This issue’s interview subject, Boston University professor Jonathan Appavoo, stands out in just about any crowd. Unless you’ve spoken with him, however, you won’t know that his unorthodox appearance finds a parallel in the research ideas he pursues to surprising conclusions. A perfect fit for Red Hat, he is passionate about the power of open source and open ideas to “enable flight, not create bonds.” Professor Appavoo is in fact one of the most interesting people I have ever met, and we’ve been very happy to support him in research ideas that run the gamut from AI-driven branch prediction for CPUs to developing open textbooks and course materials using the newly christened Red Hat OpenShift AI platform. He is driven to enable other people to do things we haven’t yet imagined, and we’re lucky to be along for the journey.
While we’re on remarkable journeys, I’m thrilled to be able to publish our in-depth survey of the RISC-V processor extension landscape. Author Rich Jones and I have been working together since before we both came to Red Hat, and I distinctly recall asking him in 2015 or so if he could maybe spend less time on building RISC-V for Fedora and more time on the work I needed him to do. Of course, he was right, and I wasn’t. Today, the RISC-V landscape is critically important not only for Red Hat Research but also for Red Hat products. If you’ve wondered what is so different about RISC-V beyond the fact that its ISA is open source, Rich’s guide to the world of RISC-V extensions will answer you. I think this is one of those articles that will become a resource for all kinds of people working in this space. If you want to understand where RISC-V is going and what makes it so different, it’s a must-read.
I’m also excited to report that our university partners BU and Harvard launched the New England Research Cloud, or NERC. NERC provides containers managed by Red Hat OpenShift, VMs managed by Red Hat OpenStack Platform, and data science training and hosting tools from Red Hat OpenShift AI. Getting these pieces put together and operating smoothly—including billing and charge-back—in a university infrastructure was no mean feat. Tzu-Mainn Chen pulls back the curtain to show how we and the university IT folks we worked with pulled it off. We’d like this to be a pattern for other research infrastructure installations; if you’d like to know more, Mainn will be glad to point you to the open source code repos with the various recipes.
Of course, research using a cloud like the NERC requires a lot of data, and it is more and more likely that some of that data will be affected by privacy constraints. Fortunately, interesting research avenues are opening up in the data privacy and data sovereignty space. Whether you’re trying to account for how much privacy you have lost, use a non-trustworthy computing platform without worry, or do research on disconnected data sets, there is probably a software solution out there that will help you. Red Hat Research staff writer Gordon Haff has put together an excellent survey of what’s available and what is happening in research in each space. Although there’s no such thing as anonymized data, at least there are ways to share some things without compromise.