Author of the publication

An Energy-Efficient Transformer Processor Exploiting Dynamic Weak Relevances in Global Attention.

, , , , , , , , , , and . IEEE J. Solid State Circuits, 58 (1): 227-242 (2023)

Please choose a person to relate this publication to

To differ between persons with the same name, the academic degree and the title of an important publication will be displayed. You can also use the button next to the name to display some publications already assigned to the person.

 

Other publications of authors with the same name

A 28nm 27.5TOPS/W Approximate-Computing-Based Transformer Processor with Asymptotic Sparsity Speculating and Out-of-Order Computing., , , , , , , , , and 1 other author(s). ISSCC, page 1-3. IEEE, (2022)Trainer: An Energy-Efficient Edge-Device Training Processor Supporting Dynamic Weight Pruning., , , , , , , , and . IEEE J. Solid State Circuits, 57 (10): 3164-3178 (2022)SWPU: A 126.04 TFLOPS/W Edge-Device Sparse DNN Training Processor With Dynamic Sub-Structured Weight Pruning., , , , and . IEEE Trans. Circuits Syst. I Regul. Pap., 69 (10): 4014-4027 (2022)A 28nm 276.55TFLOPS/W Sparse Deep-Neural-Network Training Processor with Implicit Redundancy Speculation and Batch Normalization Reformulation., , , , , , , , and . VLSI Circuits, page 1-2. IEEE, (2021)BR-CIM: An Efficient Binary Representation Computation-In-Memory Design., , , , , and . IEEE Trans. Circuits Syst. I Regul. Pap., 69 (10): 3940-3953 (2022)HPPU: An Energy-Efficient Sparse DNN Training Processor with Hybrid Weight Pruning., , , , and . AICAS, page 1-4. IEEE, (2021)A 28nm 49.7TOPS/W Sparse Transformer Processor with Random-Projection-Based Speculation, Multi-Stationary Dataflow, and Redundant Partial Product Elimination., , , , , , , , , and 3 other author(s). A-SSCC, page 1-3. IEEE, (2023)A 28nm 77.35TOPS/W Similar Vectors Traceable Transformer Processor with Principal-Component-Prior Speculating and Dynamic Bit-wise Stationary Computing., , , , , , , , , and 1 other author(s). VLSI Technology and Circuits, page 1-2. IEEE, (2023)34.1 A 28nm 83.23TFLOPS/W POSIT-Based Compute-in-Memory Macro for High-Accuracy AI Applications., , , , , , , , , and . ISSCC, page 566-568. IEEE, (2024)20.2 A 28nm 74.34TFLOPS/W BF16 Heterogenous CIM-Based Accelerator Exploiting Denoising-Similarity for Diffusion Models., , , , , , , , , and 2 other author(s). ISSCC, page 362-364. IEEE, (2024)