相关论文: swTVM: Towards Optimized Tensor Code Generation fo…

There is an increasing need to bring machine learning to a wide diversity of hardware devices. Current frameworks rely on vendor-specific operator libraries and optimize for a narrow range of server-class GPUs. Deploying workloads to new…

The growing adoption of domain-specific architectures in edge computing platforms for deep learning has highlighted the efficiency of hardware accelerators. However, integrating custom accelerators into modern machine learning (ML)…

A growing number of applications implement predictive functions using deep learning models, which require heavy use of compute and memory. One popular technique for increasing resource efficiency is 8-bit integer quantization, in which…

Subsequence Dynamic Time Warping (sDTW) is the metric of choice when performing many sequence matching and alignment tasks. While sDTW is flexible and accurate, it is neither simple nor fast to compute; significant research effort has been…

We introduce the DeTerministic Virtual Machine (DTVM) Stack, a next-generation smart contract execution framework designed to address critical performance, determinism, and ecosystem compatibility challenges in blockchain networks. Building…

Optimal deployment of deep neural networks (DNNs) on state-of-the-art Systems-on-Chips (SoCs) is crucial for tiny machine learning (TinyML) at the edge. The complexity of these SoCs makes deployment non-trivial, as they typically contain…

In the past few years, more and more AI applications have been applied to edge devices. However, models trained by data scientists with machine learning frameworks, such as PyTorch or TensorFlow, can not be seamlessly executed on edge. In…

We implemented and optimized matrix multiplications between dense and block-sparse matrices on CUDA. We leveraged TVM, a deep learning compiler, to explore the schedule space of the operation and generate efficient CUDA code. With the…

While the prominent quantum computing architectures are based on superconducting technology, new quantum hardware technologies are emerging, such as Trapped Ions, Neutral Atoms (or FPQAs), Silicon Spin Qubits, etc. This diverse set of…

This paper reports our efforts on swCaffe, a highly efficient parallel framework for accelerating deep neural networks (DNNs) training on Sunway TaihuLight, the current fastest supercomputer in the world that adopts a unique many-core…

As deep learning models nowadays are widely adopted by both cloud services and edge devices, reducing the latency of deep learning model inferences becomes crucial to provide efficient model serving. However, it is challenging to develop…

Dynamism is common in AI computation, e.g., the dynamic tensor shapes and the dynamic control flows in models. Due to the long compilation time, existing runtime compilation damages the model efficiency, while the offline compilers either…

Scaling inference-time computation has enabled Large Language Models (LLMs) to achieve strong reasoning performance, but inherently sequential decoding leads to substantial latency, especially on complex tasks. Recent work on adaptive…

While proprietary systems such as Seedance-2.0 have achieved remarkable success in omni-capable video generation, open-source alternatives significantly lag behind. Most academic models remain heavily fragmented, and the few existing…

Deep learning architectures for supervised learning on tabular data range from simple multilayer perceptrons (MLP) to sophisticated Transformers and retrieval-augmented methods. This study highlights a major, yet so far overlooked…

Surface wave tomography is essential for investigating the shear-wave velocity structure of the crust and upper mantle. The direct surface wave tomography method, DSurfTomo, has become one of the most widely adopted packages due to its…

The ShenWei many-core series processors powering multiple cutting-edge supercomputers are equipped with their unique on-chip heterogeneous architecture. They have long required programmers to write separate codes for the control part on…

Deep learning-based code generation has completely transformed the way developers write programs today. Existing approaches to code generation have focused either on the Sequence-to-Sequence paradigm, which generates target code as a…

We present a compilation flow for the generation of CNN inference accelerators on FPGAs. The flow translates a frozen model into OpenCL kernels with the TVM compiler and uses the Intel OpenCL SDK to compile to an FPGA bitstream. We improve…

The rise of data-intensive applications exposed the limitations of conventional processor-centric von-Neumann architectures that struggle to meet the off-chip memory bandwidth demand. Therefore, recent innovations in computer architecture…