SlideShare una empresa de Scribd logo
1 de 31
Descargar para leer sin conexión
No Moore Left to Give
Enterprise Computing After Moore’s Law
CTO
bryan@joyent.com
Bryan Cantrill
@bcantrill
Moore’s Law?
• Which of these is Moore’s Law?
1. “Transistor speed increases exponentially over time”
2. “Transistors per dollar grows exponentially over time”
3. “Transistor density grows exponentially over time”
4. “The number of transistors in a package grows
exponentially over time”
Moore’s Law?
• Which of these is Moore’s Law?
1. “Transistor speed increases exponentially over time”
2. “Transistors per dollar grows exponentially over time”
3. “Transistor density grows exponentially over time”
4. “The number of transistors in a package grows
exponentially over time”
• Answer: All of them — and none of them!
Moore’s Law: Primary document, ca. 1965
Gordon Moore, ca. 1965
Gordon Moore, ca. 1965 (wat)
Aside: “Chiefly North American”
Gordon Moore, ca. 1965
Gordon Moore, ca. 1965
Gordon Moore, ca. 1965
Moore’s Law?
Gordon Moore, ca. 1965
Moore’s Law!
Gordon Moore, ca. 1965
Moore’s Law?!
So… Moore’s Law?
• In the 1965 paper, there is no Moore’s Law per se — just a
bunch of incredibly astute and prescient observations
• The term “Moore’s Law” would be coined by Carver Mead in
1971 as part of his work on determining ultimate physical limits
• Moore updated the law in 1975 to be a doubling of transistor
density every two years
• Dennard scaling would be outlined in detail in 1974
• But for many years, Moore’s Law could be inferred to be
doublings of transistor density, speed, and economics…
Moore’s Law: Good old days?
• The 1980s and 1990s were seemingly great for Moore’s Law…
• But even in those halcyon years, memory was not speeding up
relative to the CPU — we were hitting the memory wall
• Symmetric multiprocessing became the clear path for delivering
single-system throughput on multi-threaded workloads…
• …but SMP didn’t help single-threaded performance, which
needed aggressive speculative execution to achieve low CPI
• SMP and speculative execution were both fraught with peril:
even the good old days of Moore’s Law were still a challenge!
Moore’s Law: Beginning of the end
• Dennard scaling ended in ~2006 due to current leakage…
• …but by then chip multiprocessing was clearly the trajectory
• CMP was enhanced by simultaneous multithreading (SMT),
which offered another factor of two on throughput…
• Thanks to experience with SMP, CMP/SMT was less of a
software performance apocalypse than many had feared — but
more of a security apocalypse than anyone anticipated!
• …but “dark silicon” threatened to limit CMP
Moore’s Law: Crossing the rubicon
• As feature sizes continued to shrink to 28nm and then 22nm, a
critical rubicon was crossed:
Source: “Why Migration to 20nm bulk CMOS and 16/14nm FinFETs is not best approach for semiconductor industry” (Handel Jones, IBS, 2014)
Moore’s Law: Crossing the rubicon
• Moving to 20nm and beyond required solving some nasty and
very expensive problems, necessitating FinFETs
• We were adhering to one definition of Rashomon’s Moore’s Law
(density) but not the more important definition (economics)
• Moore’s Law was not — or should not have been — merely
about density at any cost!
Moore’s Law: End of the end?
• In August 2018, GlobalFoundries suddenly stopped 7nm
development, citing economics and lack of demand
• GlobalFoundries departure left TSMC and Samsung on 7nm —
and Intel on 14nm, struggling to get to 10nm
• Intel’s 10nm Cannon Lake is now three years late — and for Ice
Lake/Cascade Lake, Intel will intermix 14nm and 10nm
• Moving to 3nm/5nm will require moving beyond FinFETs to
GAAFETs — and to EUV photolithography
Moore’s Law: Really, the end
• Even if narrowly interpreted to be an exponential increase in
transistor density over time at any cost, Moore’s Law is ending
• A silicon atom is 0.2nm wide — 3nm is very close to the end on
that basis alone!
• Whether or not the industry can get to 3nm/5nm or not isn’t the
question — the question is at what cost
• Economically, Moore’s Law is indisputably over
• What does this mean for the enterprise?
Beyond Moore’s Law: Quantum computing?
• Quantum computing is an interesting and laudable idea — and
becoming surprisingly real! (viz. IBM’s Q System One)
• But:
• The problem domain is very limited
• The economics are entirely unknown
• The scale is still tiny — and there doesn’t yet appear to be a
Moore’s Law for qubits
• Quantum computing may be relevant one day for the enterprise,
but no time soon
Beyond Moore’s Law: Specialized compute?
• Specialized computing is already important in several
commercially relevant problem domains, e.g. GPGPU for ML/DL
• This will continue, though hardware design/verification cycles
will demand the problem domains to be well-specified
• When implemented as ASICs and (especially) FPGAs, use of
older CMOS nodes may inhibit some performance gain
• Transistor limits will apply to specialized compute as well —
specialized compute will quickly hit the accelerator wall
Beyond Moore’s Law: 3D?
• 3D (vs. planar) has proven to work well for some components
(e.g., 3D NAND on an older CMOS node)…
• …but using 3D for CPUs may pose heat, yield and potentially
cost problems
• Intel is actively investing in 3D via Foveros
Beyond Moore’s Law: 2.5D Chiplets?
• One very promising approach is to make smaller chiplets that
are then integrated into a larger die with a through-silicon via:









