性能
A lot of deep learning applications are desired to be run on mobile devices. Both accuracy and inference time are meaningful for a lot of them. While the number of FLOPs is usually used as a proxy for neural network latency, it may be not…
Length Rate Quotient (LRQ) is the first algorithm of interleaved shaping -- a novel concept proposed to provide per-flow shaping for a flow aggregate without per-flow queuing. This concept has been adopted by Time-Sensitive Networking (TSN)…
Most large web-scale applications are now built by composing collections (from a few up to 100s or 1000s) of microservices. Operators need to decide how many resources are allocated to each microservice, and these allocations can have a…
While state-of-the-art development in CNN topology, such as VGGNet and ResNet, have become increasingly accurate, these networks are computationally expensive involving billions of arithmetic operations and parameters. To improve the…
The development of the A64FX processor by Fujitsu has been a massive innovation in vectorized processors and led to Fugaku: the current world's fastest supercomputer. We use a variety of tools to analyze the behavior and performance of…
Consolidating latency-critical (LC) and best-effort (BE) tenants at storage backend helps to increase resources utilization. Even if tenants use dedicated queues and threads to achieve performance isolation, threads are still contend for…
Consider a message processing system whose objective is to produce the most current information as measured by the quantity known as "age of information". We have argued in previous papers that if we are allowed to design the message…
Caching systems have long been crucial for improving the performance of a wide variety of network and web based online applications. In such systems, end-to-end application performance heavily depends on the fraction of objects transferred…
Analyzing performance within asynchronous many-task-based runtime systems is challenging because millions of tasks are launched concurrently. Especially for long-term runs the amount of data collected becomes overwhelming. We study HPX and…
Premier cloud service providers (CSPs) offer two types of purchase options, namely on-demand and spot instances, with time-varying features in availability and price. Users like startups have to operate on a limited budget and similarly…
The First-Come First-Served (FCFS) scheduling policy is the most popular scheduling algorithm used in practice. Furthermore, its usage is theoretically validated: for light-tailed job size distributions, FCFS has weakly optimal asymptotic…
Deficit Round-Robin (DRR) is a widespread scheduling algorithm that provides fair queueing with variable-length packets. Bounds on worst-case delays for DRR were found by Boyer et al., who used a rigorous network calculus approach and…
We analyze a tandem network of polling queues with two product types and two stations. We assume that external arrivals to the network follow a Poisson process, and service times at each station are exponentially distributed. For this…
Edge video analytics is becoming the solution to many safety and management tasks. Its wide deployment, however, must first address the tension between inference accuracy and resource (compute/network) cost. This has led to the development…
We consider a distributed storage system which stores several hot (popular) and cold (less popular) data files across multiple nodes or servers. Hot files are stored using repetition codes while cold files are stored using erasure codes.…
In this paper, we provide (i) a rigorous general theory to elicit conditions on (tail-dependent) heavy-tailed cyber-risk distributions under which a risk management firm might find it (non)sustainable to provide aggregate cyber-risk…
In this paper, we exploit results obtained in an earlier study for the Laplace transform of the sojourn time $\Omega$ of an entire batch in the $M^{[X]}/M/1$ Processor Sharing (PS) queue in order to derive the asymptotic behavior of the…
We consider the c.o.c. redundancy system with $N$ parallel servers where incoming jobs are immediately replicated to $d$ servers chosen uniformly at random (without replacement). A job finishes service as soon as the first replica is…
We show how to calculate the Age of Information in an overtake-free network of quasi-reversible queues, with exponential exogenous interarrivals of multiple classes of update packets and exponential service times at all nodes. Results are…
The development of NekRS, a GPU-oriented thermal-fluids simulation code based on the spectral element method (SEM) is described. For performance portability, the code is based on the open concurrent compute abstraction and leverages…