Related papers: Accelerating Cloud-Based Transcriptomics: Performa…

This paper presents a methodology to efficiently explore the design space of communication adapters. In most digital signal processing (DSP) applications, the overall performance of the system is significantly affected by communication…

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…

Large Language Model (LLM) inference has emerged as a fundamental paradigm, however, variations in output length cause severe workload imbalance in the decode phase, particularly for long-output reasoning tasks. Existing systems, such as PD…

Orchestrating service-oriented workflows is typically based on a design model that routes both data and control through a single point - the centralised workflow engine. This causes scalability problems that include the unnecessary…

Aligning large language models (LLMs) with human values is crucial for safe deployment. Inference-time techniques offer granular control over generation; however, they rely on model uncertainty, meaning an internal estimate of how likely…

The rapid development of cloud-native architecture has promoted the widespread application of container technology, but the optimization problems in container scheduling and resource management still face many challenges. This paper…

Recent innovations in Transformer-based large language models have significantly advanced the field of general-purpose neural language understanding and generation. With billions of trainable parameters, deployment of these large models…

This work introduces TrimTuner, the first system for optimizing machine learning jobs in the cloud to exploit sub-sampling techniques to reduce the cost of the optimization process while keeping into account user-specified constraints.…

We present STARFlow, a scalable generative model based on normalizing flows that achieves strong performance in high-resolution image synthesis. The core of STARFlow is Transformer Autoregressive Flow (TARFlow), which combines the…

Real-time embedded systems require precise timing and fault detection to ensure correct behavior. Traditional tracing tools often rely on local desktops with limited processing and storage capabilities, which hampers large-scale analysis.…

We propose an unsupervised adaptation framework, Self-TAught Recognizer (STAR), which leverages unlabeled data to enhance the robustness of automatic speech recognition (ASR) systems in diverse target domains, such as noise and accents.…

Modern large-scale computing systems distribute jobs into multiple smaller tasks which execute in parallel to accelerate job completion rates and reduce energy consumption. However, a common performance problem in such systems is dealing…

Real-world ecommerce recommender systems must deliver relevant items under strict tens-of-milliseconds latency constraints despite challenges such as cold-start products, rapidly shifting user intent, and dynamic context including…

Recent literature including our past work provide analysis and solutions for using (i) erasure coding, (ii) parallelism, or (iii) variable slicing/chunking (i.e., dividing an object of a specific size into a variable number of smaller…

Cloud platforms are increasing their emphasis on sustainability and reducing their operational carbon footprint. A common approach for reducing carbon emissions is to exploit the temporal flexibility inherent to many cloud workloads by…

Large Language Model (LLM) inference on large-scale systems is expected to dominate future cloud infrastructures. Efficient LLM inference in cloud environments with numerous AI accelerators is challenging, necessitating extensive…

We present a case study of a cloud-based computational workflow for processing large astronomical data sets from the Murchison Widefield Array (MWA) cosmology experiment. Cloud computing is well-suited to large-scale, episodic computation…

Runtime failure and performance degradation is commonplace in modern cloud systems. For cloud providers, automatically determining the root cause of incidents is paramount to ensuring high reliability and availability as prompt fault…

In this article we look at the potential of cloud containers and we provide some guidelines for companies and organisations that are starting to look at how to migrate their legacy infrastructure to something modern, reliable and scalable.…

We introduce Tuna, a static analysis approach to optimizing deep neural network programs. The optimization of tensor operations such as convolutions and matrix multiplications is the key to improving the performance of deep neural networks.…