“Machine Learning runtime library for neural network acceleration”

*Patent pending*, Apr 2019

Aaron Ng

University of Michigan

Advanced Computer Architecture Lab

M.S.E. Computer Science and Engineering

Hardware Systems, Distributed Computing

Advisor: Igor L. Markov

Advanced Computer Architecture Lab

M.S.E. Computer Science and Engineering

Hardware Systems, Distributed Computing

Advisor: Igor L. Markov

hello@aaron-ng.com

A. N. Ng, J. Zejda, E. Delaye, X. Teng, S. Santan, S. T. Soe, A. Sirasao, E. Ghasemi, S. Settle

“Machine Learning runtime library for neural network acceleration”

*Patent pending*, Apr 2019

“Machine Learning runtime library for neural network acceleration”

A. N. Ng, E. Delaye, E. Ghasemi, X. Teng, J. Zejda, Y. Wu, S. Settle, A. Sirasao

“Multi-layer neural network processing by a neural network accelerator using host-communicated merged weights and a package of per-layer instructions”

*Patent pending*, Apr 2019

“Multi-layer neural network processing by a neural network accelerator using host-communicated merged weights and a package of per-layer instructions”

A. N. Ng, E. Delaye, J. Zejda, A. Sirasao

“Host-directed multi-layer neural network processing via per-layer work requests”

*Patent pending*, Apr 2019

“Host-directed multi-layer neural network processing via per-layer work requests”

A. N. Ng, J. Zejda, E. Delaye, X. Teng, A. Sirasao

“Neural network processing system having host-controlled kernel accelerators”

*Patent pending*, Apr 2019

“Neural network processing system having host-controlled kernel accelerators”

E. Delaye, A. Sirasao, A. Ng, Y. Wu, J. Zejda

“Image preprocessing for generalized image processing”

*Patent pending*, Apr 2019

“Image preprocessing for generalized image processing”

X. Teng, A. N. Ng, A. Sirasao, E. Delaye

“Neural network processing system having multiple processors and a neural network accelerator”

*Patent pending*, Apr 2019

“Neural network processing system having multiple processors and a neural network accelerator”

S. Settle, E. Delaye, A. N. Ng, E. Ghasemi, A. Sirasao, X. Teng, J. Zejda,

“Software-driven design optimization for mapping between floating-point and fixed-point multiply accumulators”

*U.S. Patent No. 10,678,509*, Jun 2020

“Software-driven design optimization for mapping between floating-point and fixed-point multiply accumulators”

A. Sirasao, E. Delaye, A. N. Ng, E. Ghasemi,

“Inline image preprocessing for convolution operations using a matrix multiplier on an integrated circuit”

*U.S. Patent No. 10,460,416*, Oct 2019

“Inline image preprocessing for convolution operations using a matrix multiplier on an integrated circuit”

A. N. Ng, S. Krishnamurthy, G. S. Gasparyan,

“Timing Closure of Circuit Designs for Integrated Circuits”

*U.S. Patent No. 10,366,201*, Jul 2019

“Timing Closure of Circuit Designs for Integrated Circuits”

E. Delaye, A. Sirasao, A. N. Ng

“Software-Defined Memory Bandwidth Reduction by Hierarchical Stream Buffering for General Matrix Multiplication In A Programmable IC”

*U.S. Patent No. 10,354,733*, Jul 2019

“Software-Defined Memory Bandwidth Reduction by Hierarchical Stream Buffering for General Matrix Multiplication In A Programmable IC”

A. N. Ng, P. Basu, S. Das

“Neural network based Physical Synthesis for Circuit Designs”

*U.S. Patent No. 10,192,016*, Jan 2019

“Neural network based Physical Synthesis for Circuit Designs”

S. Settle, M. Bollavaram, P. D'Alberto, E. Delaye, O. Fernandez, N. Fraser, A. Ng, A. Sirasao, M. Wu

“Quantizing Convolutional Neural Networks for Low-Power High-Throughput Inference Engines”

*Conference on Neural Information Processing Systems*, May 2018

“Quantizing Convolutional Neural Networks for Low-Power High-Throughput Inference Engines”

R. Lu, Z. Wang, A. N. Ng, N. Shah, S. Das

“Fanout Optimization to Facilitate Timing Improvement in Circuit Designs”

*U.S. Patent No. 9,965,581*, May 2018

“Fanout Optimization to Facilitate Timing Improvement in Circuit Designs”

I. Ganusov, A. N. Ng, R. Plyler, S. Das and F. Revenu

“Programmable Integrated Circuit Design Flow using Timing-driven Pipeline Analysis”

*U.S. Patent No. 9,836,568*, Dec 2017

“Programmable Integrated Circuit Design Flow using Timing-driven Pipeline Analysis”

R. Lu, Z. Wang, A. N. Ng and S. Das

“Post-routing Structural Netlist Optimization for Circuit Designs”

*U.S. Patent No. 9,646,126*, May 2017

“Post-routing Structural Netlist Optimization for Circuit Designs”

I. Ganusov, H. Fraisse, A. N. Ng, R. T. Possignolo and S. Das,

“Automated Extra Pipeline Analysis of Applications Mapped to Xilinx UltraScale+ FPGAs”

*Intl. Conference on Field-Programmable Logic and Applications (FPL)*, August 2016

“Automated Extra Pipeline Analysis of Applications Mapped to Xilinx UltraScale+ FPGAs”

Q. Wang, A. N. Ng, R. Aggarwal,

“Resource Mapping of Functional Areas on an Integrated Circuit”

*U.S. Patent No. 7,840,919*, Nov 2010

“Resource Mapping of Functional Areas on an Integrated Circuit”

J. A. Roy, A. N. Ng, R. Aggarwal, V. Ramachandran, and I. L. Markov,

“Solving Modern Mixed-size Placement Instances”

*Integration, the VLSI Journal*, vol. 42 no. 2, 2009

“Solving Modern Mixed-size Placement Instances”

A. N. Ng, R. Aggarwal, V. Ramachandran and I. L. Markov,

“Solving Hard Instances of Floorplacement,”

*Intl. Symposium on Physical Design (ISPD)*, April 2006

“Solving Hard Instances of Floorplacement,”

M. Moffitt, A. N. Ng and I. L. Markov,

“Constraint-driven Floorplan Repair,”

*Design Automation Conference (DAC)*, July 2006

“Constraint-driven Floorplan Repair,”

J. A. Roy, D. A. Papa, A. N. Ng, I. L Markov,

“Satisfying Whitespace Requirements in Top-down Placement,”

*Intl. Symposium on Physical Design (ISPD)*, April 2006

“Satisfying Whitespace Requirements in Top-down Placement,”

J. A. Roy, D. A. Papa, S. N. Adya, H. H. Chan, J. F. Lu, A. N. Ng, and I. L. Markov,

“Capo: Robust and Scalable Open-Source Min-cut Floorplacer,”

*Intl. Symposium on Physical Design (ISPD)*, April 2005

“Capo: Robust and Scalable Open-Source Min-cut Floorplacer,”

A. N. Ng and I. L. Markov,

“Toward Quality Tools and Tool Flows Through High-Performance Computing,”

*Intl. Symposium on Quality Electronic Design (ISQED)*, March 2005

“Toward Quality Tools and Tool Flows Through High-Performance Computing,”