Engineers face many daily operational inefficiencies that inhibit their time-to-solution. Every day we work with engineers to provide solutions to computing resource limitations and management of HPC. Specifically, we excel at utilizing our platform to accelerate HPC engineering simulations. The impact is real: Rescale users have seen accelerated time-to-solution by 23%, allowing engineering teams to be 12% more productive overall.

In this article, we hope to give you exactly what you need to better plan for HPC in 2019.

(Your) 2019 Engineering Objectives: Measurably Improve Engineering Team Productivity

1. Shorten the turnaround time of your engineering services

2. Eliminate engineering hours spent in HPC queues

3. Increase the individual productivity of your engineers

4. Develop best practices for HPC usage by workflow

Some key issues engineers face when developing a product are simulation constraints due to queue times from lack of computing resources, software availability, architecture diversity, and departmental management. The shortage of these vital resources and tools results longer development cycles of the products that generate revenue.   

1. Shorten the turnaround time of your engineering services

By eliminating queue time and enabling engineers with the best HPC hardware and software, you can optimize your research pipeline and push innovations to the market, sooner.

The Proof:

Dinex, an automotive exhaust supplier, saw a reduction in time-to market of 25% by utilizing the Rescale platform. With abundant computing resources available through our public cloud partners, you gain the ability to mitigate queue time by immediately securing the resources as you need them. The abundant computing hardware and software diversity allows engineers to run simulation that were previously unsupported by on-premise systems (either based off intolerable queue time or software and hardware resource demand). The availability of software and computing resources, ability to innovate design of experiments, and the mitigation of queue time allow engineers to be more efficient and deliver products to market faster.

2. Eliminate engineering hours spent in HPC queues

Stop waiting to run your simulations because of limited HPC resources and/or low priority. Empower every engineer with the resources to run simulations immediately using our AWS, Azure, and IBM cloud resources.

The Proof:

Queues for running simulations can halt the research pipeline and waste valuable engineering time. A queue directly results in a delayed time-to-solution that can be critical to the progression of research. The days spent without answers can cost a company millions of dollars in engineer idle time. The ability to secure hardware as needed allows engineers to be agile with their computing resources and break the constraints of a static on-premise HPC system that limit their simulation volume and fidelity. These inefficiencies directly impact the company’s objective to bring innovations to the market and generate revenue; so, the ramifications of research inefficiencies reverberate throughout the entire organization and externally. By utilizing Rescale, you can run a single simulation on 10,000 cores, or run 10,000 simulations on 10 cores each: the availability of resources means there is no reason not to run a simulation immediately.

3. Increase the individual productivity of your engineers

Remove the constraints of static On-Premise HPC systems and engage a dynamic environment with a the latest HPC hardware and simulation software. Explore new DOE and optimize your research pipeline to achieve the fastest time-to-solutions.  

The Proof:

Rescale has over 300 ported and tuned software’s incorporated into our platform; many on a pay as you use model such as ANSYS, Siemens, CONVERGE, and LS-DYNA. Utilization of the endless, diverse computing resources allows engineers to use the best software on the best hardware, always. The coupling of the best software and hardware allows engineers to have the best results available, quickly. In addition, engineers are exposed to new software and computing resources that were previously unavailable. Some Rescale customers have seen as high as 80% reduction in time-to-answers. The freedom of architecture choices allows for the exploration of new processes in your design of experiments which can create quicker research pipelines with higher fidelity. Enabling researchers with the best tools HPC tools produces quicker results and increases productivity.

4. Develop best practices for HPC usage by workflow

Gain real time insight into your engineers activities and utilize the information to optimize your engineering departments operations and finances.

The Proof:

Scale X Enterprise allows you to fully manage your engineers by tracking expenses, allocating resources, and budgeting teams. With control of computing and software resources, budgets, projects, and access, you can fully manage how your engineering teams utilize cloud computing. In addition, access to billing summaries and real time spending dashboards allow you to monitor your computing expenses. Rescale doesn’t only provide a solution to engineering inefficiencies, it gives management the insight to innovate their own research pipeline.  

Rescale is a turn-key platform that enables access to limitless computing resources and over 300 ported and tuned softwares. With ScaleX Enterprise’s management dashboard, engineering departments are capable of fully managing and reporting on their HPC usage. Rescale has had significant impact on many of our customers; but to understand the true impact Rescale can have on your organization, it is best to reach out to us. With our confidential tools and industry leading knowledge, we can define the impact of Rescale on your engineering operations.

