Nicole Hemsoth

Author Archives: Nicole Hemsoth

Will Chapel Mark Next Great Awakening for Parallel Programmers?

Just because supercomputers are engineered to be far more powerful over time does not necessarily mean programmer productivity will follow the same curve. Added performance means more complexity, which for parallel programmers means working harder, especially when accelerators are thrown into the mix.

It was with all of this in mind that DARPA’s High Productivity Computing Systems (HPCS) project rolled out in 2009 to support higher performance but more usable HPC systems. This is the era that spawned systems like Blue Waters, for instance, and various co-design efforts from IBM and Intel to make parallelism within broader reach to the

Will Chapel Mark Next Great Awakening for Parallel Programmers? was written by Nicole Hemsoth at The Next Platform.

Another Step Toward FPGAs in Supercomputing

There has been plenty of talk about where FPGA acceleration might fit into high performance computing but there are only a few testbeds and purpose-built clusters pushing this vision forward for scientific applications.

While we do not necessarily expect supercomputing centers to turn their backs on GPUs as the accelerator of choice in favor of FPGAs anytime in the foreseeable future. there is some smart investment happening in Europe and to a lesser extent, in the U.S. that takes advantage of recent hardware additions and high-level tool development that put field programmable devices within closer reach–even for centers whose users

Another Step Toward FPGAs in Supercomputing was written by Nicole Hemsoth at The Next Platform.

Momentum for Bioinspired GPU Computing at the Edge

Bioinspired computing is nothing new but with the rise in mainstream interest in machine learning, these architectures and software frameworks are seeing fresh light. This is prompting a new wave of young companies that are cropping up to provide hardware, software, and management tools—something that has also spurred a new era of thinking about AI problems.

We most often think of these innovations happening at the server and datacenter level but more algorithmic work is being done (to suit better embedded hardware) to deploy comprehensive models on mobile devices that allow for long-term learning on single instances of object recognition

Momentum for Bioinspired GPU Computing at the Edge was written by Nicole Hemsoth at The Next Platform.

Aruba Networks Leads HPE to the Edge

When pre-split Hewlett-Packard bought Aruba Networks three years ago for $3 billion, the goal was to create a stronger and larger networking business that combined both wired and wireless networking capabilities and could challenge market leader Cisco Systems at a time when enterprises were more fully embracing mobile computing and public clouds.

Aruba was launched in 2002 and by the time of the acquisition had established itself as a leading vendor in the wireless networking market and had an enthusiastic following of users who call themselves “Airheads.” The worry among many of them was that once the deal was closed,

Aruba Networks Leads HPE to the Edge was written by Nicole Hemsoth at The Next Platform.

An Inside Look at What Powers Microsoft’s Internal Systems for AI R&D

For those who might expect Microsoft to favor its own Windows-centric platforms and tools to power comprehensive infrastructure for serving AI compute and software services for internal R&D groups, plan on being surprised.

While Microsoft does rely on some core windows features and certainly its Azure cloud services, much of its infrastructure is powered by a broad suite of open source tools. As Jim Jernigan, senior R&D systems engineer at Microsoft Research told us at the GPU Technology Conference (GTC18) this week, the highest volume of workloads running on the diverse research clusters Microsoft uses for AI development are running

An Inside Look at What Powers Microsoft’s Internal Systems for AI R&D was written by Nicole Hemsoth at The Next Platform.

A First Look at Summit Supercomputer Application Performance

Big iron aficionados packed the room when ORNL’s Jack Wells gave the latest update on the upcoming 207 petaflop Summit supercomputer at the GPU Technology Conference (GTC18) this week.

In just eight years, the folks at Oak Ridge have pushed the high performance bar from the 18.5 teraflop Phoenix system to the 27 petaflop Titan. That’s a 1000x + improvement in eight years.

Summit will deliver 5-10x more performance than the existing Titan machine, but what is noteworthy is how Summit will do this. The system is set to have far fewer nodes (18,688 for Titan vs. ~4,800 for Summit)

A First Look at Summit Supercomputer Application Performance was written by Nicole Hemsoth at The Next Platform.

Capital One Machine Learning Lead on Lessons at Scale

Machine learning has moved from prototype to production across a wide range of business units at financial services giant Capital One due in part to a centralized approach to evaluating and rolling out new projects.

