David Patterson is a Turing award winner and professor of computer science at Berkeley. He is known for pioneering contributions to RISC processor architecture used by 99% of new chips today and for co-creating RAID storage. The impact that these two lines of research and development have had on our world is immeasurable. He is also one of the great educators of computer science in the world. His book with John Hennessy "Computer Architecture: A Quantitative Approach" is how I first learned about and was humbled by the inner workings of machines at the lowest level. Support this podcast by supporting these sponsors: - Jordan Harbinger Show: https://jordanharbinger.com/lex/ - Cash App – use code "LexPodcast" and download: - Cash App (App Store): https://apple.co/2sPrUHe - Cash App (Google Play): https://bit.ly/2MlvP5w This conversation is part of the Artificial Intelligence podcast. If you would like to get more information about this podcast go to https://lexfridman.com/ai or connect with @lexfridman on Twitter, LinkedIn, Facebook, Medium, or YouTube where you can watch the video versions of these conversations. If you enjoy the podcast, please rate it 5 stars on Apple Podcasts, follow on Spotify, or support it on Patreon. Here's the outline of the episode. On some podcast players you should be able to click the timestamp to jump to that time. OUTLINE: 00:00 - Introduction 03:28 - How have computers changed? 04:22 - What's inside a computer? 10:02 - Layers of abstraction 13:05 - RISC vs CISC computer architectures 28:18 - Designing a good instruction set is an art 31:46 - Measures of performance 36:02 - RISC instruction set 39:39 - RISC-V open standard instruction set architecture 51:12 - Why do ARM implementations vary? 52:57 - Simple is beautiful in instruction set design 58:09 - How machine learning changed computers 1:08:18 - Machine learning benchmarks 1:16:30 - Quantum computing 1:19:41 - Moore's law 1:28:22 - RAID data storage 1:36:53 - Teaching 1:40:59 - Wrestling 1:45:26 - Meaning of life
This podcast features a guest who is known for his pioneering contributions to risk processor architecture used by 99% of new chips today and for co-creating RAID storage.
This episode gives a brief overview of microprocessors, including their components, functions, and history.
The use of open source hardware can simplify microprocessors as they change rapidly due to Moore's Law. This can be achieved by using standard software to generate simple instructions that can be translated by a compiler, and open-source can be used at the design architecture instruction set level or at the manufacturer level.
This podcast discusses the benefits of focusing on simple instruction sets versus relying on compilers to reduce complicated instructions. Despite executing more instructions, simple instruction sets can lead to faster processing times.
Building a successful instruction set requires a combination of being in the right place at the right time and having a design that users enjoy. The most effective way to compare the performance of different computers is to use agreed benchmark programs that allow for compatibility with a large software base.
The main difference in clock cycles between RISC and Sys instruction sets is that the latter takes 10 clock cycles per instruction while the former takes only two. Open instruction sets allow for open source implementations and open source processors for wider use.
The open and free RISC-V instruction set architecture has the potential to become a mainstream alternative to the proprietary x86 and Arm, as industries look towards unification, though the process may take several years to happen.
The RISC-V instruction set, developed by Andrew Waterman, Christo Sonovich and Yunsup Lee, features a subset of simple instructions that all software can run on. Additional, optional features can accelerate performance in the case of larger software
This podcast highlights the shift from traditional programming to machine learning, whereby creating programs that extract the desired output from large data sets as it proves to be more effective. The program emphasizes the significance of RISC-V in developing innovative technology, that entirely relies on simple instruction sets to make inferences from data.
Investing in a software stack is crucial for running machine learning benchmarks effectively. Without a proper software stack, companies will struggle to keep up with the revolution of software 2.0.
This podcast discusses the use of benchmarking in computer vision and the potential of quantum computing which may not be commercially viable for a few more years, and when it does arise, it will be specific to certain applications.
The history of disc drives points towards a trend of becoming smaller and smaller, and this matches with Gordon Moore's well-known law of doubling transistors every two years. However, experts believe that the computer industry is in need of radical change at a lower level that could affect the entire software stack due to the growth of data and technological advancements.
The people you work with, influence, and help contribute to job satisfaction, even more so than transient experiences, such as doing research. Despite preconceived notions, many individuals may enjoy teaching if they give it a chance.