Source: “Cost-Effective Design of Scalable High-Performance Systems Using Active and Passive Interposers” (Stow et al.)
• Chiplets allow different functions (e.g. CPU, GPU) and different
processes (e.g. 7nm, 14nm) on the same die
• AMD is using chiplets for EPYC; Intel is investing in EMIB
Beyond Moore’s Law: Alternative physics?
• There are tons of alternative physics out there: silicon
photonics, carbon nanotubes, phase change memory, etc.
• All of these are worthy of exploration and productization!
• History indicates that promise is insufficient: they must be able
to compete with the economics of the technologies that they will
displace
• When they break through — e.g., flash memory — they can
change the entire industry
• So watch them and cheer for them — but don’t assume them
Beyond Moore’s Law: Wright’s Law?
• In 1936, Theodore Wright studied the costs of aircraft
manufacturing, finding that the cost dropped with experience
• When volume doubled, unit costs dropped by 10-15%
• This has been dubbed the learning curve effect
• In 2013, Jessika Trancik et al. found Wright’s observation held
predictive power for semiconductors
• Could this hold true for the 7nm node or for older nodes like
14nm? Could transistors start getting cheaper again?
Beyond Moore’s Law: Compute everywhere?
• Especially if and as highest-density compute becomes cheaper,
we will see many more CPUs — and in many more places
• We are already seeing CPUs on the NIC (SmartNIC), CPUs
next to flash (e.g. open-channel SSD) and on the spindle (e.g.
Western Digital’s SweRV, based on RISC-V)
• These promise to deliver higher-performing abstractions — but
with concomitant increases in overall system complexity!
• Left unchecked, there are potentially terrifying security
ramifications; comprehensive systems design is called for!
Beyond Moore’s Law: Durable computing?
• Assuming that we are indeed on 7nm for a protracted period of
time, should we still be replacing host CPUs every 3 years?
• How long can a CPU last, anyway?
• Should we be exploring more durable computers?
• Should we be looking to other axes of improvement with respect
to density or power?
• e.g., Open Compute Project — and specifically the Intel Twin
Lakes (Xeon D-2100) for OCP Yosemite V2
Beyond Moore’s Law
• The end of Moore’s Law is a change — but the truth is that we
have always had engineering challenges at every level of the
hardware and software stack
• The future will be no different: the complexion of the problems
may change — and we may focus more on efficiency and
density as raw power — but it remains alive with promise
• For software engineers, there will be new opportunities to
understand and optimize for the hardware beneath us!
• Beyond Moore’s Law, we must increasingly think of the system:
hardware and software designed together!

Más contenido relacionado

Similar a No Moore Left to Give: Enterprise Computing After Moore's Law

Quantum Computers PART 1 & 2 by Prof Lili Saghafi
Quantum Computers  PART 1 & 2 by Prof Lili SaghafiQuantum Computers  PART 1 & 2 by Prof Lili Saghafi
Quantum Computers PART 1 & 2 by Prof Lili SaghafiProfessor Lili Saghafi
 
Quantum Computing vs Encryption: A Battle to Watch Out for
Quantum Computing vs Encryption: A Battle to Watch Out forQuantum Computing vs Encryption: A Battle to Watch Out for
Quantum Computing vs Encryption: A Battle to Watch Out forCheapSSLsecurity
 
Is 201 presentation_2_moore's_law[1]
Is 201 presentation_2_moore's_law[1]Is 201 presentation_2_moore's_law[1]
Is 201 presentation_2_moore's_law[1]skroh
 
resonant tunneling devices
 resonant tunneling devices resonant tunneling devices
resonant tunneling devicesnagisetty2
 
Quantum Computers New Generation of Computers part 6 by Prof Lili Saghafi
Quantum Computers New Generation of Computers part 6 by Prof Lili SaghafiQuantum Computers New Generation of Computers part 6 by Prof Lili Saghafi
Quantum Computers New Generation of Computers part 6 by Prof Lili SaghafiProfessor Lili Saghafi
 