This is no easy task given the scale and scope of the enterprise but according to Zachary Hanif who is director of Capitol One’s machine learning “center for excellence”, the trick is to define use cases early that touch as broad of a base within the larger organization as possible and build outwards. This is encapsulated in the philosophy Hanif spearheads—locating machine learning talent in

Capital One Machine Learning Lead on Lessons at Scale was written by Nicole Hemsoth at The Next Platform.

The Future of Programming GPU Supercomputers

There are few people as visible in high performance computing programming circles as Michael Wolfe—and fewer still with level of experience. With 20 years working on PGI compilers and another 20 years before that working on languages and HPC compilers in industry, when he talks about the past, present and future of programming supercomputers, it is worthwhile to listen.

In his early days at PGI (formerly known as The Portland Group) Wolfe focused on building out the company’s suite of Fortran, C, and C++ compilers for HPC, a role that changed after Nvidia Tesla GPUs came onto the scene and

The Future of Programming GPU Supercomputers was written by Nicole Hemsoth at The Next Platform.

What’s Ahead for Supercomputing’s Balanced Benchmark

We all know about the Top 500 supercomputing benchmark, which measures raw floating point performance. But over the several years there has been talk that this no longer represents real-world application performance.

This has opened the door for a new benchmark to come to the fore, in this case the high performance conjugate gradients benchmark, or HPCG, benchmark.

Here to talk about this on today’s episode of “The Interview” with The Next Platform is one of the creators of HPCG, Sandia National Lab’s Dr. Michael Heroux. Interestingly, Heroux co-developed HPCG with one of the founders of the Top

What’s Ahead for Supercomputing’s Balanced Benchmark was written by Nicole Hemsoth at The Next Platform.

Singularity Containers for HPC & Deep Learning

Containerization as a concept of isolating application processes while sharing the same operating system (OS) kernel has been around since the beginning of this century. It started its journey from as early as Jails from the FreeBSD era. Jails heavily leveraged the chroot environment but expanded capabilities to include a virtualized path to other system attributes such as storage, interconnects and users. Solaris Zones and AIX Workload Partitions also fall into a similar category.

Since then, the advent and advancement in technologies such as cgroups, systemd and user-namespaces greatly improved the security and isolation of containers when compared to their

Singularity Containers for HPC & Deep Learning was written by Nicole Hemsoth at The Next Platform.

Using Python to Snake Closer to Simplified Deep Learning

On today’s episode of “The Interview” with The Next Platform, we discuss the role of higher level interfaces to common machine learning and deep learning frameworks, including Caffe.

Despite the existence of multiple deep learning frameworks, there is a lack of comprehensible and easy-to-use high-level tools for the design, training, and testing of deep neural networks (DNNs) according to this episode’s guest, Soren Klemm, one of the creators of Python based Barista, which is an open-source graphical high-level interface for the Caffe framework.

While Caffe is one of the most popular frameworks for training DNNs, editing prototxt files in

Using Python to Snake Closer to Simplified Deep Learning was written by Nicole Hemsoth at The Next Platform.

Japan Invests in Fusion Energy Future with New Cray Supercomputer

There are a number of key areas where exascale computing power will be required to turn simulations into real-world good. One of these is fusion energy research with the ultimate goal of building efficient plants that can safely deliver hundreds of megawatts of clean, renewable fusion energy.

Japan has announced that it will install its top-end XC50 supercomputer at the at the Rokkasho Fusion Institute.

The new system will achieve four petaflops, which is over double the capability of the current machine for international collaborations in fusion energy, Helios, which was built by European supercomputer maker, Bull. The Helios system

Japan Invests in Fusion Energy Future with New Cray Supercomputer was written by Nicole Hemsoth at The Next Platform.

Volkswagen Refining Machine Learning on D-Wave System

Researchers at Volkswagen have been at the cutting edge of implementing D-Wave quantum computers for a number of complex optimization problems, including traffic flow optimization, among other potential use cases.

These efforts are generally focused on developing algorithms suitable for the company’s recently purchased 2000-qubit quantum system and have expanded to a range of new machine learning possibilities, including what a research team at the company’s U.S. R&D office and the Volkswagen Data:Lab in Munich are calling quantum-assisted cluster analysis.

The art and science of clustering is well known for machine learning on classical computing architectures, but the VW approach

Volkswagen Refining Machine Learning on D-Wave System was written by Nicole Hemsoth at The Next Platform.

Open Source Data Management for All

On today’s episode of “The Interview” with The Next Platform, we talk about an open source data management platform (and related standards group) called iRODS, which many in scientific computing already know—but that also has applicability in enterprise.

