Capabilities / Multi-Party Computation
Compute on shared data, reveal nothing
Multi-Party Computation (MPC) is a cryptographic method that lets multiple parties jointly analyze data while keeping their individual inputs private, so partners can collaborate without ever exposing sensitive information.
Why it matters
Transparency without giving away your data
Semiconductor supply chains struggle with product complexity, fragmented data and long lead times. MPC lets partners securely aggregate demand signals and other KPIs, increasing end-to-end demand & supply transparency while every company keeps its raw figures private.
How it works
Four steps, zero exposure
Step 1
Secure infrastructure
Scalable, secure infrastructure within trusted digital ecosystems, the foundation for joint computation.
Step 2
Encrypt & split
Each party encrypts its data and splits it into secret shares. No raw input ever leaves the owner.
Step 3
Compute jointly
The shares are processed together to compute the agreed result, without exposing any individual input.
Step 4
Reveal only results
Only the computed output is reconstructed into actionable insights. The inputs stay private.
MPC in action
CryptSurvey
Surveys with zero knowledge of individual data.
Collect sensitive data and compute aggregates, sums, averages, min/max, variance, without ever seeing individual responses. Responses are split into encrypted shares across independent servers, so no single party can view raw data. Ideal for salary benchmarks, industry comparisons and employee satisfaction surveys.
Opaix leads the “Survey” subgroup within the Supply Chain working group of the EU Industrial Alliance for Processors & Semiconductor Technologies, and uses CryptSurvey there.
Built by Opaix together with Circuit Analytics GmbH.
Visit cryptsurvey.io
Under the hood
Proven cryptographic methods
Shamir's Secret Sharing
Splits a secret into shares so that only a defined threshold can reconstruct it.
Yao's Garbled Circuits
Lets two parties evaluate a function jointly without revealing their inputs.
Homomorphic Encryption
Enables computation directly on encrypted data.
Zero-Knowledge Proofs (ZKPs)
Prove a statement is true without revealing the underlying data.
Research & ecosystems
We develop MPC within EU- and German-funded projects. It is designed to be integrable with the Eclipse Data Space Connector (EDC) for sovereign, standardized data exchange, an integration we are actively working on in these projects.
SC4EU
Supply Chains for Europe, secure data sharing for semiconductor resilience.
GAIA-X 4 PLC
Federated, sovereign data infrastructure for the product life cycle.
Semiconductor-X
Data ecosystems and standards for the semiconductor domain.
FAQ
Common questions
Does anyone ever see our raw data?
No. Your data is split into encrypted shares before it leaves your systems, and those shares are processed across independent nodes. No single node, and not Opaix, ever reconstructs a complete individual input. Only the agreed result is revealed.
What can we compute together?
Aggregates such as sums, averages, minimums, maximums and benchmarks across partners. Typical uses are salary and cost benchmarking, demand and capacity signals, and shared KPIs, all without exposing any individual figure.
How is privacy actually guaranteed?
By cryptography, not just policy. Individual responses are never reconstructed (zero knowledge) and only aggregates are kept (zero retention), so the guarantee rests on mathematical proofs rather than trust in a provider.
Where does this run?
Within trusted digital ecosystems with EU data residency. The encrypted shares are processed across independent nodes, so no single party ever holds a complete input.
Can we try MPC for surveys today?
Yes. CryptSurvey, built by Opaix together with Circuit Analytics GmbH, lets you run privacy-preserving surveys and benchmarks powered by MPC.