A new compute-in-memory design has cleared 28nm validation, supporting post-quantum cryptography while reducing data movement, latency and power demands for secure edge and embedded applications.
BTQ Technologies and the Industrial Technology Research Institute (ITRI) have validated the QCIM compute-in-memory chip architecture in a 28nm test environment, marking a key step towards hardware designed to accelerate post-quantum cryptography. The architecture is intended to support both conventional and post-quantum cryptographic functions in a compact, low-power design.
Testing covered algorithms associated with FIPS 203, 204 and 205, which form part of the US National Institute of Standards and Technology’s emerging post-quantum cryptography standards. The architecture is also designed to accommodate additional algorithms as security requirements and standards develop.
At the heart of the approach is the integration of cryptographic operations within the memory subsystem, rather than repeatedly moving data between separate processing and memory components. This is intended to reduce data movement, latency and power consumption, potentially making the architecture suitable for applications where efficient security processing is important.
BTQ describes QCIM as a soft intellectual-property cryptographic accelerator, giving system designers flexibility to support different cryptographic functions as requirements change. The architecture is also designed to be “crypto-agile”, allowing it to adapt to evolving security standards.
The development involves three areas of semiconductor expertise. BTQ contributes its cryptographic architecture, while ICTK provides secure semiconductor and physical unclonable-function capabilities. ITRI contributes semiconductor design, integration and validation expertise.
The companies said the completed milestone reduces technical uncertainty surrounding the architecture and establishes a foundation for broader system integration. The next phase will focus on module-level integration, verification and validation, examining how the QCIM core can be incorporated into larger chip and system architectures while maintaining performance, interoperability and functional correctness.
The technology is being targeted at applications including edge computing, industrial systems, secure elements and AI-enabled devices. BTQ also expects to provide test chips to selected customers and strategic partners for performance and functional validation.
