Why the Cloud is Ideal for HPC

By Nicole Hemsoth

September 5, 2013

Cloud computing has emerged as a model to address a broad range of computing needs and promises to solve all but world peace. The idea of utility or on-demand computing is hardly new but the business models and technology have matured sufficiently to propel the concept firmly back into the limelight. High Performance Computing (HPC) is where most progressive businesses should be focusing their Big Data and Big Compute efforts.

Unfortunately, some companies pigeon hole HPC as scientists geeking out with massive supercomputers. However, there is a far more significant market evolving outside of traditional HPC with growing computational needs. Big Data and Big Compute is an excellent example where the opportunity for deep analytical insight is an extremely attractive proposition to a wide range of vertical markets. This new segment of the market is often referred to as the “Missing Middle,” and disruptive technology solutions such as those offered by GreenButton put supercomputing power within progressive companies’ reach.

With effective job management as demonstrated below, thousands of HPC activities can be managed by companies such as GreenButton though innovative technology and simple to use web interfaces.

 GreenButton Web Interface

Keep in mind, when we talk about HPC in the cloud, it’s important to remember that cloud computing is ultimately a business model, not a technology. However, there are a common set of technical capabilities (e.g. virtualization) that are realized in the cloud, and these, as well as the business model provide certain benefits or challenges to HPC applications.

The Positives

CFO’s Pay Attention to this!!

Cost: For spikey workloads, the cost savings in the cloud can be significant. What’s more, the cost is an Operational Expense rather than Cap-Ex so is often more palatable for many businesses as costs can be attributed to a particular project.

Ease of use: The cloud can make dynamic provisioning of specified workloads very easy. The ability to have OS/Software configurations particular to a workload is a key advantage.

Speed of deployment: The ability to rapidly provision new environments/clusters in minutes is incredibly valuable to many businesses.

Scalability: Elastically scaling out to meet increased capacity demands is a powerful concept. The public cloud promises “infinite” scale. The reality is somewhat different: there are some real limits even in the cloud. But the computing capacity that you can get from large providers such as AWS and Azure is far greater than what most customers can fathom with internal hardware. Some companies – such as GreenButton work with a variety of cloud providers such as Windows Azure, HP Cloud Services, OpenStack, Amazon Web Services, and VMware for global access to resources.

 GreenButton Map

Resiliency: The ability to snapshot workloads as they are running can allow for check-pointing of MPI workloads. Combine this with active monitoring and the ability to dynamically move a guest VM from one physical host to another, and your workloads can keep running even in the face of hardware failure.

Portability: The ability to move a workload from one cloud platform to another on the fly without any application changes presents powerful options such as for bursting from a private cloud out to a public cloud, High Availability where a workload is run on multiple clouds simultaneously, scaling across multiple clouds to meet extremely high resource requirements or, to take advantage of shifts in the spot pricing market.

Challenges

Security: this remains a significant barrier to adoption today, but the issue is primarily in trust and perception rather than real limitations of the cloud platforms. One could argue that in some cases your data is safer in the hands of Amazon or Microsoft than your own data center. That said, data isn’t sufficiently secure by default so some effort commensurate to the sensitivity and risks needs to be applied. For example;

  • Encryption at rest of cloud-bound data.
  • Limiting the time window that data is resident in the cloud.
  • Anonymizing data. A great example is running risk models for the financial services sector where sensitive customer data can easily be stripped out prior to sending to the cloud.

Performance: There is no single answer to the question of performance, though in general the cloud offers massive performance gains in most cases (and therefore is generally a positive), this does depend on the workload in question and presents some challenges today.

Some workloads scale in a linear fashion i.e. embarrassingly parallel, and these scale extremely well to the cloud. Even many MPI workloads scale perfectly well on cloud infrastructure.

