| Название | : | David Patterson: Computer Architecture and Data Storage | Lex Fridman Podcast #104 |
| Продолжительность | : | 1.49.51 |
| Дата публикации | : | |
| Просмотров | : | 129 rb |
 |
I really enjoyed this conversation with David Here's the outline:br0:00 - Introductionbr3:28 - How have computers changed?br4:22 - What's inside a computer?br10:02 - Layers of abstractionbr13:05 - RISC vs CISC computer architecturesbr28:18 - Designing a good instruction set is an artbr31:46 - Measures of performancebr36:02 - RISC instruction setbr39:39 - RISC-V open standard instruction set architecturebr51:12 - Why do ARM implementations vary?br52:57 - Simple is beautiful in instruction set designbr58:09 - How machine learning changed computersbr1:08:18 - Machine learning benchmarksbr1:16:30 - Quantum computingbr1:19:41 - Moore's lawbr1:28:22 - RAID data storagebr1:36:53 - Teachingbr1:40:59 - Wrestlingbr1:45:26 - Meaning of life Comment from : Lex Fridman |
 |
Not only does this man have a great mind, he has a great heart as well Great interview Comment from : gra |
 |
Looks like Bryan Cranston from breaking bad😂 Comment from : Vedanta Madhav Shivam |
 |
BASED Comment from : Natalia Murillo |
 |
Thanks Alex for this amazing podcasts, I have a request if you could put books suggested by your guest in the description as wellbrThanks a lot! Comment from : SOFTWARE ENGINEERING |
 |
i am buy his book but didn't know that it is david peterson Comment from : budiardjo |
 |
Awesome content! Comment from : Rafael de Lucena Valle |
 |
I thought it was Acorn Risc Machine? Comment from : CMan356 |
 |
He looks like Walter from Breaking Bad equally genius! Comment from : Zero Political Correctness Lets Talk Straight! |
 |
great on the basics, please do more tech videos like this Comment from : Home Design Austin |
 |
53:25 "Forgive me for writing a long letter, I didn't have time to write a short one" - Oscar Wilde Comment from : Richard Norris |
 |
great pleasure to watch this wonderful interview! Comment from : __flow__ kudo |
 |
My first real job was for HP and I remember the buzz when the new machines with RISC architecture was introduced Yes I'm that old 😉 Great discussion guys Comment from : Mickael Jansson |
 |
I am constantly amazed how these brilliant people are always humble, and willing to explain and capable of explaining things really well in such simple terms Comment from : Burke Kelleher |
 |
In terms of architectures for AI, I think there might be place within the ISA for graph computations and traversals Comment from : Tharal Pius |
 |
Since 2020, Lex has grown exponentially in the podcasting space with interviews of more popular figures Although this interview is less popular and interests a more niche crowd, I think it is one of, if not the greatest, podcasts on the channel With David's expertise and knowledge, I think this podcast will be a much more prolific history snippet for future generations of computer engineers who read Patterson & Hennessy great textbooks Comment from : Bit Banger |
 |
On teaching - at my alum (IITK), one of the professors would talk about a PhD student (Shirish) taking an exam in the faculty room: "he is done answering the questions, now he is asking questions"! Comment from : Puneet Singh |
 |
Euv lithography, fusion production of electric energy to the grid, and quantum computing are all amazing pieces of technology One of these is not like the others 😁😁 Comment from : Cyrus Tabery |
 |
So THAT'S what MEMVAR does Comment from : Shoalstone |
 |
I would imagine hardware people were pro cisc for job preservation Comment from : Ean Erickson |
 |
Thank you so much! Reading his book on MIPS Comment from : AmCan Tech |
 |
He is a genius in his field but still so humble Comment from : Jurrasic Grant |
 |
If you know nothing of micro processor architecture, this podcast episode is mostly meaningless Comment from : Stapleman 007 |
 |
He looks a lot like Heisenberg in brba Comment from : The First Principles Guy |
 |
