Search papers, labs, and topics across Lattice.
This paper introduces probabilistic memory (p-MEM), a novel memory primitive designed to enhance the efficiency of probabilistic computation in edge intelligence by storing distribution parameters and allowing direct sampling at high throughput. The authors demonstrate that p-MEM can achieve over 1000 GSa/s/mm虏 in Gaussian random number generation throughput while significantly reducing instruction counts, sampling latency, and energy consumption in Bayesian neural network workloads. These advancements address the critical data throughput gap in probabilistic computation, making it feasible to implement trustworthy AI systems at the edge.
p-MEM achieves over 1000 GSa/s/mm虏 throughput, drastically cutting down latency and energy use for probabilistic AI applications.
Probabilistic computation plays an important role in trustworthy edge intelligence to quantify uncertainty, enhance robustness, reconstruct data, and protect privacy, but its adoption is limited by the orders-of-magnitude data throughput gap between Gaussian random number generation (GRNG) and computation, as well as instruction overhead. This paper introduces probabilistic memory (p-MEM), a unified memory primitive that stores distribution parameters, such as mean and standard deviation, and samples directly at the native memory bandwidth, where deterministic data becomes the zero-variance special case. Using a layout-validated p-MEM simulator, we comprehensively explore device choices, memory specifications, and technology nodes, showing that p-MEM can achieve more than 1000 GSa/s/mm^2 GRNG throughput, including memory-array access. Integrated into CPU/GPU systems, p-MEM reduces instruction count by up to 2.19x/4.37x, sampling latency by 562x/3.45x, and energy by 295.5x/3.53x for Bayesian neural network workloads, providing a scalable hardware substrate for trustworthy probabilistic AI.
