benhg Goto Github PK

I am a hardware/software architect and researcher primarily interested in the intersection of computation and science. Some of the most interesting things I've been able to do are helping domain scientists meet their computational needs. Much of the public work I've done has to do with making computational scientific resources easier to access and utilize in order to streamline the process of computation-based science.

Regarding computers, I am interested in novel hardware architectures (especially advanced memory and I/O systems), high-performance computing and networking, and (distributed) operating systems. Other academic interests interests include solid-state physics and thermodynamics, high-performance computing and parallelism, cloud and grid computing, discrete mathematics and combinatorics, and linear algebra. Outside of work, I enjoy flying airplanes, cooking (and consuming) food, ham radio, and various sports.

In general, my work has boiled down to three things - OS/Accelerator development, HPC system design, development, and management, and research (a broad category).

I have experience everywhere from programming low-level custom hardware to high-level system management to core, memory, and IO system architecture. I am passionate about making cutting-edge systems accessible to everyone, regardless of their experience levels. I enjoy guiding systems through all phases of their life cycle, from hardware design to sustained operation and long-term support.

As a day job, I work as a HW/SW architect at SAPEON. I work in the "hardware/software co-design" bucket. Some examples include designing a next-generation memory and I/O subsystem, desinging a distributed OS to support large ML training and inference, and consulting on our system's chip and system architecture.

Most of my research has had to do with making scientific workflows more efficient and easy to operate, in order to help make domain scientists' computational lives easier. I have worked in several domains, including bioinformatics, computational phylogenetics, and quantum chromodynamics. Some of the code I have written for those applications is right here on my GitHub. Additionally, (almost) all of my publications are listed on my website.

Finally, I enjoy designing, building, operating, and otherwise working with small high-performance computing systems, such as campus clusters, small supercomputers, and grid systems. I am talented at extracting the maximum possible performance out of small systems in a user-friendly manner. I am experienced at designing and building custom middleware and system software to make scientific computing tasks performant, easy, and useful for a wide audience, ranging from undergraduate students to scientific researchers to faculty. If you are curious about any of this, please get in touch!

Some more details about me and my work are available at my website.

If you want to contact me, here are some ways to get in touch:

• Email [email protected]
• LinkedIn [https://linkedin.com/in/benjamin-glick]