Hi! I am a systems engineer, working on various levels of the software stack.
In this space, you can find more information about the projects I am currently
working on and how to contact me.
I’m working on building a lightweight, ultra-performant hypervisor for x86 and ARM architectures. Essentially, this is a research project inspired by Serverless Computing, Unikernels and modern hypervisor design for low-power devices.
More information is available in our FOSDEM 2020 talk.
Accelerating parts of an application can really boost performance. But is it something we can easily do without being too intrusive to the application code? Can we do that in the cloud? I’m working on getting (some of) these answers.
We are building vAccel, a framework that abstracts hardware implementations of operations to functions.
vAccel offers more predictable execution time than device passthrough, as all the complicated software execution is pushed to the Host/Hypervisor. vAccel ensures interoperability as it is hardware/vendor-agnostic by providing a simple interface to add any acceleration framework.
More information about vAccel is available at https://docs.vaccel.org. Additionally, aspects of the vAccel framework are presented at the following FOSDEM 2021 talks:
I’m also looking into unikernels, and how this technology can facilitate deployment of cloud-native apps, without sacrificing performance, isolation, and security.
We are exploring various unikernel frameworks, mainly Rumprun, Solo5, and Unikraft.
Our main goal is to make it easier for end-users to deploy their apps in a unikernel, as well as run a unikernel efficiently, either in the Cloud or at the Edge. To this end, we have been doing some work regarding aarch64 support for rumprun over solo5, optimizing container runtime support for unikernels, while porting unikernel frameworks to lightweight VMMs.