Rebooting Computing- The next phase of Moores law(Quantum Computing,Biologica...
Rebooting Computing- The next phase of Moores law(Quantum Computing,Biologica...Rebooting Computing- The next phase of Moores law(Quantum Computing,Biologica...
Rebooting Computing- The next phase of Moores law(Quantum Computing,Biologica...santhosh sreeram
 
Things to come - possible futures for electronics
Things to come - possible futures for electronicsThings to come - possible futures for electronics
Things to come - possible futures for electronicschrised
 
Introduction to Parallel Computing
Introduction to Parallel ComputingIntroduction to Parallel Computing
Introduction to Parallel ComputingRoshan Karunarathna
 
Technology trends Moore’s law
Technology trends Moore’s lawTechnology trends Moore’s law
Technology trends Moore’s lawSyed Zaid Irshad
 
The Parallel Computing Revolution Is Only Half Over
The Parallel Computing Revolution Is Only Half OverThe Parallel Computing Revolution Is Only Half Over
The Parallel Computing Revolution Is Only Half Overinside-BigData.com
 
Quantum Computing by Elisha.pptx
Quantum  Computing by Elisha.pptxQuantum  Computing by Elisha.pptx
Quantum Computing by Elisha.pptxAsheMathew
 
Performance beyond moore's law
Performance beyond moore's lawPerformance beyond moore's law
Performance beyond moore's lawAnand Haridass
 
What is telecom?
What is telecom?What is telecom?
What is telecom?TSSC
 
VLSI Design-Lecture2 introduction to ic technology
VLSI Design-Lecture2 introduction to ic technologyVLSI Design-Lecture2 introduction to ic technology
VLSI Design-Lecture2 introduction to ic technologysritulasiadigopula
 
MT48 A Flash into the future of storage….  Flash meets Persistent Memory: The...
MT48 A Flash into the future of storage….  Flash meets Persistent Memory: The...MT48 A Flash into the future of storage….  Flash meets Persistent Memory: The...
MT48 A Flash into the future of storage….  Flash meets Persistent Memory: The...Dell EMC World
 

Similar a No Moore Left to Give: Enterprise Computing After Moore's Law (20)

Quantum Computers PART 1 & 2 by Prof Lili Saghafi
Quantum Computers  PART 1 & 2 by Prof Lili SaghafiQuantum Computers  PART 1 & 2 by Prof Lili Saghafi
Quantum Computers PART 1 & 2 by Prof Lili Saghafi
 
Quantum Computing vs Encryption: A Battle to Watch Out for
Quantum Computing vs Encryption: A Battle to Watch Out forQuantum Computing vs Encryption: A Battle to Watch Out for
Quantum Computing vs Encryption: A Battle to Watch Out for
 
Is 201 presentation_2_moore's_law[1]
Is 201 presentation_2_moore's_law[1]Is 201 presentation_2_moore's_law[1]
Is 201 presentation_2_moore's_law[1]
 
resonant tunneling devices
 resonant tunneling devices resonant tunneling devices
resonant tunneling devices
 
Quantum Computers New Generation of Computers part 6 by Prof Lili Saghafi
Quantum Computers New Generation of Computers part 6 by Prof Lili SaghafiQuantum Computers New Generation of Computers part 6 by Prof Lili Saghafi
Quantum Computers New Generation of Computers part 6 by Prof Lili Saghafi
 
Rebooting Computing- The next phase of Moores law(Quantum Computing,Biologica...
Rebooting Computing- The next phase of Moores law(Quantum Computing,Biologica...Rebooting Computing- The next phase of Moores law(Quantum Computing,Biologica...
Rebooting Computing- The next phase of Moores law(Quantum Computing,Biologica...
 
Quantum computers
Quantum computersQuantum computers
Quantum computers
 
Things to come - possible futures for electronics
Things to come - possible futures for electronicsThings to come - possible futures for electronics
Things to come - possible futures for electronics
 
Introduction to Parallel Computing
Introduction to Parallel ComputingIntroduction to Parallel Computing
Introduction to Parallel Computing
 
Technology trends Moore’s law
Technology trends Moore’s lawTechnology trends Moore’s law
Technology trends Moore’s law
 
Nano computing.
Nano computing.Nano computing.
Nano computing.
 
The Parallel Computing Revolution Is Only Half Over
The Parallel Computing Revolution Is Only Half OverThe Parallel Computing Revolution Is Only Half Over
The Parallel Computing Revolution Is Only Half Over
 
Quantum Computing by Elisha.pptx
Quantum  Computing by Elisha.pptxQuantum  Computing by Elisha.pptx
Quantum Computing by Elisha.pptx
 
Future-of-computers
Future-of-computersFuture-of-computers
Future-of-computers
 
Performance beyond moore's law
Performance beyond moore's lawPerformance beyond moore's law
Performance beyond moore's law
 
What is telecom?
What is telecom?What is telecom?
What is telecom?
 
