Trial data and R&D IP on a machine only your team touches.
Private LLM hosting on a dedicated machine in EU-Central, for pharmaceutical and biotech teams working with clinical trial data, molecule research, and regulatory submissions.
Not compliance, regulatory, or legal advice, and not a GxP-certified product. This page describes GPUwerk's infrastructure and contractual commitments only. Whether a given use of trial or research data meets GxP, GDPR, or your sponsor's requirements is for your own quality and regulatory function to assess.
Why trial data and research IP need a different answer
Clinical trial data carries patient information and sponsor obligations. Molecule research and formulation data are the company's core asset. Neither belongs in a general-purpose AI tool with unclear data handling.
"Our researchers already use public AI tools"
A scientist summarizing trial protocols or pasting assay results into a consumer AI tool sends that content to a third party outside your control. On a dedicated instance, the same workflow, same chat interface, runs on hardware assigned only to you. See the private ChatGPT setup →
"The provider could see our unpublished results"
GPUwerk runs infrastructure, not models: content processed on your dedicated machine is handled by software you installed, behind SSH keys only you hold. Under the data processing agreement, GPUwerk does not access your workspace's content except at your request for support or where a legal obligation requires it.
"We need this validated for GxP use"
We don't hold or claim GxP certification, and infrastructure alone can't provide it: GxP validation applies to your processes and systems as a whole. What we can supply is documentation of the physical location, access controls, and change history your validation effort would need as inputs.
Where the infrastructure facts sit
If your program has GxP or sponsor-specific obligations, keep those separate from what infrastructure alone can answer.
| Question | Answer |
|---|---|
| Is this GxP compliant? | Not claimed. GxP applies to your validated processes and systems; consult your quality function about what your specific submission or trial requires. |
| Where is data physically processed? | EU-Central, on a dedicated single-tenant machine assigned to you |
| Who can access instance content? | Through the instance itself, only holders of your SSH keys; password login is disabled fleet-wide. GPUwerk keeps infrastructure administrator access, as on any hosted service, and under the DPA does not use it on your content except at your request or where a legal obligation requires it. |
| Sub-processors for the workload? | None, listed at /legal/sub-processors |
| DPA (GDPR Art. 28)? | Published at /legal/dpa, no charge |
| Certifications held? | None claimed on this page or elsewhere on the site. If your procurement requires a specific certification, verify directly with us before relying on this page. |
| Data on termination? | Container and workspace volume deleted from the node, then the node is sanitised before reassignment. Filesystem deletion, not a cryptographic erase; export what you need before terminating, since it isn't reversible. |
Pricing
On-demand spark-1x is $0.79/hour, spark-2x is $1.79/hour. Held (reserved) capacity is 75% of the on-demand rate: $0.59/hour and $1.34/hour. Full detail on the pricing page.
Questions we get from pharma and biotech teams
Is GPUwerk GxP compliant?
We don't claim GxP compliance or certification. GxP covers your validated processes and systems, not the underlying infrastructure alone. If your workload runs in a GxP-validated environment, that validation is your responsibility; GPUwerk can provide documentation of the physical location, access controls, and change history that your validation effort would need as inputs.
Can we run trial data analysis without it touching a third-party API?
Yes. Your instance is a dedicated single-tenant machine reachable only over SSH keys you hold; there is no third-party model API in the path unless you choose to call one yourself.
Is it safe to run clinical trial data through an LLM here?
That depends on where the model runs and who can reach it, not on the model itself. On a GPUwerk dedicated instance, the machine is assigned to you alone, in EU-Central, reachable only through your SSH keys; GPUwerk does not access, read, copy, index or analyse what runs on it except at your request for support or where a legal obligation requires it. Whether a specific trial's protocol or your IRB and ethics approvals permit this kind of processing at all is a separate question for your own clinical operations and legal teams.
Does R&D IP leave the EU on your platform?
No. The hardware is physically in EU-Central, operated by PRINT IT! SE, a Societas Europaea registered in Prague. There is no US parent or US-incorporated entity in the chain.
Do you sign a DPA covering clinical or research data?
Yes, a standard GDPR Article 28 DPA is published at /legal/dpa at no charge, and there are no sub-processors for instance workloads. Email hello@gpuwerk.com with questions.
Related pages
See how other regulated teams use the same infrastructure: healthcare, insurance, or back to private LLM hosting overview.
Related pages
Bring your research and trial data workload to a machine you control.
Talk to the people who run the racks, or start with a pilot and a practical rollout plan.
Talk to us Deploy an instance