{"id":1321,"date":"2021-12-08T05:02:00","date_gmt":"2021-12-08T05:02:00","guid":{"rendered":"https:\/\/salarydistribution.com\/machine-learning\/2021\/12\/08\/silicon-express-lanes-ai-gpus-pave-fast-routes-for-chip-designers\/"},"modified":"2021-12-08T05:02:00","modified_gmt":"2021-12-08T05:02:00","slug":"silicon-express-lanes-ai-gpus-pave-fast-routes-for-chip-designers","status":"publish","type":"post","link":"https:\/\/salarydistribution.com\/machine-learning\/2021\/12\/08\/silicon-express-lanes-ai-gpus-pave-fast-routes-for-chip-designers\/","title":{"rendered":"Silicon Express Lanes: AI, GPUs Pave Fast Routes for Chip Designers"},"content":{"rendered":"
\n

AI can design chips no human could, said Bill Dally in a virtual keynote today at the Design Automation Conference (DAC), one of the world\u2019s largest gatherings of semiconductor engineers.<\/p>\n

The chief scientist of NVIDIA discussed research in accelerated computing<\/a> and machine learning that\u2019s making chips smaller, faster and better.<\/p>\n

\u201cOur work shows you can achieve orders-of-magnitude improvements in chip design using GPU-accelerated systems. And when you add in AI, you can get superhuman results \u2014 better circuits than anyone could design by hand,\u201d said Dally, who leads a team of more than 200 people at NVIDIA Research<\/a>.<\/p>\n

Gains Span Circuits, Boards<\/b><\/h2>\n

Dally cited improvements GPUs and AI deliver across the workflow of chip and board design. His examples spanned the layout and placement of circuits to faster ways to render images of printed-circuit boards.<\/p>\n

In one particularly stunning example of speedups with GPUs, he pointed to research NVIDIA plans to present at a conference next year. The GATSPI tool accelerates detailed simulations of a chip\u2019s logic by more than 1,000x compared to commercial tools running on CPUs today.<\/p>\n

<\/p>\n

\"GATSPI<\/p>\n

<\/a>

GATSPI, a GPU-accelerated simulation tool, completes work in seconds that currently takes a full day running on a CPU.<\/figcaption><\/figure>\n

A paper at DAC this year describes how NVIDIA collaborated with Cadence Design Systems, a leading provider of EDA software, to render board-level designs using graphics techniques on NVIDIA GPUs. Their work boosted performance up to 20x for interactive operations on Cadence\u2019s Allegro X platform<\/a>, announced in June.<\/p>\n

\u201cEngineers used to wait for programs to respond after every edit or pan across an image \u2014 it was an awkward, frustrating way to work. But with GPUs, the flow becomes truly interactive,\u201d said Dally, who chaired Stanford University\u2019s computer science department before joining NVIDIA in 2009.<\/p>\n

Reinforcement Learning Delivers Rewards<\/b><\/h2>\n

A technique called NVCell, described in a DAC session this week, uses reinforcement learning to automate the job of laying out a standard cell, a basic building block of a chip.<\/p>\n

The approach reduces work that typically takes months for a 10-person team to an automated process that runs in a couple days. \u201cThat lets the engineering team focus on a few challenging cells that need to be designed by hand,\u201d said Dally.<\/p>\n

In another example of the power of reinforcement learning, NVIDIA researchers will describe at DAC a new tool called PrefixRL. It discovers how to design a circuit such as an adder, encoder or custom design.<\/p>\n

PrefixRL treats the design process like a game where the high score is in finding the smallest area and power consumption for the circuit.<\/p>\n

By letting AI optimize the process, engineers get a device that\u2019s more efficient than what\u2019s possible with today\u2019s tools. It\u2019s a good example of how AI can deliver designs no human could.<\/p>\n

Leveraging AI\u2019s Tool Box<\/b><\/h2>\n

NVIDIA worked with the University of Texas at Austin on a research project called DREAMPlace that made novel use of PyTorch, a popular software framework for deep learning. It adapted the framework used to optimize weights in a neural network to find the best spot to place a block with 10 million cells inside a larger chip.<\/p>\n

It\u2019s a routine job that currently takes nearly four hours using today\u2019s state-of-the-art techniques on CPUs. Running on NVIDIA Volta architecture GPUs in a data center or cloud service, it can finish in as little as five minutes, a 43x speedup.<\/p>\n

Getting a Clearer Image Faster<\/b><\/h2>\n

To make a chip, engineers use a lithography machine to project their design onto a semiconductor wafer. To make sure the chip performs as expected, they must accurately simulate that image, a critical challenge.<\/p>\n

NVIDIA researchers created a neural network that understands the optical process. It simulated the image on the wafer 80x faster and with higher accuracy, using a 20x smaller model than current state-of-the-art machine learning methods.<\/p>\n

It\u2019s one more example of how the combination of accelerated computing and AI are helping engineers design better chips faster.<\/p>\n

An AI-Powered Future<\/b><\/h2>\n

\u201cNVIDIA used some of these techniques to make our existing GPUs, and we plan to use more of them in the future,\u201d Dally said.<\/p>\n

\u201cI expect tomorrow\u2019s standard EDA tools will be AI-powered to make the chip designer\u2019s job easier and their results better than ever,\u201d he said.<\/p>\n

To watch Dally\u2019s keynote, register<\/a> for a complimentary pass to DAC using the code ILOVEDAC, then view the talk here<\/a>.<\/p>\n<\/div>\n","protected":false},"excerpt":{"rendered":"

https:\/\/blogs.nvidia.com\/blog\/2021\/12\/07\/ai-acceleration-dally-dac\/<\/p>\n","protected":false},"author":0,"featured_media":1322,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":[],"categories":[3],"tags":[],"_links":{"self":[{"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/posts\/1321"}],"collection":[{"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/types\/post"}],"replies":[{"embeddable":true,"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/comments?post=1321"}],"version-history":[{"count":0,"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/posts\/1321\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/media\/1322"}],"wp:attachment":[{"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/media?parent=1321"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/categories?post=1321"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/tags?post=1321"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}