We found that several of our readers had heard of iRODS and knew it was associated with a scientific computing base, but few understood what the technology was and were not aware that there was a consortium. To dispel any confusion, we spoke with Jason Coposky, executive director of the iRODS Consortium about both the technology itself and the group’s role

Open Source Data Management for All was written by Nicole Hemsoth at The Next Platform.

Networks Within Networks: Optimization at Massive Scale

On today’s episode of “The Interview” with The Next Platform we talk about the growing problem of networks within networks (within networks) and what that means for future algorithms and systems that will support smart cities, smart grids, and other highly complex and interdependent optimization problems.

Our guest on this audio interview episode (player below) is Hadi Amini, a researcher at Carnegie Mellon who has focused on the interdependency of many factors for power grids and smart cities in a recent book series on these and related interdependent network topics. Here, as in the podcast, the focus is on the

Networks Within Networks: Optimization at Massive Scale was written by Nicole Hemsoth at The Next Platform.

Changing HPC Workloads Mean Tighter Storage Stacks for Panasas

Changes to workloads in HPC mean alterations are needed up and down the stack—and that certainly includes storage. Traditionally these workloads were dominated by large file handling needs, but as newer applications (OpenFOAM is a good example) bring small file and mixed workload requirements to the HPC environment, it means storage approaches need to shift to meet the need.

With these changing workload demands in mind, recall that in the first part of our series on future directions for storage for enterprise HPC shops we focused on the ways open source parallel file systems like Lustre fall short for users

Changing HPC Workloads Mean Tighter Storage Stacks for Panasas was written by Nicole Hemsoth at The Next Platform.

Pushing Greater Stream Processing Platform Evolution

Today’s episode of “The Interview” with The Next Platform is focused on the evolution of stream processing—from the early days to more recent times with vast volumes of social, financial, and other data challenging data analysts and systems designers alike.

Our guest is Nathan Trueblood, a veteran of several companies like Mirantis, Western Digital, EMC, and current VP of product management at DataTorrent—a company comprised of many ex-Yahoo employees who worked with the Hadoop platform and have pushed the evolution of that framework to include more real-time requirements with Apache Apex.

Trueblood’s career has roots in high performance computing

Pushing Greater Stream Processing Platform Evolution was written by Nicole Hemsoth at The Next Platform.

Expanding Use Cases Mean Tape Storage is Here to Stay

On today’s episode of “The Interview” with The Next Platform we talk about the past, present, and future of tape storage with industry veteran Matt Starr.

Starr is CTO at tape giant, Spectra Logic and has been with the company for almost twenty-five years. He was the lead engineer and architect forthe design and production of Spectra’s enterprise tape library family, which is still a core product.

We talk about some of the key evolutions in tape capacity and access speeds over the course of his career before moving into where the new use cases at massive scale are. In

Expanding Use Cases Mean Tape Storage is Here to Stay was written by Nicole Hemsoth at The Next Platform.

Leverage Extreme Performance with GPU Acceleration

Hewlett Packard Enterprise (HPE) and NVIDIA have partnered to accelerate innovation, combining the extreme compute capabilities of high performance computing (HPC) with the groundbreaking processing power of NVIDIA GPUs.

In this fast-paced digital climate, traditional CPU technology is no longer sufficient to support growing data centers. Many enterprises are struggling to keep pace with escalating compute and graphics requirements, particularly as computational models become larger and more complex. NVIDIA GPU accelerators for HPC seamlessly integrate with HPE servers to achieve greater speed, optimal power efficiency, and dramatically higher application performance than CPUs. High-end data centers rely on these high performance

Leverage Extreme Performance with GPU Acceleration was written by Nicole Hemsoth at The Next Platform.

Machine Learning for Auto-Tuning HPC Systems

On today’s episode of “The Interview” with The Next Platform we discuss the art and science of tuning high performance systems for maximum performance—something that has traditionally come at high time cost for performance engineering experts.

While the role of performance engineer will not disappear anytime soon, machine learning is making tuning systems—everything from CPUs to application specific parameters—less of a burden. Despite the highly custom nature of systems and applications, reinforcement learning is allowing new leaps in time-saving tuning as software learns what works best for user applications and architectures, freeing up performance engineers to focus on the finer

Machine Learning for Auto-Tuning HPC Systems was written by Nicole Hemsoth at The Next Platform.

1 2 3 18