Vlsi
VlsiVlsi
Vlsi
 
VLSI Design-Lecture2 introduction to ic technology
VLSI Design-Lecture2 introduction to ic technologyVLSI Design-Lecture2 introduction to ic technology
VLSI Design-Lecture2 introduction to ic technology
 
MT48 A Flash into the future of storage….  Flash meets Persistent Memory: The...
MT48 A Flash into the future of storage….  Flash meets Persistent Memory: The...MT48 A Flash into the future of storage….  Flash meets Persistent Memory: The...
MT48 A Flash into the future of storage….  Flash meets Persistent Memory: The...
 
Technology Disruption
Technology DisruptionTechnology Disruption
Technology Disruption
 

Más de bcantrill

Predicting the Present
Predicting the PresentPredicting the Present
Predicting the Presentbcantrill
 
Sharpening the Axe: The Primacy of Toolmaking
Sharpening the Axe: The Primacy of ToolmakingSharpening the Axe: The Primacy of Toolmaking
Sharpening the Axe: The Primacy of Toolmakingbcantrill
 
Coming of Age: Developing young technologists without robbing them of their y...
Coming of Age: Developing young technologists without robbing them of their y...Coming of Age: Developing young technologists without robbing them of their y...
Coming of Age: Developing young technologists without robbing them of their y...bcantrill
 
I have come to bury the BIOS, not to open it: The need for holistic systems
I have come to bury the BIOS, not to open it: The need for holistic systemsI have come to bury the BIOS, not to open it: The need for holistic systems
I have come to bury the BIOS, not to open it: The need for holistic systemsbcantrill
 
Towards Holistic Systems
Towards Holistic SystemsTowards Holistic Systems
Towards Holistic Systemsbcantrill
 
Tockilator: Deducing Tock execution flows from Ibex Verilator traces
Tockilator: Deducing Tock execution flows from Ibex Verilator tracesTockilator: Deducing Tock execution flows from Ibex Verilator traces
Tockilator: Deducing Tock execution flows from Ibex Verilator tracesbcantrill
 
Andreessen's Corollary: Ethical Dilemmas in Software Engineering
Andreessen's Corollary: Ethical Dilemmas in Software EngineeringAndreessen's Corollary: Ethical Dilemmas in Software Engineering
Andreessen's Corollary: Ethical Dilemmas in Software Engineeringbcantrill
 
Visualizing Systems with Statemaps
Visualizing Systems with StatemapsVisualizing Systems with Statemaps
Visualizing Systems with Statemapsbcantrill
 
Platform values, Rust, and the implications for system software
Platform values, Rust, and the implications for system softwarePlatform values, Rust, and the implications for system software
Platform values, Rust, and the implications for system softwarebcantrill
 
Is it time to rewrite the operating system in Rust?
Is it time to rewrite the operating system in Rust?Is it time to rewrite the operating system in Rust?
Is it time to rewrite the operating system in Rust?bcantrill
 
dtrace.conf(16): DTrace state of the union
dtrace.conf(16): DTrace state of the uniondtrace.conf(16): DTrace state of the union
dtrace.conf(16): DTrace state of the unionbcantrill
 
The Hurricane's Butterfly: Debugging pathologically performing systems
The Hurricane's Butterfly: Debugging pathologically performing systemsThe Hurricane's Butterfly: Debugging pathologically performing systems
The Hurricane's Butterfly: Debugging pathologically performing systemsbcantrill
 
Papers We Love: ARC after dark
Papers We Love: ARC after darkPapers We Love: ARC after dark
Papers We Love: ARC after darkbcantrill
 
Principles of Technology Leadership
Principles of Technology LeadershipPrinciples of Technology Leadership
Principles of Technology Leadershipbcantrill
 
Zebras all the way down: The engineering challenges of the data path
Zebras all the way down: The engineering challenges of the data pathZebras all the way down: The engineering challenges of the data path
Zebras all the way down: The engineering challenges of the data pathbcantrill
 
Platform as reflection of values: Joyent, node.js, and beyond
Platform as reflection of values: Joyent, node.js, and beyondPlatform as reflection of values: Joyent, node.js, and beyond
Platform as reflection of values: Joyent, node.js, and beyondbcantrill
 
Debugging under fire: Keeping your head when systems have lost their mind
Debugging under fire: Keeping your head when systems have lost their mindDebugging under fire: Keeping your head when systems have lost their mind
Debugging under fire: Keeping your head when systems have lost their mindbcantrill
 
