Lambda Compute
State-of-the-art compute, rented by the day.
A high-performance remote workstation for genomics, machine learning, and protein modeling — without buying a workstation or paying enterprise cloud rates.
Coming soon
The wet lab, over HTTPS
Lambda Compute is a high-performance workstation today (specs and software below). Tomorrow it's also the terminal for Lambda Biolab's automated bench — schedule pipetting protocols, watch incubators, and pull plate-reader data over HTTPS, from anywhere with a network connection.
Endpoints, not bench visits
Trigger Gibson assemblies, serial dilutions, OD reads, and colony imaging over HTTPS. A webhook fires when your run finishes and data lands back on your workstation.
Every run is a script
Protocol, reagents, and scheduling are code. Replay the same experiment six months later and get the same result — diff two runs the way you diff two commits.
Wet ↔ dry, same hostname
Chain a variant call straight into a knockout validation. Code and reagents on the same shelf — no airdrop dance between cloud regions.
Private alpha with design partners — get in touch to help shape the first endpoints.
Real workloads on the workstation — debugging a pipette driver in PyCharm.
Hardware
What you're renting
CPU
Intel Core i9-13900K
24 cores, boost clock 5.80 GHz
RAM
192 GB DDR5
Corsair Vengeance, 5600 MHz
GPU
NVIDIA RTX 4090
24 GB GDDR6X, 1,008 GB/s bandwidth
Storage
Up to 1 TB NVMe
7,450 MB/s read · 6,900 MB/s write
Software
Ready when you log in
GATK 4 suite
Sandboxed Docker container preloaded with the Broad's GATK best-practices pipelines. SSH or web login.
Galaxy server
Customized Galaxy instance for point-and-click genomic analyses — DeepVariant, variant annotation, and friends.
Bring your own stack
Need PyTorch, AlphaFold, or a custom CUDA build? Container-friendly host with sudo on your slice.
Galaxy on Lambda Compute — DeepVariant workflow loaded
Need a workstation for a project?
Tell us what you'd run, for how long, and we'll quote a daily rate or a weekly bundle.