相关论文: The APENEXT project

The best performing learning algorithms devised for event cameras work by first converting events into dense representations that are then processed using standard CNNs. However, these steps discard both the sparsity and high temporal…

Graph Neural Networks (GNNs) have recently gained attention due to their performance on non-Euclidean data. The use of custom hardware architectures proves particularly beneficial for GNNs due to their irregular memory access patterns,…

In this paper, we describe the architecture and performance of the GRAPE-6 system, a massively-parallel special-purpose computer for astrophysical $N$-body simulations. GRAPE-6 is the successor of GRAPE-4, which was completed in 1995 and…

PEGs are a formal grammar foundation for describing syntax, and are not hard to generate parsers with a plain recursive decent parsing. However, the large amount of C-stack consumption in the recursive parsing is not acceptable especially…

This work introduces novel training and post-training compression schemes to reduce external memory access during transformer model inference. Additionally, a new control flow mechanism, called dynamic batching, and a novel buffer…

In this paper, we present EdgeFace, a lightweight and efficient face recognition network inspired by the hybrid architecture of EdgeNeXt. By effectively combining the strengths of both CNN and Transformer models, and a low rank linear…

The explosive arrival of OpenAI's ChatGPT has fueled the globalization of large language model (LLM), which consists of billions of pretrained parameters that embodies the aspects of syntax and semantics. HyperAccel introduces latency…

In this paper we describe the present status and our plans for the realization of next phases of the CalcPHEP project aimed at the theoretical support of experiments at modern and future accelerators: TEVATRON, LHC, electron Linear…

In recent years, transformer-based models have shown state-of-the-art results for Natural Language Processing (NLP). In particular, the introduction of the BERT language model brought with it breakthroughs in tasks such as question…

In recent years, a great deal of attention has been paid to the Transformer network for speech recognition tasks due to its excellent model performance. However, the Transformer network always involves heavy computation and large number of…

Sparse tensor programs are essential in deep learning and graph analytics, driving the need for optimized processing. To meet this demand, specialized hardware accelerators are being developed. Optimizing these programs for accelerators is…

Live traffic analysis at the first aggregation point in the ISP network enables the implementation of complex traffic engineering policies but is limited by the scarce processing capabilities, especially for Deep Learning (DL) based…

Mixture-of-Experts (MoE) architectures offer the promise of larger model capacity without the prohibitive costs of fully dense designs. However, in real-world inference serving, load skew across experts often leads to suboptimal device…

The processor accelerators are effective because they are working not (completely) on principles of stored program computers. They use some kind of parallelism, and it is rather hard to program them effectively: a parallel architecture by…

Long Term Evolution-Advanced (LTE-Advanced) is the most recent mobile telecommunication technology proposed by 3GPP. LTE-Advanced is applied in some countries, but still in development and testing phase, because of that, a simulation model…

GPU architectures have become popular for executing general-purpose programs. Their many-core architecture supports a large number of threads that run concurrently to hide the latency among dependent instructions. In modern GPU…

We introduce ABEL, the Adaptable Beginning-to-End Linac simulation framework developed for agile design studies of plasma-based accelerators and colliders. ABEL's modular architecture allows users to simulate particle acceleration across…

Transformers have revolutionized deep learning and generative modeling to enable unprecedented advancements in natural language processing tasks and beyond. However, designing hardware accelerators for executing transformer models is…

The project, aimed at the theoretical support of experiments at modern and future accelerators -- TEVATRON, LHC, electron Linear Colliders (TESLA, NLC, CLIC) and muon factories, is presented. Within this project a four-level computer system…

This whitepaper proposes the design and adoption of a new generation of Tensor Processing Unit which has the performance of Google's TPU, yet performs operations on wide precision data. The new generation TPU is made possible by…