Down Memory Lane: Two Decades with the Slab Allocator
Down Memory Lane: Two Decades with the Slab AllocatorDown Memory Lane: Two Decades with the Slab Allocator
Down Memory Lane: Two Decades with the Slab Allocatorbcantrill
 
Oral tradition in software engineering: Passing the craft across generations
Oral tradition in software engineering: Passing the craft across generationsOral tradition in software engineering: Passing the craft across generations
Oral tradition in software engineering: Passing the craft across generationsbcantrill
 
The Container Revolution: Reflections after the first decade
The Container Revolution: Reflections after the first decadeThe Container Revolution: Reflections after the first decade
The Container Revolution: Reflections after the first decadebcantrill
 

Más de bcantrill (20)

Predicting the Present
Predicting the PresentPredicting the Present
Predicting the Present
 
Sharpening the Axe: The Primacy of Toolmaking
Sharpening the Axe: The Primacy of ToolmakingSharpening the Axe: The Primacy of Toolmaking
Sharpening the Axe: The Primacy of Toolmaking
 
Coming of Age: Developing young technologists without robbing them of their y...
Coming of Age: Developing young technologists without robbing them of their y...Coming of Age: Developing young technologists without robbing them of their y...
Coming of Age: Developing young technologists without robbing them of their y...
 
I have come to bury the BIOS, not to open it: The need for holistic systems
I have come to bury the BIOS, not to open it: The need for holistic systemsI have come to bury the BIOS, not to open it: The need for holistic systems
I have come to bury the BIOS, not to open it: The need for holistic systems
 
Towards Holistic Systems
Towards Holistic SystemsTowards Holistic Systems
Towards Holistic Systems
 
Tockilator: Deducing Tock execution flows from Ibex Verilator traces
Tockilator: Deducing Tock execution flows from Ibex Verilator tracesTockilator: Deducing Tock execution flows from Ibex Verilator traces
Tockilator: Deducing Tock execution flows from Ibex Verilator traces
 
Andreessen's Corollary: Ethical Dilemmas in Software Engineering
Andreessen's Corollary: Ethical Dilemmas in Software EngineeringAndreessen's Corollary: Ethical Dilemmas in Software Engineering
Andreessen's Corollary: Ethical Dilemmas in Software Engineering
 
Visualizing Systems with Statemaps
Visualizing Systems with StatemapsVisualizing Systems with Statemaps
Visualizing Systems with Statemaps
 
Platform values, Rust, and the implications for system software
Platform values, Rust, and the implications for system softwarePlatform values, Rust, and the implications for system software
Platform values, Rust, and the implications for system software
 
Is it time to rewrite the operating system in Rust?
Is it time to rewrite the operating system in Rust?Is it time to rewrite the operating system in Rust?
Is it time to rewrite the operating system in Rust?
 
dtrace.conf(16): DTrace state of the union
dtrace.conf(16): DTrace state of the uniondtrace.conf(16): DTrace state of the union
dtrace.conf(16): DTrace state of the union
 
The Hurricane's Butterfly: Debugging pathologically performing systems
The Hurricane's Butterfly: Debugging pathologically performing systemsThe Hurricane's Butterfly: Debugging pathologically performing systems
The Hurricane's Butterfly: Debugging pathologically performing systems
 
Papers We Love: ARC after dark
Papers We Love: ARC after darkPapers We Love: ARC after dark
Papers We Love: ARC after dark
 
Principles of Technology Leadership
Principles of Technology LeadershipPrinciples of Technology Leadership
Principles of Technology Leadership
 
Zebras all the way down: The engineering challenges of the data path
Zebras all the way down: The engineering challenges of the data pathZebras all the way down: The engineering challenges of the data path
Zebras all the way down: The engineering challenges of the data path
 
Platform as reflection of values: Joyent, node.js, and beyond
Platform as reflection of values: Joyent, node.js, and beyondPlatform as reflection of values: Joyent, node.js, and beyond
Platform as reflection of values: Joyent, node.js, and beyond
 
Debugging under fire: Keeping your head when systems have lost their mind
Debugging under fire: Keeping your head when systems have lost their mindDebugging under fire: Keeping your head when systems have lost their mind
Debugging under fire: Keeping your head when systems have lost their mind
 
Down Memory Lane: Two Decades with the Slab Allocator
Down Memory Lane: Two Decades with the Slab AllocatorDown Memory Lane: Two Decades with the Slab Allocator
Down Memory Lane: Two Decades with the Slab Allocator
 
Oral tradition in software engineering: Passing the craft across generations
Oral tradition in software engineering: Passing the craft across generationsOral tradition in software engineering: Passing the craft across generations
Oral tradition in software engineering: Passing the craft across generations
 
