It has been a long time that computer architecture and systems are optimized for efficient
execution of machine learning (ML) models. Now, it is time to reconsider the relationship …

Public cloud providers seek to meet stringent performance requirements and low hardware
cost. A key driver of performance and cost is main memory. Memory pooling promises to …

Past research has proposed numerous hardware prefetching techniques, most of which rely
on exploiting one specific type of program context information (eg, program counter …

We present Decoupled Vector Runahead (DVR), an in-core prefetching technique,
executing separately to the main application thread, that exploits massive amounts of …

Prefetching which predicts future memory accesses and preloads them from main memory,
is a widely-adopted technique to overcome the processor-memory performance gap …

Machine learning algorithms have shown potential to improve prefetching performance by
accurately predicting future memory accesses. Existing approaches are based on the …

Integrating compute express link (CXL) with SSDs allows scalable access to large memory
but has slower speeds than DRAMs. We present ExPAND, an expander-driven CXL …

Online Reinforcement Learning (RL) has been adopted as an effective mechanism in
various decision-making problems in microarchitecture. Its high adaptability and the ability to …

Memory disaggregation is a promising direction to mitigate memory contention in
datacenters. To make memory disaggregation practical, prior efforts expose remote memory …

Data prefetching is a technique that plays a crucial role in modern high-performance
processors by hiding long latency memory accesses. Several state-of-the-art hardware …