Related papers: AXI-Pack: Near-Memory Bus Packing for Bandwidth-Ef…

The existence of considerable amount of redundancy in the Internet traffic at the packet level has stimulated the deployment of packet-level redundancy elimination techniques within the network by enabling network nodes to memorize data…

Personalized recommendation systems leverage deep learning models and account for the majority of data center AI cycles. Their performance is dominated by memory-bound sparse embedding operations with unique irregular memory access patterns…

Packet trimming is a primitive that has been proposed for datacenter networks: to minimize latency, switches run small queues; when the queue overflows, rather than dropping packets the switch trims off the packet payload and either…

A growing trend in modern data analysis is the integration of data management with learning, guided by accuracy, latency, and cost requirements. In practice, applications draw data of different formats from many sources. In the meanwhile,…

Emerging interconnects, such as CXL and NVLink, have been integrated into the intra-host topology to scale more accelerators and facilitate efficient communication between them, such as GPUs. To keep pace with the accelerator's growing…

Approximate nearest neighbor (ANN) search in AI systems increasingly handles sensitive data on third-party infrastructure. Trusted execution environments (TEEs) offer protection, but cost-efficient deployments must rely on external SSDs,…

Point clouds are increasingly important in intelligent applications, but frequent off-chip memory traffic in accelerators causes pipeline stalls and leads to high energy consumption. While conventional line buffer techniques can eliminate…

In most modern systems, the memory subsystem is managed and accessed at multiple different granularities at various resources. We observe that such multi-granularity management results in significant inefficiency in the memory subsystem.…

We present HeartStream, a 64-RV-core shared-L1-memory cluster (410 GFLOP/s peak performance and 204.8 GBps L1 bandwidth) for energy-efficient AI-enhanced O-RAN. The cores and cluster architecture are customized for baseband processing,…

The exact Markov modeling analysis of erasure networks with finite buffers is an extremely hard problem due to the large number of states in the system. In such networks, packets are lost due to either link erasures or blocking by the full…

Over-the-air computation (AirComp) enables fast wireless data aggregation at the receiver through concurrent transmission by sensors in the application of Internet-of-Things (IoT). To further improve the performance of AirComp under…

Specialized hardware like application-specific integrated circuits (ASICs) remains the primary accelerator type for cryptographic kernels based on large integer arithmetic. Prior work has shown that commodity and server-class GPUs can…

In the landscape of High-Performance Computing (HPC), the quest for efficient and scalable memory solutions remains paramount. The advent of Compute Express Link (CXL) introduces a promising avenue with its potential to function as a…

With the growing number of data-intensive workloads, GPU, which is the state-of-the-art single-instruction-multiple-thread (SIMT) processor, is hindered by the memory bandwidth wall. To alleviate this bottleneck, previously proposed…

This paper try to better understand the performance of ContikiMAC compared to X-MAC. ContikiMAC achieves a transmission by repeatedly transmitting a data packet until the reception of an ACK from the destination. While X-MAC uses a stream…

While Transformers are dominated by Floating-Point (FP) Matrix-Multiplications, their aggressive acceleration through dedicated hardware or many-core programmable systems has shifted the performance bottleneck to non-linear functions like…

The AC Optimal Power Flow (AC-OPF) problem is central to power system operation but challenging to solve efficiently due to its nonconvex and nonlinear nature. Neural networks (NNs) offer fast surrogates, yet their black-box behavior raises…

Edge applications generate a large influx of sensor data on massive scales, and these massive data streams must be processed shortly to derive actionable intelligence. However, traditional data processing systems are not well-suited for…

As computing demand and memory footprint of deep learning applications accelerate, clusters of cores sharing local (L1) multi-banked memory are widely used as key building blocks in large-scale architectures. When the cluster's core count…

The substantial memory requirements of Large Language Models (LLMs), particularly for long-context fine-tuning, have renewed interest in CPU offloading to augment limited GPU memory. However, as context lengths grow, relying on CPU memory…