Legend Comment from : Till Zemann |
 |
He looks like a combination of Bryan Cranston and Patrick Stewart Comment from : RAJAT CHOPRA 🇮🇳 |
 |
Thanks! Comment from : Anthony Alexander |
 |
Damn! You have had more legebds on your podcast than perhaps any other podcast This is a gift Comment from : Batman Fan |
 |
2 hours of my life well spent Comment from : Tyler Sehon |
 |
He's really number one Thanks for the interview Lex! Comment from : Filaxsan |
 |
the one thing missing in Lex interviews is raising discussion on whether the trends of these technologies are actually good for the human species or negative This is more so in relation to technological outcomes of AI than most - indeed, perhaps that questioning is actually more pertinent in respect to AI than any other technological endeavor of Mankind Comment from : TheSulross |
 |
Hi Lex, it’s my humble request to you to keep working on such podcasts on technology I am learning and gathering ideas through your podcasts which I never had in my life Comment from : _—- |
|
|
Hi Lex, your videos are bringing a new perspective in my life towards technology Please continue the hard and good work Comment from : _—- |
 |
Thank you for this podcast with Mr white Comment from : Kareem Hallak |
 |
Amazing episode Lex Comment from : Daniel Fiori |
 |
The closed cell prudently film because energy lately consider past a protective kidney steep, obscene surfboard Comment from : esteban collazo |
 |
If Walter White was an engineer Comment from : Tejas M |
 |
great talk, thank you! Comment from : Roman Dzhadan |
 |
Walter white never died he just changed occupation from crystals to chip design Comment from : Chitranjan Baghi |
 |
Walter white?! What are you doin' here? Comment from : Mimo |
 |
He looks like the guy from Breaking Bad lol Very informative Comment from : Taliesins Personal Vlog - Taliesin McKnight |
 |
Ngl the thumbnail got me thinking this was professor X Comment from : King Kong |
 |
the following is a conversation with captain picardbrbrat least the bookmark looked like that xD Comment from : Fox Fews |
 |
The first microprocessor, the 4004, from 1971 is right at 50 years old now Comment from : perseverance8 |
 |
This podcast is absolute gold Comment from : Prabhat Kumar |
 |
The assorted coil tinctorially scribble because jewel conspicuously enter between a descriptive beach zany, befitting step-uncle Comment from : Chic Rondon |
 |
Heisenberg Comment from : DASHDV |
 |
"there has to be something special about these 40 or 50 simple instructions that you can build something intelligent from"brI'm not a computer scientist, but aren't you able to produce all logical instructions from any other set of logical instructions? In the same way you can build nand gates from and, and so on Comment from : Joe Gregory |
 |
Thank you , Lex Great conversation Comment from : Rakesh ram |
 |
Amazing to get glimpses at these individualsbrLex, you are amazing for bringing this content to the world! Comment from : Charlie Palmer |
 |
I didn't know Walter White did computer science Comment from : CLIn7 l33tW00d |
 |
In swedish, Fridman translates to "The man of peace" Comment from : Variator |
 |
What a great interview! Very interesting Thank you Lex, thank you David Comment from : TheBronx |
 |
Habana Labs AI Chipbr youtube/PfyGCbwOB_o Comment from : Han Han |
|
|
27:00brbrMarwellous ConfessionbrHow Intel won PC era?brAnd RISC bring mobile era Comment from : Han Han |
 |
It is amazing I can listen to David in my bed Comment from : Eung Jin Lee |
 |
Given that recommending books and audio books is a common occurance on your content, I would approach an audiobook company as a sponsor Comment from : Nothin Happened |
 |
Another great one from the Legendary Lex Comment from : khronos |
 |
This was an amazing interview! It's great to hear from such a knoledgeable guest to get some more insight on this interesting topics! Comment from : Nuno Lopes |
 |
Walter White is a way better teacher than the show let on Comment from : BradMyrick |
