Decentriq v3: Confidential Data Clean Rooms for scale
By utilizing Confidential Computing technology from AMD, we are able to provide a safe and neutral space for sensitive data collaboration that meets the demands of the modern data enterprise without any compromises.
Today we wrap up the successful preview of our biggest release yet with some of our early users, making it generally available for everyone.
The v3 of our Data Clean Rooms is marking the release of the world’s first analytics platform deployed in production properly using AMD SEV-SNP. We pair this new technology with our existing Confidential Computing setup to offer the most scalable data analysis platform using encryption in-use. For the first time, data scientists can use their existing Machine Learning workloads and frameworks to collaborate on production-size datasets with Confidential Computing guarantees.
A first for Confidential Computing, powered by AMD
We are proud to announce that we have successfully integrated AMD's new Confidential Computing technology, known as AMD SEV-SNP, into our platform. This marks a significant milestone, as our platform is the first fully attestable deployment of AMD SEV-SNP. By enabling this technology, we are able to scale a wider range of use cases to meet the high standards of production-level scalability, reliability, and resilience that our customers expect. Furthermore, our users are now able to independently attest to our platform, eliminating the need for them to trust us with their sensitive data. This achievement required us to re-examine and innovate upon some fundamental aspects of computing.
Enhanced Resource Distribution and Reduced Latency
In addition to the AMD SEV-SNP technology, we have also enhanced our resource distribution system to reduce latency and allow for faster computation speeds, even when the platform is heavily used. We have revamped the internals of our platform to make it future-proof for upcoming feature development, ensuring that Decentriq remains a reliable and evolving platform for Confidential Computing analytics.
Larger Dataset Support and Improved Synthetic Data Generation
We have added the ability to provision Gigabytes of data directly in the Decentriq UI and run computations on larger workloads. Our synthetic data generation feature has also been improved, giving you the option to choose between differential privacy-based or standard models to synthesize your dataset. The standard model has better pairwise correlations and supports larger datasets.
S3 Bucket Integration and enhanced security
You can now upload your computation results directly from the Data Clean Room to your S3 bucket, following the strongest security standards. A Data Clean Room password can also be set, requiring knowledge of the password for any interaction with the Data Clean Room.
New Documentation page and data residency options
To help you get the most out of our platform, we have revamped our documentation page with step-by-step guides and advanced API references. Additionally, all sensitive data is now hosted and processed exclusively in Switzerland, with auxiliary services based in the EU only.
For a more detailed view of what the v3 release brings you can see our full release notes here.
At Decentriq, trust is at the very core of who we are and what we do. That's why the release of version 3 of our Data Clean Rooms is so significant for us. By utilizing Confidential Computing technology from AMD, we are able to provide a safe and neutral space for sensitive data collaboration that meets the demands of the modern data enterprise without any compromises.
Our commitment to trust is reflected in every aspect of our platform, from its scalability and performance to its ability to safeguard sensitive information. We understand that trust is essential in today's data-driven world, which is why we have made it our mission to provide a reliable and secure environment for data collaboration.
Subscribe to Decentriq
Stay connected with Decentriq. Receive email notifications about industry news and product updates.