SRI International has secured an $11.5 million contract with the Defense Advanced Research Projects Agency to develop a chip-based technology in an effort to accelerate data analytics, artificial intelligence training and other processes that require a large amount of computing power.
The company said Monday it will create a hardware accelerator chip that allows for fully homomorphic encryption, which has the potential to speed up data-heavy computing processes such as machine learning.
FHE technology allows encrypted data to be computed, eliminates the need for decryption and reduces exposure risks, but produces noise that may render the data undecryptable. SRI will address this issue through a new architecture for central processing units and corresponding software.
Work under the contract supports the Data Protection in Virtual Environments or DPRIVE program, a larger effort to deliver a hardware accelerator that significantly reduces FHE's processing overhead.
“Creating a new hardware accelerator for FHE encrypted data is a unique technical challenge that requires expertise in co-processor architectures, hardware design, computer-aided verification of hardware, software, mathematics and FHE algorithms,“ said Karim Eldefrawy, principal computer scientist for SRI International's Computer Science Lab.