 |
One of my favorites, what a great guy to listen to in more ways than one Comment from : Hop Music |
 |
If we ignore Intel and look specifically at Nvidia and AMD (Ryzen and Radeon) moore's law has made a little bit of a comeback, but probably not sustainable using silicon It's gotten so expensive and complicated to do die shrinks that wafer cost has actually begun to rise Anyway, TSMC is killing it, they are two full nodes ahead of Intel Strange times Comment from : HypnoticSuggestion |
 |
great man! Comment from : viruslab1 |
 |
I clicked because I thought he had Patrick Stewart on Comment from : Aaron Sayd |
 |
So Walter White decided to switch from Chemistry to Computer Science Comment from : Mederbek Buzurmankulov |
 |
Charles Xavier?) Comment from : ned msen |
 |
1:26:50 Thank you! I am a computer technician in the field and the rubber meeting the road on this is devices being released with inefficient software (developers assuming the hardware is getting better) One of the worst offenders is microsoft, releasing some PCs without the hardware features to support their seemingly endless software suite! Comment from : Olympic Garden Crafts |
 |
Wonderful interview, I wonder if you could make segment pointing out books you recommend that gives the philosophy of programming Pattersons book seems interesting in that respect Comment from : Bako S |
 |
I sort of feel like the class clown saying this and I absolutely mean no disrespect to David Patterson but how many of you watching keep seeing Bryan Cranston when you see and listen to David Pattersonbr enwikipediaorg/wiki/Bryan_Cranston Comment from : Jeffrey Hudson |
 |
Aren't the "accelerators" being added to CPUs keep up with Moore's Law basically a throwback to CISC? Comment from : see tortle |
 |
love it, thanks! Comment from : watashiwahatchi |
 |
Great podcast Advertising is starting to be a bit too much, though Comment from : riiad |
 |
Pure RISC has been dead for a long time, as instructions like "fused multiply-add" show Pure CISC is also dead Abominations like SPARC were RISC Try to program early SPARC, but stop before you go madbrI wish RISC-V more successbrbrOne amazing feature of x86 is its backwards compatibility and the non-existent fragmentation ARM is more fragmented, not providing binary compatibilitybrIf RISC-V will also provide this diverse ecosystem of a gazillion incompatible instruction set extensions, it will also be dead in the long run Comment from : Bernadette Treual |
 |
As good as he is with the computers he isn't terribly good at explaining them Comment from : Sauvik Dhammakirti |
 |
Ah this is the dude from the books Comment from : sir meme-a-lot |
 |
Respect! Comment from : Ben Dover |
 |
1:02:29 Sounded like caustic from Apex Legends Comment from : gopal v |
 |
Great inteview David Patterson has worked on and been involved with a lot of amazing stuff He's a good mix of a visionary and a good engineering craftsman I guess that's some keys to working in computer architecture Comment from : Gull Lars |
 |
The "simple" instruction Risc is a little confusing It's not mean that that ISA has a fewer instructions (i think on the AVR micros here) it's just meaning that the instruction is doing a simple taskI like the load store explanation better The x86 from the Pentium till now is between RISC and CISC blurred line Comment from : Ferenc Szabo |
 |
Who came here for Picard? Comment from : DrDress |
 |
Badass, both of em' Comment from : Dangelo Shakur |
 |
It would be nice if you have a conversation with Andrea Morello Comment from : Vincent |
 |
you lost me at hello Comment from : August Reigns |
 |
Why is Captain Picard on here? Comment from : HoMaster Zerg |
 |
Lex is truly one of the great people of our time Thank you for exposing so many fascinating and impactful people to the world at large Comment from : SpenserFL |
 |
Excellent interview Easy on your top five Comment from : Mårten Justrell |
 |