The Container Revolution: Reflections after the first decade
The Container Revolution: Reflections after the first decadeThe Container Revolution: Reflections after the first decade
The Container Revolution: Reflections after the first decade
 

Último

A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)Gabriella Davis
 
Presentation on how to chat with PDF using ChatGPT code interpreter
Presentation on how to chat with PDF using ChatGPT code interpreterPresentation on how to chat with PDF using ChatGPT code interpreter
Presentation on how to chat with PDF using ChatGPT code interpreternaman860154
 
IAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI SolutionsIAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI SolutionsEnterprise Knowledge
 
Boost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdfBoost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdfsudhanshuwaghmare1
 
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptxHampshireHUG
 
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...Drew Madelung
 
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...apidays
 
How to convert PDF to text with Nanonets
How to convert PDF to text with NanonetsHow to convert PDF to text with Nanonets
How to convert PDF to text with Nanonetsnaman860154
 
What Are The Drone Anti-jamming Systems Technology?
What Are The Drone Anti-jamming Systems Technology?What Are The Drone Anti-jamming Systems Technology?
What Are The Drone Anti-jamming Systems Technology?Antenna Manufacturer Coco
 
Raspberry Pi 5: Challenges and Solutions in Bringing up an OpenGL/Vulkan Driv...
Raspberry Pi 5: Challenges and Solutions in Bringing up an OpenGL/Vulkan Driv...Raspberry Pi 5: Challenges and Solutions in Bringing up an OpenGL/Vulkan Driv...
Raspberry Pi 5: Challenges and Solutions in Bringing up an OpenGL/Vulkan Driv...Igalia
 
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...Miguel Araújo
 
08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking Men08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking MenDelhi Call girls
 
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
08448380779 Call Girls In Diplomatic Enclave Women Seeking MenDelhi Call girls
 
08448380779 Call Girls In Friends Colony Women Seeking Men
08448380779 Call Girls In Friends Colony Women Seeking Men08448380779 Call Girls In Friends Colony Women Seeking Men
08448380779 Call Girls In Friends Colony Women Seeking MenDelhi Call girls
 
Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024The Digital Insurer
 
A Call to Action for Generative AI in 2024
A Call to Action for Generative AI in 2024A Call to Action for Generative AI in 2024
A Call to Action for Generative AI in 2024Results
 
Real Time Object Detection Using Open CV
Real Time Object Detection Using Open CVReal Time Object Detection Using Open CV
Real Time Object Detection Using Open CVKhem
 
Exploring the Future Potential of AI-Enabled Smartphone Processors
Exploring the Future Potential of AI-Enabled Smartphone ProcessorsExploring the Future Potential of AI-Enabled Smartphone Processors
Exploring the Future Potential of AI-Enabled Smartphone Processorsdebabhi2
 
Tata AIG General Insurance Company - Insurer Innovation Award 2024
Tata AIG General Insurance Company - Insurer Innovation Award 2024Tata AIG General Insurance Company - Insurer Innovation Award 2024
Tata AIG General Insurance Company - Insurer Innovation Award 2024The Digital Insurer
 
Histor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slideHistor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slidevu2urc
 

Último (20)

A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)A Domino Admins Adventures (Engage 2024)
A Domino Admins Adventures (Engage 2024)
 
Presentation on how to chat with PDF using ChatGPT code interpreter
Presentation on how to chat with PDF using ChatGPT code interpreterPresentation on how to chat with PDF using ChatGPT code interpreter
Presentation on how to chat with PDF using ChatGPT code interpreter
 
IAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI SolutionsIAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI Solutions
 
Boost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdfBoost Fertility New Invention Ups Success Rates.pdf
Boost Fertility New Invention Ups Success Rates.pdf
 
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
 
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...
Strategies for Unlocking Knowledge Management in Microsoft 365 in the Copilot...
 
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
Apidays Singapore 2024 - Building Digital Trust in a Digital Economy by Veron...
 
How to convert PDF to text with Nanonets
How to convert PDF to text with NanonetsHow to convert PDF to text with Nanonets
How to convert PDF to text with Nanonets
 
What Are The Drone Anti-jamming Systems Technology?
What Are The Drone Anti-jamming Systems Technology?What Are The Drone Anti-jamming Systems Technology?
What Are The Drone Anti-jamming Systems Technology?
 
Raspberry Pi 5: Challenges and Solutions in Bringing up an OpenGL/Vulkan Driv...
Raspberry Pi 5: Challenges and Solutions in Bringing up an OpenGL/Vulkan Driv...Raspberry Pi 5: Challenges and Solutions in Bringing up an OpenGL/Vulkan Driv...
Raspberry Pi 5: Challenges and Solutions in Bringing up an OpenGL/Vulkan Driv...
 
