Related papers: Old is Gold: Optimizing Single-threaded Applicatio…
Memory allocation is a fairly mature field of computer science. However, we challenge a prevailing assumption in the literature over the last 50 years which, if reconsidered, necessitates a fundamental reevaluation of many classical memory…
Edge Computing (EC) allows users to access computing resources at the network frontier, which paves the way for deploying delay-sensitive applications such as Mobile Augmented Reality (MAR). Under the EC paradigm, MAR users connect to the…
IoT applications increasingly rely on on-device AI accelerators to ensure high performance, especially in low-connectivity and safety-critical scenarios. However, the limited on-chip memory of these accelerators forces inference runtimes to…
We consider the problem of jointly optimizing users' offloading decisions, communication and computing resource allocation in a sliced multi-cell mobile edge computing (MEC) network. We minimize the weighted sum of the gap between the…
Since very few contributions to the development of an unified memory orchestration framework for efficient management of both host and remote idle memory have been made, we present Valet, an efficient approach to orchestration of host and…
Future multiprocessor chips will integrate many different units, each tailored to a specific computation. When designing such a system, the chip architect must decide how to distribute limited system resources such as area, power, and…
Transformer networks, driven by self-attention, are central to Large Language Models. In generative Transformers, self-attention uses cache memory to store token projections, avoiding recomputation at each time step. However, GPU-stored…
The growing performance gap between multi-core CPUs and main memory necessitates hardware-aware software design paradigms. This study provides a comprehensive performance analysis of Data Oriented Design (DOD) versus the traditional…
Large Language Models (LLMs) based on autoregressive, decoder-only Transformers generate text one token at a time, where a token represents a discrete unit of text. As each newly produced token is appended to the partial output sequence,…
PIM architectures aim to reduce data transfer costs between processors and memory by integrating processing units within memory layers. Prior PIM architectures have shown potential to improve energy efficiency and performance. However, such…
Relaxed retention (or volatile) spin-transfer torque RAM (STT-RAM) has been widely studied as a way to reduce STT-RAM's write energy and latency overheads. Given a relaxed retention time STT-RAM level one (L1) cache, we analyze the impacts…
Cloud computing is a new archetype that provides dynamic computing services to cloud users through the support of datacenters that employs the services of datacenter brokers which discover resources and assign them Virtually. The focus of…
This paper introduces REDC, a comprehensive strategy for offloading computational tasks within mobile Edge Networks (EN) to Distributed Computing (DC) after Rateless Encoding (RE). Despite the efficiency, reliability, and scalability…
Spiking Neural Networks (SNNs) are bio-plausible models that hold great potential for realizing energy-efficient implementations of sequential tasks on resource-constrained edge devices. However, commercial edge platforms based on standard…
With the increasing popularity of recommendation systems (RecSys), the demand for compute resources in datacenters has surged. However, the model-wise resource allocation employed in current RecSys model serving architectures falls short in…
Industries are recently considering the adoption of cloud computing for hosting safety critical applications. However, the use of multicore processors usually adopted in the cloud introduces temporal anomalies due to contention for shared…
The growing memory footprints of cloud and big data applications mean that data center CPUs can spend significant time waiting for memory. An attractive approach to improving performance in such centralized compute settings is to employ…
Performance of distributed optimization and learning systems is bottlenecked by "straggler" nodes and slow communication links, which significantly delay computation. We propose a distributed optimization framework where the dataset is…
We present MaxMem, a tiered main memory management system that aims to maximize Big Data application colocation and performance. MaxMem uses an application-agnostic and lightweight memory occupancy control mechanism based on fast memory…
Software-defined networks (SDNs) are a huge evolution in simplifying implementation and network operation which have reduced costs and made the network programmable. Although SDNs are a suitable option for solving some of the previous…