System-Level Performance Modeling of Photonic In-Memory Computing
🤖 AI Summary
This work addresses the lack of comprehensive system-level performance modeling for photonic in-memory computing, which has hindered realistic assessment under actual high-performance computing (HPC) workloads. The study presents the first performance model that incorporates system-level overheads—including off-chip memory accesses and electro-optical conversions—and integrates an algorithm-hardware co-mapping methodology. A 1×256-bit single-wavelength photonic SRAM array is implemented using a standard silicon photonics process (GlobalFoundries). Evaluated on representative HPC benchmarks—Sod shock tube, MTTKRP, and Vlasov-Maxwell equations—the system achieves 1.5, 0.9, and 1.3 TOPS, respectively, with an average energy efficiency of 2.5 TOPS/W, offering the first quantitative evaluation of photonic in-memory computing in real-world applications.
Application Category
📝 Abstract
Photonic in-memory computing is a high-speed, low-energy alternative to traditional transistor-based digital computing that utilizes high photonic operating frequencies and bandwidths. In this work, we develop a comprehensive system-level performance model for photonic in-memory computing, capturing the effects of key latency sources such as external memory access and opto-electronic conversion. We perform algorithm-to-hardware mapping across a range of workloads, including the Sod shock tube problem, Matricized Tensor Times Khatri-Rao Product (MTTKRP), and the Vlasov-Maxwell equation, to evaluate how the latencies impact real-world high-performance computing workloads. Our performance model shows that, while accounting for system overheads, a compact 1×256 bit single-wavelength photonic SRAM array, fabricated using the standard silicon photonics process by GlobalFoundries, sustains up to 1.5 TOPS, 0.9 TOPS, and 1.3 TOPS on the Sod shock tube problem, MTTKRP, and the Vlasov-Maxwell equation with an average energy efficiency of 2.5 TOPS/W.
Problem
Research questions and friction points this paper is trying to address.
photonic in-memory computing
system-level performance modeling
latency sources
high-performance computing workloads
opto-electronic conversion
Innovation
Methods, ideas, or system contributions that make the work stand out.
photonic in-memory computing
system-level performance modeling
opto-electronic conversion latency
silicon photonics
energy efficiency
🔎 Similar Papers
No similar papers found.
