AI & ML interests
Multiplier-free balanced ternary ({−1, 0, +1}) arithmetic for BitNet b1.58 inference on commodity FPGAs. We build the open hardware (CERN-OHL-S v2) and distil the ternary-weight students that run on it — silicon-validated on a low-cost Artix-7, with whole-board energy measured rather than estimated.
Recent Activity
View all activity
Ternarycore 's datasets
None public yet