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
Mastering MySQL Database Architecture: Deep Dive into MySQL Shell and MySQL R...
 
08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking Men08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking Men
 
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
08448380779 Call Girls In Diplomatic Enclave Women Seeking Men
 
08448380779 Call Girls In Friends Colony Women Seeking Men
08448380779 Call Girls In Friends Colony Women Seeking Men08448380779 Call Girls In Friends Colony Women Seeking Men
08448380779 Call Girls In Friends Colony Women Seeking Men
 
Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024Finology Group – Insurtech Innovation Award 2024
Finology Group – Insurtech Innovation Award 2024
 
A Call to Action for Generative AI in 2024
A Call to Action for Generative AI in 2024A Call to Action for Generative AI in 2024
A Call to Action for Generative AI in 2024
 
Real Time Object Detection Using Open CV
Real Time Object Detection Using Open CVReal Time Object Detection Using Open CV
Real Time Object Detection Using Open CV
 
Exploring the Future Potential of AI-Enabled Smartphone Processors
Exploring the Future Potential of AI-Enabled Smartphone ProcessorsExploring the Future Potential of AI-Enabled Smartphone Processors
Exploring the Future Potential of AI-Enabled Smartphone Processors
 
Tata AIG General Insurance Company - Insurer Innovation Award 2024
Tata AIG General Insurance Company - Insurer Innovation Award 2024Tata AIG General Insurance Company - Insurer Innovation Award 2024
Tata AIG General Insurance Company - Insurer Innovation Award 2024
 
Histor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slideHistor y of HAM Radio presentation slide
Histor y of HAM Radio presentation slide
 

No Moore Left to Give: Enterprise Computing After Moore's Law

  • 1. No Moore Left to Give Enterprise Computing After Moore’s Law CTO bryan@joyent.com Bryan Cantrill @bcantrill
  • 2. Moore’s Law? • Which of these is Moore’s Law? 1. “Transistor speed increases exponentially over time” 2. “Transistors per dollar grows exponentially over time” 3. “Transistor density grows exponentially over time” 4. “The number of transistors in a package grows exponentially over time”
  • 3. Moore’s Law? • Which of these is Moore’s Law? 1. “Transistor speed increases exponentially over time” 2. “Transistors per dollar grows exponentially over time” 3. “Transistor density grows exponentially over time” 4. “The number of transistors in a package grows exponentially over time” • Answer: All of them — and none of them!
  • 4. Moore’s Law: Primary document, ca. 1965
  • 6. Gordon Moore, ca. 1965 (wat)
  • 16. So… Moore’s Law? • In the 1965 paper, there is no Moore’s Law per se — just a bunch of incredibly astute and prescient observations • The term “Moore’s Law” would be coined by Carver Mead in 1971 as part of his work on determining ultimate physical limits • Moore updated the law in 1975 to be a doubling of transistor density every two years • Dennard scaling would be outlined in detail in 1974 • But for many years, Moore’s Law could be inferred to be doublings of transistor density, speed, and economics…
  • 17. Moore’s Law: Good old days? • The 1980s and 1990s were seemingly great for Moore’s Law… • But even in those halcyon years, memory was not speeding up relative to the CPU — we were hitting the memory wall • Symmetric multiprocessing became the clear path for delivering single-system throughput on multi-threaded workloads… • …but SMP didn’t help single-threaded performance, which needed aggressive speculative execution to achieve low CPI • SMP and speculative execution were both fraught with peril: even the good old days of Moore’s Law were still a challenge!
  • 18. Moore’s Law: Beginning of the end • Dennard scaling ended in ~2006 due to current leakage… • …but by then chip multiprocessing was clearly the trajectory • CMP was enhanced by simultaneous multithreading (SMT), which offered another factor of two on throughput… • Thanks to experience with SMP, CMP/SMT was less of a software performance apocalypse than many had feared — but more of a security apocalypse than anyone anticipated! • …but “dark silicon” threatened to limit CMP
  • 19. Moore’s Law: Crossing the rubicon • As feature sizes continued to shrink to 28nm and then 22nm, a critical rubicon was crossed: Source: “Why Migration to 20nm bulk CMOS and 16/14nm FinFETs is not best approach for semiconductor industry” (Handel Jones, IBS, 2014)
  • 20. Moore’s Law: Crossing the rubicon • Moving to 20nm and beyond required solving some nasty and very expensive problems, necessitating FinFETs • We were adhering to one definition of Rashomon’s Moore’s Law (density) but not the more important definition (economics) • Moore’s Law was not — or should not have been — merely about density at any cost!
  • 21. Moore’s Law: End of the end? • In August 2018, GlobalFoundries suddenly stopped 7nm development, citing economics and lack of demand • GlobalFoundries departure left TSMC and Samsung on 7nm — and Intel on 14nm, struggling to get to 10nm • Intel’s 10nm Cannon Lake is now three years late — and for Ice Lake/Cascade Lake, Intel will intermix 14nm and 10nm • Moving to 3nm/5nm will require moving beyond FinFETs to GAAFETs — and to EUV photolithography
  • 22. Moore’s Law: Really, the end • Even if narrowly interpreted to be an exponential increase in transistor density over time at any cost, Moore’s Law is ending • A silicon atom is 0.2nm wide — 3nm is very close to the end on that basis alone! • Whether or not the industry can get to 3nm/5nm or not isn’t the question — the question is at what cost • Economically, Moore’s Law is indisputably over • What does this mean for the enterprise?
  • 23. Beyond Moore’s Law: Quantum computing? • Quantum computing is an interesting and laudable idea — and becoming surprisingly real! (viz. IBM’s Q System One) • But: • The problem domain is very limited • The economics are entirely unknown • The scale is still tiny — and there doesn’t yet appear to be a Moore’s Law for qubits • Quantum computing may be relevant one day for the enterprise, but no time soon
  • 24. Beyond Moore’s Law: Specialized compute? • Specialized computing is already important in several commercially relevant problem domains, e.g. GPGPU for ML/DL • This will continue, though hardware design/verification cycles will demand the problem domains to be well-specified • When implemented as ASICs and (especially) FPGAs, use of older CMOS nodes may inhibit some performance gain • Transistor limits will apply to specialized compute as well — specialized compute will quickly hit the accelerator wall
  • 25. Beyond Moore’s Law: 3D? • 3D (vs. planar) has proven to work well for some components (e.g., 3D NAND on an older CMOS node)… • …but using 3D for CPUs may pose heat, yield and potentially cost problems • Intel is actively investing in 3D via Foveros
  • 26. Beyond Moore’s Law: 2.5D Chiplets? • One very promising approach is to make smaller chiplets that are then integrated into a larger die with a through-silicon via:
 
