Search papers, labs, and topics across Lattice.
4
0
5
VIPER slashes design-space exploration time from hours to under a minute while achieving less than 10% error in performance predictions for Processing-in-Memory systems.
p-MEM achieves over 1000 GSa/s/mm虏 throughput, drastically cutting down latency and energy use for probabilistic AI applications.
Bayesian decision trees get a 100x speedup and 10,000x energy efficiency boost thanks to a novel hardware platform leveraging ferroelectric FETs.
Finally, a way to make Retrieval-Augmented Generation (RAG) on edge devices practical: CQ-CiM jointly compresses and quantizes embeddings to fit the constraints of diverse Compute-in-Memory (CiM) architectures.
