Related papers: Understanding and Improving the Latency of DRAM-Ba…
This article summarizes the idea of "refresh-access parallelism," which was published in HPCA 2014, and examines the work's significance and future potential. The overarching objective of our HPCA 2014 paper is to reduce the significant…
Energy consumption dictates the cost and environmental impact of deploying Large Language Models. This paper investigates the impact of on-chip SRAM size and operating frequency on the energy efficiency and performance of LLM inference,…
Solid-state drives (SSDs) have revolutionized data storage with their high performance, energy efficiency, and reliability. However, as storage demands grow, SSDs face critical challenges in scalability, endurance, latency, and security.…
The continuing advancement of memory technology has not only fueled a surge in performance, but also substantially exacerbate reliability challenges. Traditional solutions have primarily focused on improving the efficiency of protection…
Today's systems have diverse needs that are difficult to address using one-size-fits-all commodity DRAM. Unfortunately, although system designers can theoretically adapt commodity DRAM chips to meet their particular design goals (e.g., by…
The efficiency of Large Language Model~(LLM) inference is often constrained by substantial memory bandwidth and capacity demands. Existing techniques, such as pruning, quantization, and mixture of experts/depth, reduce memory capacity…
Emerging non-volatile main memory (NVRAM) technologies provide byte-addressability, low idle power, and improved memory-density, and are likely to be a key component in the future memory hierarchy. However, a critical challenge in achieving…
Due to the scaling problem of the DRAM technology, non-volatile memory devices, which are based on different principle of operation than DRAM, are now being intensively developed to expand the main memory of computers. Disaggregated memory…
Many modern workloads, such as neural networks, databases, and graph processing, are fundamentally memory-bound. For such workloads, the data movement between main memory and CPU cores imposes a significant overhead in terms of both latency…
It is often said that one of the biggest limitations on computer performance is memory bandwidth (i.e."the memory wall problem"). In this position paper, I argue that if historical trends in computing evolution (where growth in available…
NAND flash memory is ubiquitous in everyday life today because its capacity has continuously increased and cost has continuously decreased over decades. This positive growth is a result of two key trends: (1) effective process technology…
We measured the energy efficiency of information erasure using silicon DRAM cells capable of counting charges on capacitors at the single-electron level. Our measurements revealed that the efficiency decreased as the erasure error…
The memory subsystem has always been a bottleneck in performance as well as significant power contributor in memory intensive applications. Many researchers have presented multi-layered memory hierarchies as a means to design energy and…
This paper summarizes our work on experimentally characterizing, mitigating, and recovering read disturb errors in multi-level cell (MLC) NAND flash memory, which was published in DSN 2015, and examines the work's significance and future…
Memory bandwidth is critical in today's high performance computing systems. The bandwidth is particularly paramount for GPU workloads such as 3D Gaming, Imaging and Perceptual Computing, GPGPU due to their data-intensive nature. As the…
Many modern workloads such as neural network inference and graph processing are fundamentally memory-bound. For such workloads, data movement between memory and CPU cores imposes a significant overhead in terms of both latency and energy. A…
Hybrid memory systems comprised of dynamic random access memory (DRAM) and non-volatile memory (NVM) have been proposed to exploit both the capacity advantage of NVM and the latency and dynamic energy advantages of DRAM. An important…
Recent nano-technological advances enable the Monolithic 3D (M3D) integration of multiple memory and logic layers in a single chip, allowing for fine-grained connections between layers and significantly alleviating main memory bottlenecks.…
Lattice Boltzmann method (LBM) is a promising approach to solving Computational Fluid Dynamics (CFD) problems, however, its nature of memory-boundness limits nearly all LBM algorithms' performance on modern computer architectures. This…
The increasing adoption of large language models (LLMs) necessitates inference serving systems that can deliver both high throughput and low latency. Deploying LLMs with hundreds of billions of parameters on memory-constrained GPUs exposes…