RAG-Enabled Large Language Model Inference Solution Utilizes NVIDIA H100 Tensor Core GPUs
SAN FRANCISCO, May 7, 2024 /PRNewswire/ — RSA Conference 2024 – Inpher, pioneers in privacy-enhanced computation, today announced the general availability release of SecurAI, a leading solution that protects the privacy and security of user inputs on large language models. This enterprise-ready release of SecurAI leverages the NVIDIA H100 Tensor Core GPU for maximum speed and performance.
“Enterprises need to harness the power of generative AI while seamlessly and securely incorporating it into their business processes,” stated Daniel Rohrer, vice president of Product Security at NVIDIA. “Inpher’s SecurAI coupled with NVIDIA confidential computing technology enables organizations to use confidential GPUs for improved data utility and business productivity.”
Powered by confidential GPUs, this much anticipated release of SecurAI also integrates retrieval-augmented generation (RAG). SecurAI ensures that user prompts are never exposed to the model provider while enabling organizations to improve model outputs by seamlessly integrating with core business services, data sources and applications such as Slack, Zendesk, Salesforce, Snowflake, et al. SecurAI also enables enterprise users to upload their own files based on their organizational needs by leveraging secure RAG.
“Security is a critical element for every enterprise to consider when building a generative AI deployment strategy,” stated Dr. Jordan Brandt, chief executive officer and cofounder at Inpher. “Our commitment to privacy and the responsible use of AI is foundational to our team and by working closely with NVIDIA we deliver the most robust, enterprise-ready generative AI solution.”
Inpher SecurAI addresses the privacy and security concerns that users have when trying to leverage AI and open models where prompts might contain sensitive data that needs to be protected from the model and/or cloud service provider. The intricate details of the prompts received and the subsequent data processing occur within a Trusted Execution Environment that is isolated from the underlying infrastructure. This isolation shields sensitive information, ensuring that the specifics of user inputs, such as, proprietary business data, personal details, or other confidential information, remain inaccessible to external entities, including SecurAI’s own hosting infrastructure.
By keeping the intricate details of user inputs and data processing within a protected space, SecurAI significantly reduces the risk of data breaches, leaks, or unauthorized access, making it a strong choice for safeguarding IP and privacy. SecurAI’s distinct focus on data protection and privacy offers a secure choice for organizations emphasizing the privacy of their data and intellectual property. With SecurAI, users can incorporate large language models into their organization without compromising privacy.
To begin harnessing the power of AI for increased productivity and data utility and for more information on Inpher’s SecurAI solution for the privacy-preserving use of LLMs, including RAG systems, click here.
About Inpher
Inpher, Inc. is the leader in privacy-enhancing computation that empowers organizations to harness the potential of AI securely and with trust. Inpher makes privacy-preserving machine learning easy, enabling organizations to unlock the value of sensitive data across teams, organizations, and geographies without moving it. For organizations seeking to leverage new generative AI technologies such as ChatGPT, Inpher SecurAI ensures privacy and compliance. Founded by world-renowned cryptographers and engineers, Inpher has long been recognized as a pioneer in the fields of secure Multiparty Computation (MPC), Fully Homomorphic Encryption (FHE), Federated Learning (FL), and other combinations of privacy-enhancing technologies (PETs). Inpher continues to deliver enterprise ready capabilities and real world examples.
For more information on Inpher, please visit us at www.inpher.io and follow us on LinkedIn and Twitter.
SOURCE Inpher
Wanda Parisien is a computing expert who navigates the vast landscape of hardware and software. With a focus on computer technology, software development, and industry trends, Wanda delivers informative content, tutorials, and analyses to keep readers updated on the latest in the world of computing.