However, I/O bound MPI processes will often run into performance challenges due to their heavy demands on network infrastructure or sensitivity to latency. Many traditional HPC applications are tuned for very low-latency Infiniband interconnects and take advantage of RDMA technology. These applications just won’t scale on the 10 GigE networks within the cloud.   This will change as cloud providers roll out Infiniband or RDMA over Ethernet but for the time being remains an issue.

Other challenges lie in certain cloud platforms intentionally distributing your deployed instances across the data center to increase availability. This can negatively affect performance through increased latency. But this is increasingly being rectified with increased control over physical placement of VMs – e.g. AWS Placement Groups.

I’m not going to dwell on the overhead of virtualization as there is a lot of material on the web covering this topic. I will say that modern virtualization technologies have such a small overhead on CPU performance today that it is effectively negligible. The I/O hit in some cases can be more noticeable but this depends on the characteristics of the workload. Josh Simons of VMware has posted extensively about this so check out his posts at http://cto.vmware.com/author/joshsimons/

Management: One of the challenges when spreading workloads across more than one platform is management of the workloads and resources being utilized. Being able to consolidate management within a single tool becomes critical for effective use of the cloud.

Data: Moving large datasets to the cloud still presents some challenges. In the Oil & Gas sector we physically ship 50TB+ to AWS where it undergoes weeks/months of processing, and the entire workflow lives in the cloud using visualization technology. RenderMan workloads also present challenges with large datasets (up to 1GB per frame). There are also technologies such as Aspera or GreenButton’s own CloudSync which optimize throughput over the internet.

Managing Costs: There is obviously some level of fear when moving from a known and understood capital expenditure model to one of pay-by-the-drink where costs could spiral out of control. Trust me, this has many CFOs awake at night in a cold sweat. At GreenButton, we’ve addressed this by predicting job execution time and committing to users on runtime and cost. We also support cost monitoring and chargebacks down to the departmental or user level so the CFO never has to get any nasty surprises!

Cloud Lock-in:   Different cloud vendors have different APIs and deployment mechanisms, so you may be concerned about being locked into a particular cloud and being unable to take advantage of improved pricing or services becoming available in other clouds.  I’ve written about how to avoid cloud lock-in before so I won’t repeat it here!

Cloud Bursting

One advantage of moving HPC workloads to on-demand virtualized infrastructure is that Enterprise customers can take advantage of internal hardware investment in the form of a private cloud. The private cloud obviously solves some of the issues around security and data transfers, at the cost of limited capacity. But throw in the ability to seamlessly burst to nominated public clouds and you have something pretty compelling indeed. Below is an example of how this can be implemented effectively.

 GreenButton Cloud

Conclusion

Not only is the cloud an ideal platform for many HPC (and non-“HPC”) workloads today, but current limitations are constantly being whittled away by the platform providers themselves or by software vendors such as GreenButton. There is a common perception that HPC is so complex and expensive that ordinary businesses are not able to tap into the massive benefits and business value that can be obtained.  With the advent of the cloud HPC is accessible and affordable to the mass market for any type of application. Do your research to find the solutions that work best for you!

Dave Fellow, CTO, GreenButton™

 

 

Dave Fellows, CTO, GreenButtonAbout the Author

Dave Fellows is the Chief Technology Officer of GreenButton ™ Limited. Dave has extensive experience designing massively scalable PaaS applications in a variety of technology industries. He has a passion for the Cloud and High Performance Computing (HPC) and creating innovative technologies to bring unique and compelling solutions to GreenButton’s global customers.

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!

