WOA Issue 12
In this issue: Fedora 26 on ARM release Dennis Chen…
Linaro is pleased to announce the first speaker at our Arm Architecture HPC workshop in Santa Clara on July 26th. Simon McIntosh-Smith will give a talk at about Isambard: the world’s first production Arm-based supercomputer.
Michael Feldman interviews Simon McIntosh-Smith of Isambar for the “Top 500” news.
Along similar lines, Mcintosh-Smith also points to what he believes is Arm’s biggest advantage, namely that the licensing of its shrink-wrapped IP enables people to build customized processors at lower cost than would be possible with a commodity chip business model. The implication is that some enterprising startup or startups could construct specialized Arm processors for the HPC market, incorporating custom circuitry for vector processing, AI, or other interesting types of acceleration.
“These processors will be highly differentiated from high-volume mainstream datacenter parts, and should bring significant steps forward in performance for scientists around the world who have become increasingly frustrated with the relatively small improvements in performance and performance per dollar we’ve seen in recent years,” writes Mcintosh-Smith. “As such, Arm’s entry into the HPC market, and the injection of new ideas, innovation and competition this brings, could trigger a revolution in scientific computing of the kind not seen since the commodity CPU revolution of the late 1990’s. Exciting times are ahead.”
The Works on Arm cluster offers resources for projects doing development, testing, and CI/CD on Arm. Periodically I report back here on new projects that are underway and some sense of their status. Details of project engagement are kept in a Github issue tracker at http://github.com/WorksOnArm/cluster .
The Electron build system relies on Microsoft’s Visual Studio Team Services (VSTS), which in turn relies on the .NET Core CLR runtime. Work is progressing to get full arm64 support for all of this development tooling.
Adélie Linux is a Free, Libre operating system based on the Linux kernel. Our goals of full POSIX® compliance, compatibility with a wide variety of computers, and ease of use without sacrificing features set us apart from most other Linux distributions.
Adélie Linux uses the Works on Arm infrastructure to support a port to arm64. The project uses Gitlab for collaboration, source code management, and testing.
The Works on Arm directory is a reference guide to software and systems that have been developed for or ported to arm64 server-class hardware. While it is by its nature incomplete – new software appears daily – it does provide a handy reference for identifying more information when bringing your favorite codes to a new arm64 based platform.
Entries for the following databases have been updated or added in the Works on Arm directory.
In addition, we are working on updates for the following databases, each of which have some level of support for arm64 systems. Look for news in future editions of this newsletter.
Packet provides a highly automated environment for setting up and managing bare metal servers, including the servers found in the Works on Arm cluster. Program participants can make use of programmable tools like Terraform, Ansible, cloud-init and fog to set up and configure servers.
If you are interested in automation at Packet, see the #integration channel on the Packet Community Slack, which you can join via http://slack.packet.net .
A new provisioner in early testing for Packet is support for Hashicorp’s Vagrant. HashiCorp Vagrant provides the same, easy workflow regardless of your role as a developer, operator, or designer. It leverages a declarative configuration file which describes all your software requirements, packages, operating system configuration, users, and more.
The new provider,
vagrant-packet, was developed by Jeffrey Sica at the University of Michigan. This code leverages the work that was done at Packet on the
fog provisioning tool, and lets you set up servers at Packet as easily as typing
Grafana is an open source metric analytics & visualization suite. It is most commonly used for visualizing time series data for infrastructure and application analytics but many use it in other domains including industrial sensors, home automation, weather, and process control.
As of version 5.2, Grafana is now doing native builds of these tools for arm64. They write:
We’ve been longing for native ARM build support for ages. With the help from our amazing community this is now finally available. Please try it out and let us know what you think.
Nick Smith has built a high performance 12-node NanoPi-Fire3 supercomputer for under £100 (£550 including twelve Fire3s).
After the interest in my cluster of Raspberry Pi 3s last year, I was keen to try building clusters with some of the other excellent SBCs now on the market. FriendlyARM in China very generously sent me 12 of their latest NanoPi-Fire3 64-bit ARM boards, each with an EIGHT core ARM A53 SoC running at 1.4GHz with gigabit Ethernet.
His benchmarks for scientific computing use the Linpack linear equation solver to generate 60000 MFLOPS, about 6 times the performance of the 1992-era 16-core Cray C90 supercomputer.
60,000 MFLOPS isn’t all that fast by current performance standards, but back in 2000 the 12-node Fire3 cluster would have made it into the Top250 fastest supercomputers in the world (!).
Nick has been building quite a variety of clusters out of single-board computers, and his builds are very tidy through custom-built enclosures and cabling.
Subscribe to DevOps’ish for updates on DevOps, Cloud Native, and Open Source news. Editor Chris Short has just taken a new job at Red Hat on the Ansible team as Principal Product Marketing Manager.
Issue 082 has capsule reviews of dozens of interesting stories from the DevOps world, including an interview with Slack’s head of engineering, Julia Grace; a report on Comcast’s fiber cuts; Chick-fil-A’s edge computing strategy; and technical debt.
Chris Short is a 20+ year veteran of the IT industry and 11 year veteran of the US Air Force. He helps people and companies embrace DevOps practices and tools through writing and public speaking. He is a staunch advocate for transparency and open source solutions to problems. Follow him on Twitter as @ChrisShort .
Arm-based Linux systems are often used in embedded control systems for radio applications. Outernet provides a platform for receiving broadcast data transmissions from satellite-based transmitters, using the highly robust LoRa protocol to encode transmissions for terrestrial reception.
The receiver is called ‘Dreamcatcher 3’, and is a custom PCB containing a hardware receiver (non-SDR based) with a LoRa decoder, as well as an embedded ARM computer capable of running Linux.