 
 
 
 Source: “Cost-Effective Design of Scalable High-Performance Systems Using Active and Passive Interposers” (Stow et al.) • Chiplets allow different functions (e.g. CPU, GPU) and different processes (e.g. 7nm, 14nm) on the same die • AMD is using chiplets for EPYC; Intel is investing in EMIB
  • 27. Beyond Moore’s Law: Alternative physics? • There are tons of alternative physics out there: silicon photonics, carbon nanotubes, phase change memory, etc. • All of these are worthy of exploration and productization! • History indicates that promise is insufficient: they must be able to compete with the economics of the technologies that they will displace • When they break through — e.g., flash memory — they can change the entire industry • So watch them and cheer for them — but don’t assume them
  • 28. Beyond Moore’s Law: Wright’s Law? • In 1936, Theodore Wright studied the costs of aircraft manufacturing, finding that the cost dropped with experience • When volume doubled, unit costs dropped by 10-15% • This has been dubbed the learning curve effect • In 2013, Jessika Trancik et al. found Wright’s observation held predictive power for semiconductors • Could this hold true for the 7nm node or for older nodes like 14nm? Could transistors start getting cheaper again?
  • 29. Beyond Moore’s Law: Compute everywhere? • Especially if and as highest-density compute becomes cheaper, we will see many more CPUs — and in many more places • We are already seeing CPUs on the NIC (SmartNIC), CPUs next to flash (e.g. open-channel SSD) and on the spindle (e.g. Western Digital’s SweRV, based on RISC-V) • These promise to deliver higher-performing abstractions — but with concomitant increases in overall system complexity! • Left unchecked, there are potentially terrifying security ramifications; comprehensive systems design is called for!
  • 30. Beyond Moore’s Law: Durable computing? • Assuming that we are indeed on 7nm for a protracted period of time, should we still be replacing host CPUs every 3 years? • How long can a CPU last, anyway? • Should we be exploring more durable computers? • Should we be looking to other axes of improvement with respect to density or power? • e.g., Open Compute Project — and specifically the Intel Twin Lakes (Xeon D-2100) for OCP Yosemite V2
  • 31. Beyond Moore’s Law • The end of Moore’s Law is a change — but the truth is that we have always had engineering challenges at every level of the hardware and software stack • The future will be no different: the complexion of the problems may change — and we may focus more on efficiency and density as raw power — but it remains alive with promise • For software engineers, there will be new opportunities to understand and optimize for the hardware beneath us! • Beyond Moore’s Law, we must increasingly think of the system: hardware and software designed together!