performance close to the no-vibration state.

Another set of tests performed by Q-Associates, a system integrator, compared I/O performance of the Green Platform AVP racks and steel racks in a Tier 1 datacenter environment. These tests used the more robustly built SAS drives (2.5-inch), which, by design, should be more resistant to vibrational stress. What they found was the AVP racks increased IOPS performance significantly: 56 to 246 percent for random reads and 34 to 88 percent for random writes, with the larger file sizes benefitting the most.

According to Malek-Madani, the overall result is that the Green Platform racks boost storage throughput by as much as 250 percent, while proportionally decreasing energy consumption related to powering and cooling the storage equipment. There is also evidence that suggests vibration dampening will lengthen MTBF (mean time between failure) for storage hardware, thus reducing support costs.

The price premium for a Green Platform rack compared to a traditional metal rack is significant. According to Gordon, an AVP will cost four to five times that of a steel rack (which runs around $2,000). But to Gordon, that’s not the way to look at this solution. Since the AVP-1000 improves performance and lowers energy costs, the rack can pay for itself in less than 12 months — sometimes much less. Where data throughput, as opposed to data capacity, is the limiting factor of the storage infrastructure, an AVP solution means you will need fewer storage servers. If you can cut your storage infrastructure by a third, payback happens on the first day, says Gordon.

In fact, according to him, the AVP delivers the type of performance enhancement that solid state drive (SSD) caching provides. “Datacenters are spending tens of thousands of dollars on high-speed caching to improve storage performance,” notes Gordon. “Essentially, we provide similar benefits, and we’re much lower cost than storage caching.”

Although Gordon and Malek-Madani think their AVP solution has broad applicability to the enterprise storage space, their initial entry point into the market will be in high performance computing, where storage performance is often the bottleneck. This is especially true in applications like seismic data analysis, which can take days, weeks or even months to process. Shortening the cycle time for that analysis is a critical path to revenue for these oil and gas companies. Beyond HPC, they see a $5 billion dollar market opportunity in the enterprise.

Today, the company is offering the AVP rack in limited quantities for pilot studies in partnership with certain storage OEMs and system integrators. (For the time being, the company is keeping the names of these interested parties to itself.) The initial business model will involve partnerships with some of these same integrators and OEMs, who will bundle their product with storage servers. According to Gordon, they have already begun a number of these relationships and are making “new connections” on a daily basis, and some of these relationships are in the process of being formalized.

They’re also looking for an influx of funding, which to date has been accomplished internally, presumably through Malek-Madani’s Composite Products business. External funding is being sought to continue testing, add a vice president of sales and help scale up manufacturing.

In the short-term, the company is looking to develop some case studies from the pilot studies and expand the program to additional user sites. “We realize that we’re making some pretty extraordinary claims,” admits Gordon. “At least early in the market, we expect people will want to prove this to themselves.”

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