David doesn't seem knowledgeable about the semiconductor industry: spectrumieeeorg/nanoclast/semiconductors/processors/intel-now-packs-100-million-transistors-in-each-square-millimeter Comment from : George Martin 59 |
 |
I see Moore's law as a self fulfilling prophecy What I cannot quite decide is whether it accelerated or slowed down the progress, though I lean more towards the latter It seems to me that the large semiconductor companies have for the past thirty or forty years set their targets to just about match Moore's law as if not reaching it would be an embarrassing failure while just reaching it would be enough I cannot help thinking that by aiming to satisfy Moore's law rather than aiming as high as they can progress was considerably slowed down On the other hand it could be that without that threat of embarrassing failure might have given that extra motivation needed Comment from : Lister Dave |
 |
Am I the only one getting a Walter White flashback? Seriously though what an intelligent and loving person Good interview Comment from : Nomadic Brian |
 |
CISC, RISC Hit by DISC? Comment from : pat bonny |
 |
OK this one did it; I finally subscribed Comment from : Dame Frances Yates Fan Club |
 |
Great episode Comment from : clayton surgeon |
 |
Simplicity is genius Comment from : John Vardas |
 |
Thanks Lex! This is by far the best interview from your podcast You and Mr Paterson had a very intelligent and entertaining conversation It takes two to dance Tango Comment from : Sergio Vianna do Rio |
 |
Thanks Lex for this video It is worth a lot of money :) Comment from : Gnichi Mohamed |
 |
I don't believe that Patterson's statement related to slowing improvement impacting cloud computing hold true:br1) Cloud computing will drive efficiency, anyhow, due to the transactional model that is coming with serverless computingbr2) The concept of cloud is elasticity, which just means you scale to more systems; you're rarely ever consuming entire processorsbrbrEDIT: Which the concept of RAID, which followed in conversation cloud is like that You can have redundancy and pool the performance of multiple resources Comment from : Dustin |
 |
Beautiful explanations! I've enjoyed listening to this very much Thank you Both of you Comment from : MadCatAttack1 |
 |
RISC vs CISC Computer Architectures (David Patterson) | AI Podcast Clips with Lex Fridman РѕС‚ : Lex Clips Download Full Episodes | The Most Watched videos of all time |
 |
How Machine Learning Changed Computer Architecture Design (David Patterson) | AI Clips with Lex РѕС‚ : Lex Clips Download Full Episodes | The Most Watched videos of all time |
 |
Michael Saylor: Bitcoin, Inflation, and the Future of Money | Lex Fridman Podcast #276 РѕС‚ : Lex Fridman Download Full Episodes | The Most Watched videos of all time |
 |
Jordan Peterson: Life, Death, Power, Fame, and Meaning | Lex Fridman Podcast #313 РѕС‚ : Lex Fridman Download Full Episodes | The Most Watched videos of all time |
 |
Nic Carter: Bitcoin Core Values, Layered Scaling, and Blocksize Debates | Lex Fridman Podcast #173 РѕС‚ : Lex Fridman Download Full Episodes | The Most Watched videos of all time |
 |
Vitalik Buterin: Ethereum, Cryptocurrency, and the Future of Money | Lex Fridman Podcast #80 РѕС‚ : Lex Fridman Download Full Episodes | The Most Watched videos of all time |
 |
Andrew Huberman: Sleep, Dreams, Creativity, Fasting, and Neuroplasticity | Lex Fridman Podcast #164 РѕС‚ : Lex Fridman Download Full Episodes | The Most Watched videos of all time |
 |
Elon Musk: SpaceX, Mars, Tesla Autopilot, Self-Driving, Robotics, and AI | Lex Fridman Podcast #252 РѕС‚ : Lex Fridman Download Full Episodes | The Most Watched videos of all time |
 |
Sergey Nazarov: Chainlink, Smart Contracts, and Oracle Networks | Lex Fridman Podcast #181 РѕС‚ : Lex Fridman Download Full Episodes | The Most Watched videos of all time |
 |
Jim Keller: The Future of Computing, AI, Life, and Consciousness | Lex Fridman Podcast #162 РѕС‚ : Lex Fridman Download Full Episodes | The Most Watched videos of all time |