If you have any questions or interest in seeing how Rescale can improve your engineering department, please reach out to our specialists today.

This article was written by Thomas Helmonds.

Cambridge, MA – April 26, 2018 – The 15th annual MIT Sloan CIO Symposium just announced Rescale as one of the ten finalists for the 2018 Innovation Showcase, as its platform represents a cutting edge solution that provides both strong value and innovation to the enterprise IT space. Rescale will receive key exposure to many of the world’s most creative and influential IT executives at the Symposium on May 23, 2018.

“We are proud to have been selected by MIT as one of the top 10 outstanding early stage companies with cutting edge solutions that combine both value and innovation to Enterprise IT” said Joris Poort, Founder and CEO at Rescale “We look forward to continuing to work with enterprise customers to instantly enable them to extend their high performance infrastructure to the cloud.”

Rescale offers a platform that manages the world’s largest high performance computing infrastructure. It is designed for engineers and scientists, and is built to help companies accelerate innovation on demand. Continue reading

This article was written by Rescale.

When software engineers look at a piece of code, the first question they ask themselves is “what is it doing?” The next question is “why is it doing it?” In fact, there is a deep connection between these two – why becomes what at the next level of abstraction. That is, what and why are mirrors of each other across an abstraction boundary. Understanding this can help engineers write more maintainable, readable software. Continue reading

This article was written by Alex Kudlick.

Servers image

We have made a number of blog posts over the years where we have run some MPI microbenchmarks against the offerings from the major public cloud providers. All of these providers have made a number of networking improvements during this time so we thought it would be useful to rerun these microbenchmarks against the latest generation of VMs. In particular, AWS has released a new version of “Enhanced Networking” that supports up to 20Gbps, and Azure has released the H-series family of VMs which offers virtualized FDR InfiniBand.

My colleague Irwen recently ran the point-to-point latency (osu_latency) and bisection bandwidth (osu_bibw) tests from the OSU Microbenchmarks library (version 5.3.2) against a number of different VM types from Google Compute Engine. For consistency, we’ll use the same library here with Azure and AWS.  The table below includes the best performing machine from Irwen’s post: the n1-highmem-32. The c4.8xlarge represents an AWS VM type from the previous Enhanced Networking generation and the newer m4.32xlarge VM is running the newer version of Enhanced Networking.

In the table below, we list the averaged results over 3 trials. A new pair of VMs were  provisioned from scratch for each trial:

0-byte Latency (us) 1MB bisection bandwidth (MB/s)
GCE (n1-highmem-32) 41.04 1076
AWS (c4.8xlarge) 37.07 1176
AWS (m4.32xlarge) 32.43 1152
Azure (H16r) 2.63 10807

As you might expect, the Azure H-series VMs seriously outpace the non-InfiniBand equipped competition in these tests. One of the frequent criticisms levied against using the public cloud for HPC is that networking performance is not up to the task of running a tightly-coupled workload. Microsoft’s Azure has shown that it is possible to run a virtualized high-performance networking fabric at hyperscale.

That said, while this is interesting from a raw networking performance perspective, it is important to avoid putting too much stock into synthetic benchmarks like this. Application benchmarks are generally a much better representation of real-world performance. It is certainly possible to achieve strong scaling with some CFD solvers with virtualized 10GigE. AWS has published STAR-CCM+ benchmarks showing close to linear scaling on a 16M cell model on runs up to 700 MPI processes. Microsoft has also published some STAR-CCM+ benchmarks showing close to linear scaling on up to 1,024 MPI processes with an older generation of InfiniBand equipped VMs (note that this is not an apples-to-apples comparison because Microsoft used a larger 100M cell model in their tests). It’s also important to highlight that specialized networking fabric typically comes at a higher price point. Additionally, keep in mind is that network speed is just one dimension of performance. Disk IO, RAM, CPU core count and generation, as well as the type of simulation and model size all need to be taken into consideration when making a decision about what hardware profiles to use. One of the advantages of using a multi-cloud platform like Rescale’s ScaleX Platform is that it makes it easy to run benchmarks and moreover, enterprise HPC workloads, across a variety of hardware configurations by simply changing the core type in your job submission request.

Finally, it is impressive to note how far things have come from the original Magellan report. There is a fierce battle going on right now between the public cloud heavyweights and we are starting to see hardware refresh cycles including not only high-performance interconnect but also modern CPU generations (Skylake) as well as GPU and FPGA availability at large scale. The “commodity” public cloud is increasingly viable for a growing number of HPC workloads.

This article was written by Ryan Kaneshiro.