AI in the News: Rao in at Intel, Ng out at Baidu, Nvidia on at Tencent Cloud

March 26, 2017

Just as AI has become the leitmotif of the advanced scale computing market, infusing much of the conversation about HPC in commercial and industrial spheres, it also is impacting high-level management changes in the industry. Read more…

By Doug Black

Scalable Informatics Ceases Operations

March 23, 2017

On the same day we reported on the uncertain future for HPC compiler company PathScale, we are sad to learn that another HPC vendor, Scalable Informatics, is closing its doors. Read more…

By Tiffany Trader

‘Strategies in Biomedical Data Science’ Advances IT-Research Synergies

March 23, 2017

“Strategies in Biomedical Data Science: Driving Force for Innovation” by Jay A. Etchings is both an introductory text and a field guide for anyone working with biomedical data. Read more…

By Tiffany Trader

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

HPE Extreme Performance Solutions

HFT Firms Turn to Co-Location to Gain Competitive Advantage

High-frequency trading (HFT) is a high-speed, high-stakes world where every millisecond matters. Finding ways to execute trades faster than the competition translates directly to greater revenue for firms, brokerages, and exchanges. Read more…

Google Launches New Machine Learning Journal

March 22, 2017

On Monday, Google announced plans to launch a new peer review journal and “ecosystem” Read more…

By John Russell

Swiss Researchers Peer Inside Chips with Improved X-Ray Imaging

March 22, 2017

Peering inside semiconductor chips using x-ray imaging isn’t new, but the technique hasn’t been especially good or easy to accomplish. Read more…

By John Russell

LANL Simulation Shows Massive Black Holes Break ‘Speed Limit’

March 21, 2017

A new computer simulation based on codes developed at Los Alamos National Laboratory (LANL) is shedding light on how supermassive black holes could have formed in the early universe contrary to most prior models which impose a limit on how fast these massive ‘objects’ can form. Read more…

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

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

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

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

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

New Japanese Supercomputing Project Targets Exascale

March 14, 2017

Another Japanese supercomputing project was revealed this week, this one from emerging supercomputer maker, ExaScaler Inc., and Keio University. The partners are working on an original supercomputer design with exascale aspirations. Read more…

By Tiffany Trader

Nvidia Debuts HGX-1 for Cloud; Announces Fujitsu AI Deal

March 9, 2017

On Monday Nvidia announced a major deal with Fujitsu to help build an AI supercomputer for RIKEN using 24 DGX-1 servers. Read more…

By John Russell

HPC4Mfg Advances State-of-the-Art for American Manufacturing

March 9, 2017

Last Friday (March 3, 2017), the High Performance Computing for Manufacturing (HPC4Mfg) program held an industry engagement day workshop in San Diego, bringing together members of the US manufacturing community, national laboratories and universities to discuss the role of high-performance computing as an innovation engine for American manufacturing. Read more…

By Tiffany Trader

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

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

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

Lighting up Aurora: Behind the Scenes at the Creation of the DOE’s Upcoming 200 Petaflops Supercomputer

December 1, 2016

In April 2015, U.S. Department of Energy Undersecretary Franklin Orr announced that Intel would be the prime contractor for Aurora: Read more…

By Jan Rowell

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

Enlisting Deep Learning in the War on Cancer

December 7, 2016

Sometime in Q2 2017 the first ‘results’ of the Joint Design of Advanced Computing Solutions for Cancer (JDACS4C) will become publicly available according to Rick Stevens. He leads one of three JDACS4C pilot projects pressing deep learning (DL) into service in the War on Cancer. 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

Leading Solution Providers

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

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

CPU Benchmarking: Haswell Versus POWER8

June 2, 2015

With OpenPOWER activity ramping up and IBM’s prominent role in the upcoming DOE machines Summit and Sierra, it’s a good time to look at how the IBM POWER CPU stacks up against the x86 Xeon Haswell CPU from Intel. Read more…

By Tiffany Trader

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

IDG to Be Bought by Chinese Investors; IDC to Spin Out HPC Group

January 19, 2017

US-based publishing and investment firm International Data Group, Inc. (IDG) will be acquired by a pair of Chinese investors, China Oceanwide Holdings Group Co., Ltd. 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

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

Intel and Trump Announce $7B for Fab 42 Targeting 7nm

February 8, 2017

In what may be an attempt by President Trump to reset his turbulent relationship with the high tech industry, he and Intel CEO Brian Krzanich today announced plans to invest more than $7 billion to complete Fab 42. Read more…

By John Russell

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