NVIDIA Tech Talks @ UIUC

Event Information

Share this event

Date and Time

Location

Location

ECEB 1013

306 N Wright St

Urbana, IL 61801

View Map

Event description

Description

NVIDIA

TECH TALKS


Come join us for on Tuesday, January 29 at UIUC's ECEB 1013. Explore the future of AI computing in engaging, informative tech talks with experts who are creating new realities with AI. One lucky winner will even walk away with an NVIDIA product!


When: January 29 (Tuesday)

Time: 8:15 p.m. – 9:30 p.m.

Location: ECEB 1013

Who it’s for: All UIUC CS, CE, and EE students (undergraduate and graduate)


AGENDA

8:15 p.m. – 8:20 p.m. | Arrive, grab a seat

8:20 p.m. – 8:45 p.m. | Po-Han Huang (Software Engineer, TensorRT)

8:45 p.m. – 9:10 p.m. | Timur Rvachov (R&D Engineer, Autonomous Vehicles)

9:10 p.m. – 9:20 p.m. | Q&A


SPEAKER INFORMATION

Po-Han Huang
UIUC Alum & current Software Engineer in TensorRT team

Abstract: With deep learning, we can train machines to learn, perceive, reason, and solve problems, covering domains from language translation to speech recognition to real time object detection. When deploying these trained neural networks, often in real time environments, the central problem is to maximize throughput at low latency. NVIDIA’s TensorRT is a library for optimized inference on GPUs in datacenter, embedded, and automotive systems, providing up to 100x performance of CPU inference.

In this talk, we will explore how we use GPUs to accelerate matrix multiplication, an essential component of deep neural networks, and how TensorRT further improves the performance with optimizations on the network graph. We will also give some insight about what it’s like to work on the multidisciplinary TensorRT software engineering team.

Bio: Po-Han Huang is a software engineer on the TensorRT team at NVIDIA. Before NVIDIA, he received the M.S. degree in Electrical and Computer Engineering from University of Illinois at Urbana-Champaign, where he focused on deep learning and computer vision. His current position focuses on the performance of the TensorRT engine and on influencing future GPU architecture designs for inference applications.

Timur Rvachov
R&D Engineer, Autonomous Vehicles

Abstract:
There is no shortage of hype and money behind the tech industry's push for creating a self-driving car. Despite much media attention, automation of a safety-critical consumer product still poses many unsolved problems. I will overview our team's work--which heavily employs machine learning--and discuss its benefits and challenges in creating an autonomous product.

Bio: Timur Rvachov is a development engineer at the New Jersey research office. Before joining the autonomous vehicle project, he received a doctorate degree in physics from MIT. His work with the New Jersey team focuses on a learned approach to vehicle autonomy.


About NVIDIA:

NVIDIA has continuously reinvented itself over two decades. Our invention of the GPU in 1999 sparked the growth of the PC gaming market, redefined modern computer graphics, and revolutionized parallel computing. More recently, GPU computing ignited the era of AI. NVIDIA is a “learning machine” that constantly evolves by adapting to new opportunities that are hard to solve, that only we can tackle, and that matter to the world. Our life’s work is to amplify human imagination and intelligence. Make the choice to join us today.

Date and Time

Location

ECEB 1013

306 N Wright St

Urbana, IL 61801

View Map

Save This Event

Event Saved