flxflx
Joined 5 years ago
Karma
68
Submissions
19
Member since
Jan 18
Recent submissions
Comment · 1 year ago
Comment · 2 years ago
Ian Buck, VP Hyperscale at Nvidia, also participates. There's also a keynote by Phil Venables, CISO Google Cloud, on the same topic.
Comment · 4 years ago
This is the corresponding repo: https://github.com/edgelesssys/constellation
Comment · 4 years ago
Thanks! There is some performance impact. How much depends on the workload. The impact comes from network, storage, and memory encryption. For most workloads the overhead should be rather small. We have a performance comparison with AKS and GKE: https://docs.edgeless.systems/constellation/overview/perform...
Comment · 4 years ago
Thanks! Target audiences are: people running sensitive clusters on-prem for security reasons and wanting to move to the cloud. There's also regulation in Europe in the eHealth space that requires runtime encryption. Ultimately, I believe that keeping workloads always encrypted and verifiable can be valuable in many scenarios.
Comment · 4 years ago
"I remember as a kid looking at the images on the back of video game cases and wishing I could explore these worlds". Same here. Love the idea and execution.
We are a team of ~20 people, building cutting-edge open-source tools for confidential computing and a "confidential GenAI" service on top of those.
You can find the GenAI service here: https://www.privatemode.ai/ and most of our code here: https://github.com/edgelesssys/
Our products span unusually far across the tech stack, starting at measured boot, through Kubernetes operators, finishing somewhere in our JavaScript frontend. We don't expect you to do all of it - these are just possibilities.
Tech stack: Go (mostly), Rust (some), JavaScript (some), Nix / NixOS, Kubernetes, Terraform, vLLM, AMD SEV-SNP, Intel TDX.
Website: https://www.edgeless.systems/ Contact: Apply via the posting on our careers page. Feel free to email me (founder) via fs [at] edgeless.systems.