Related papers: PreSto: An In-Storage Data Preprocessing System fo…
Understanding user interests is crucial for Click-Through Rate (CTR) prediction tasks. In sequential recommendation, pre-training from user historical behaviors through self-supervised learning can better comprehend user dynamic…
When IP-packet processing is unconditionally carried out on behalf of an operating system kernel thread, processing systems can experience overload in high incoming traffic scenarios. This is especially worrying for embedded real-time…
Robotic bin packing is widely deployed in warehouse automation, with current systems achieving robust performance through heuristic and learning-based strategies. These systems must balance compact placement with rapid execution, where…
Training transformer models requires substantial GPU compute and memory resources. In homogeneous clusters, distributed strategies allocate resources evenly, but this approach is inefficient for heterogeneous clusters, where GPUs differ in…
Recent years have witnessed the unprecedented achievements of large-scale pre-trained models, especially the Transformer models. Many products and services in Tencent Inc., such as WeChat, QQ, and Tencent Advertisement, have been opted in…
This paper proposes a reinforcement learning-based method for microservice resource scheduling and optimization, aiming to address issues such as uneven resource allocation, high latency, and insufficient throughput in traditional…
Efficiency in instruction fetching is critical to performance, and this requires the primary structures--L1 instruction caches (L1i), branch target buffers (BTB) and instruction TLBs (iTLB)--to have the requisite information when needed.…
There has been much interest in deploying deep learning algorithms on low-powered devices, including smartphones, drones, and medical sensors. However, full-scale deep neural networks are often too resource-intensive in terms of energy and…
The server central processing unit (CPU) market continues to exhibit robust demand due to the rising global need for computing power. Against this backdrop, CPU benchmark performance prediction is crucial for architecture designers. It…
Many state-of-the-art Segment Routing (SR) Traffic Engineering (TE) algorithms rely on Linear Program (LP)-based optimization. However, the poor scalability of the latter and the resulting high computation times impose severe restrictions…
Large-scale deep learning workloads increasingly suffer from I/O bottlenecks as datasets grow beyond local storage capacities and GPU compute outpaces network and disk latencies. While recent systems optimize data-loading time, they…
Deep learning models are widely used across computer vision and other domains. When working on the model induction, selecting the right architecture for a given dataset often relies on repetitive trial-and-error procedures. This procedure…
Click-Through Rate (CTR) prediction on cold users is a challenging task in recommender systems. Recent researches have resorted to meta-learning to tackle the cold-user challenge, which either perform few-shot user representation learning…
Fine-tuning large pre-trained models on downstream tasks has been adopted in a variety of domains recently. However, it is costly to update the entire parameter set of large pre-trained models. Although recently proposed parameter-efficient…
While other areas of machine learning have seen more and more automation, designing a high-performing recommender system still requires a high level of human effort. Furthermore, recent work has shown that modern recommender system…
Distributed training using multiple devices (e.g., GPUs) has been widely adopted for learning DNN models over large datasets. However, the performance of large-scale distributed training tends to be far from linear speed-up in practice.…
Transformer architectures serve as the backbone for most modern Large Language Models, therefore their pretraining stability and convergence speed are of central concern. Motivated by the logical dependency of sequentially stacked layers,…
Computing has a huge memory problem. The memory system, consisting of multiple technologies at different levels, is responsible for most of the energy consumption, performance bottlenecks, robustness problems, monetary cost, and hardware…
Critic-free reinforcement learning methods, particularly group policies, have attracted considerable attention for their efficiency in complex tasks. However, these methods rely heavily on multiple sampling and comparisons within the policy…
Today's systems are overwhelmingly designed to move data to computation. This design choice goes directly against at least three key trends in systems that cause performance, scalability and energy bottlenecks: (1) data access from memory…