Nubificus develops a highly-optimized hypervisor based on KVM. Our home-brewed hypervisor delivers native performance for I/O-intensive workloads, while maintaining secure execution in multi-tenant environments.
At Nubificus, we minimize the application execution overhead. We achieve this with the use of unikernels while building on the concept of library operating systems.
We bridge the gap between hardware accelerators and virtualized workloads. We define the simplest interface for common snippets of code that can be accelerated and transparently execute the code, in hardware, while the rest of the application execution flow is progressing.
We bring containerized storage to unikernel containers. Most of the benefits of container storage are now available on lightweight, single-app microVM instances, spawned as containers.
We collaborate with academic institutions and R&D departments, to conduct systems research and move beyond the state-of-the-art on systems design and implementation.
With our expertise in low-level systems software, distributed systems and security, we are building a powerful set of tools which we combine to deliver bespoke virtualisation solutions.
We enhance your monolithic application workflow with satellite micro-services running as unikernels, with minimal overhead and near-instant response times.
We design, deploy, and maintain your application in various environments: the cloud, on-prem or at the edge. With more than 20 years of experience in systems administration, we relieve you from the burden of deploying, monitoring and maintaining your application.
At Nubificus, we explore lightweight virtualization solutions for edge, low-power devices. Specifically, we enable hardware acceleration for Serverless computing, by integrating the vAccel framework on lightweight hypervisors such as AWS firecracker! Moreover, we enable the execution of such hypervisors on a RPi4 by coding in the necessary support for running on proprietary GiCs. Furthermore, we bring aarch64 support for Nabla containers.
This way, deployment and management of lightweight workloads on low-power devices is feasible, where resources are scarce and valuable.