Architecture overviews, security and certification briefs, performance benchmarks, and integration guides for QCIM cryptographic IP.

Brief
Security
In-depth look into emerging threats and how hardware solutions need to respond.

Datasheet
Performance
Technical deep dive into the innovation behind QCIM and its measured performance numbers.
QCIM executes cryptographic algorithms as versioned kernels inside a single compute-in-memory block. New classical, post-quantum, and future algorithms are deployed as authorized kernel updates, so hardware stays constant while cryptographic capability expands.
Protection is a property of the architecture rather than of an individual algorithm implementation: configurable hardware masking, a two-share masked data interface, SECDED transport protection, fault detection, and zeroization paths apply to every algorithm the engine runs.
As a 32-bit AHB-Lite target with a 256 KB address aperture and single-clock integration. It ships in accelerator mode against an existing host, in subsystem mode with an integrated Ibex RV32 controller and firmware loader, and as an end-to-end reference platform for the ChipWhisperer CW340.
The current benchmark configuration uses a 55,000-gate cryptographic core, with a configurable FPGA memory backend using inferred BRAM and synthesizable RTL supporting ASIC SRAM macros.
QCIM Engine is available now as soft IP: synthesizable RTL licensed for integration into your own ASIC or FPGA, rather than a packaged part. QCIM RoT, developed with ICTK, is in development, as is QCIM Platform Security. Roadmap algorithms include ECDSA P-256, HAETAE, HQC, and LMS/XMSS.