Via Scientific announces Workbench. Launch a high-performance cloud workspace in seconds. No more being limited by your machine. Start fast. Stay fast. 

Laptops are remarkable. Portable, fast, and essential for day-to-day science.

But the demands of modern life sciences research quickly outgrow even the best laptop. Genomics datasets balloon into the hundreds of gigabytes. Imaging projects stall as files refuse to load. Machine learning models sit half trained without access to GPUs. Even routine exploratory work such as tweaking parameters, debugging results, or chasing an unexpected signal grinds to a halt when hardware runs out of steam. Instead of advancing discovery, you spend hours watching progress bars, killing frozen sessions, or rerunning code that worked yesterday but fails today.

Trying to do modern science on a laptop is like trying to cook a wedding feast in a dorm-room microwave. But what if you had a portal between your dorm-room microwave and an industrial kitchen: one that gives you immediate access to gas ranges, convection ovens, large containers and power on demand? That scale doesn’t just make the feast possible – it lets you try more recipes, test variations, and follow ideas you couldn’t even attempt before.

You could buy a 20,000 dollar workstation with 256 GB of RAM, but even that can be pushed to its limits as datasets grow. And owning such a machine still leaves you with the burden of setup and maintenance. 

Unless you are a DevOps pro, you will also need an IT team to install the right packages, manage dependencies, and keep everything secure. The alternative, spinning up raw cloud machines from AWS or GCP, is just as painful: provisioning VMs, configuring access, and troubleshooting mismatched tools. Either way, you lose time, momentum, and focus. 

Even if you manage to get everything running, software versions drift, dependencies clash, and your own code may suddenly stop working – reproducibility becomes an uphill battle before the science even starts.

What you need is power without having to DIY or wait for IT.

Enter Workbench

Workbench, Via Scientific’s latest offering, is a high-performance cloud workstation built directly  into Foundry. It is your personal cloud powerhouse, ready in seconds, accessible from any laptop.

Open a workbook session – your own dedicated cloud environment – in AWS or GCP, customize the CPU, GPU, and memory you need, and launch JupyterLab, RStudio, QuPath, a full remote desktop, or other third-party apps in seconds. You pay only for the hours you use. And because Workbench launches directly from Foundry, you can move seamlessly from running a pipeline to opening an RStudio or QuPath session for deeper exploration – all without ever leaving the platform.

With Foundry, your laptop becomes a supercomputer.

Powerful Research Without the Hardware Burden

Workbench unlocks a way of working no laptop or DIY setup can match. It delivers four core advantages: power, cost, flexibility, and security.

Power

Single-cell workflows can overwhelm laptops long before the science gets interesting. Even the first steps may require 20–30 GB of RAM, and full analyses can scale into the hundreds. Imaging datasets climb just as high, while ML models demand GPUs that most laptops will never have. With Workbench, these limits disappear. You choose the exact CPU, GPU, and memory for the task, scaling up as your project evolves instead of reshaping questions to fit your machine.

Cost

Instead of investing in oversized hardware that sits idle most of the time, along with the IT staff to maintain it,  Workbench lets you pay only for what you consume. Analyses that would once demand major capital costs now run for just a few dollars, saving countless dollars while giving you enterprise-grade compute on demand.

Flexibility

Workbench has two primary modes. Start with a single-tool session if you only need quick exploration in RStudio or Jupyter. Or open the full remote desktop to run multiple applications side by side in one session. Each work session is containerized and preloaded with the tools you need, so you can stop wrestling with setup and focus on your science. You can also pause and resume sessions at will, or run multiple concurrent sessions to juggle projects or datasets in parallel, all with a simple click and no waiting for resources to free up.

Security‍

Security is paramount, and each Workbench you launch runs inside your own AWS or GCP environment with containerized isolation. This ensures that data is secure and prevents the misconfigurations that often plague DIY servers. Additionally, individual authentication protects each user, while persistent storage keeps work safe between sessions. Governance features enforce user-level access boundaries and team-wide compliance standards, so even collaborative projects remain auditable and under control.

Made for Those Who Explore, Test, and Build

Workbench is built for the people doing the hands-on work of science.

• Computational biologists can debug thresholds, visualize results, or test new workflows without fighting environment setup. These kinds of follow-up questions often require far more RAM than a laptop can handle – with Workbench, you can scale instantly instead of abandoning the thread.
• Data scientists and analysts can run exploratory data analysis on large datasets using pandas, tidyverse, and seaborn.
• Machine learning scientists can prototype models with scikit-learn, caret, or PyTorch, then scale up to GPU-backed sessions in seconds.
• Imaging researchers can launch QuPath or napari to process massive whole-slide or microscopy images at cloud scale.
• Researchers focused on exploration can test parameters, inspect outputs, or chase unexpected signals without stalling on local limits. Instead of reshaping the question to fit the machine, Workbench lets the machine flex to fit the question.
• Non-coders and collaborators gain reproducible, shareable environments that make analysis auditable and transparent.

Workbench is especially valuable for small and mid-sized teams, CROs, and early-stage biotechs that need enterprise-grade infrastructure without the cost of an IT department. It is equally powerful for academics who want to move faster than their institution’s shared servers allow.

Tools and Features Ready the Moment You Log In

Workbench goes beyond raw compute with features that make it practical for everyday research, and delightful to use.

• One-click launch: Open a ready-to-use cloud environment in seconds, already configured for your work.
• Prebuilt templates: Use environments tailored for genomics, machine learning, or imaging, with tools like Seurat, tidyverse, scikit-learn, or QuPath already installed. The Foundry SDK and CLI are also preloaded, so you can run pipelines, test integrations, or script workflows immediately.
• Custom environments: Beyond the prebuilt templates, you can bring your own software stack. Install the packages, libraries, or tools you prefer and save them as a reproducible environment for future sessions.
• Pause and resume: Suspend a session when you are not using it, then restart later without losing progress.
• Multiple sessions: Run more than one Workbench at a time for parallel projects. For example, you can keep a single-cell workflow running in one environment while prototyping an ML model in another – no waiting for one job to finish before starting the next.
• Cost-aware controls: You can set custom shutdown timers for each session, so resources stop when you decide. Paused sessions stop incurring compute charges while preserving your work with minimal storage costs. Once a workbench is terminated, both compute and storage costs stop entirely.
• Reproducibility: Containers keep environments consistent across reruns, so results are always verifiable and shareable with collaborators. 

Future-proof and Disposable by Design

Workbench environments are both reliable and flexible because they’re powered by Foundry. Each session is containerized and version-locked, so analyses can be rerun months or years later with identical results. Foundry’s infrastructure ensures that updates, security patches, and new libraries arrive without breaking past work, keeping your science future-proof.

At the same time, Workbench is disposable – similar to a plastic cup from a food store. You can reuse it a few times, but unlike a personal laptop or desktop, where mistakes can leave you stuck with problems until you fix them, a Workbench session allows a fresh start at any time. That disposability makes it possible, and even desirable, to use Workbench in a more experimental way, encouraging researchers to explore their data freely without the fear of breaking or ruining the entire system.

A better way to support research

Science moves forward when barriers fall. Workbench clears those obstacles away and puts the power of a supercomputer directly in your browser, so you can focus on discovery. From genomics to imaging to machine learning, the resources you need are just a click away.

With Foundry, Workbench turns your laptop into the launchpad for limitless research.

Ready to try it?

Workbench is available now. You can launch it directly from your workspace, or ask your admin to enable access for your team.

Want a closer look at how it works? Schedule a call here.