In daily life, our express information items are stored in the server of express company. To avoid the leakage of the users' private information, this paper proposes a fast encryption algorithm to protect them. Th...
详细信息
Efficient utilization of computing resources in a stream computing environment is crucial for system performance. Existing scheduling strategies can hardly guarantee latency under performance constraints, let alone ac...
详细信息
As the core hub for carrying various blockchain applications upstream and connecting network infrastructure downstream, blockchain infrastructure provides necessary storage, transmission, computing, development, and t...
详细信息
With the continuous development of China's power industry, regional energy distribution has gradually become an important way to promote social and economic operation. How to scientifically manage distributed ener...
详细信息
This research paper explores nuances of large language models along with a detailed analysis of various model types, and their associated advantages and drawbacks. The primary objective of this paper is to provide a b...
详细信息
ISBN:
(数字)9798331521349
ISBN:
(纸本)9798331521356
This research paper explores nuances of large language models along with a detailed analysis of various model types, and their associated advantages and drawbacks. The primary objective of this paper is to provide a brief knowledge about prevailing popular LLMs. This study further offers a thorough summary of the most recent survey articles that have been published on Large Language Models (LLMs) till date. Furthermore, a comparative study of the latest LLMs launched by popular organizations is also done, along with their evolution over a period of time. Challenges of LLMs are also outlined in this work.
parallelcomputing of real-time data is an effective solution to the problem of timely analysis with the rapid increase of real-time data in today's power grid. Currently, most data parallel processing methods do ...
详细信息
Large-scale deep learning models are widely deployed in many application domains with remarkable performance improvements. However, training these models with immense parameters calls for unprecedented computing and c...
详细信息
ISBN:
(纸本)9798350322255
Large-scale deep learning models are widely deployed in many application domains with remarkable performance improvements. However, training these models with immense parameters calls for unprecedented computing and communication capabilities. Recently, chiplet-based architectures have shown much promise in scaling Deep Neural Network (DNN) inference, but their applications in the training phase remain unexplored and challenging. In this paper, we posit, beyond scaling computing capability, chiplet-based architectures could also be leveraged to enable new optimization opportunities for existing parallel training algorithms (e.g., Ring and Tree-based all-reduce). Specifically, we aim to explore a variety of topological characteristics, along with the interposer technology, to sustain the performance scaling of parallel training in chiplet-based systems. We propose ARIES, a versatile chiplet-based communication architecture supporting various parallel training algorithms using a flexible interconnect design. The proposed design can adapt to various collective operations such as reduce and gather across a wide diversity of training algorithms. Moreover, such flexibility is also leveraged to further enhance existing all-reduce algorithms depending on the latency and bandwidth requirements of the DNN model and dataset size. Simulation results show that the proposed ARIES can achieve up to 3.92x speedup in execution time and 38.8% reduction in Network-on-Chip (NoC) energy consumption when compared to prior work.
VisionMaster algorithm development platform is a powerful machine vision software. The platform aims to provide users with efficient and convenient algorithm tools to quickly build visual applications and solve comple...
详细信息
The reliability of the power system can be improved by the inclusion of distributed Generation (DG) in the distribution network. The integration of DG into the power system can cause the malfunction of the existing pr...
详细信息
Electric Vehicles (EV) are increasingly popular due to environmental benefits but can strain the power grid. Optimizing EV charging in large parking lots is a challenge. This study uses a metaheuristic approach on a g...
详细信息
暂无评论