Behind The Momentum Performance, 16-bit FLOP/s +150% / Year Enabled by 1.6x annual growth in chips per cluster and 1.6x annual growth in performance per chip Performance of Leading AI Supercomputers MM Revenue ($B) Revenue, $B $0 $2 $4 2022 2023 2024Time to 365B Annual Searches = ChatGPT 5.5x Faster vs. Google Note: Dashed-line bars are for years where Google did not disclose annual search Models AI Development Trending = Unprecedented38 AI Developer Growth (NVIDIA Ecosystem as Proxy) = +6x to 6MM Developers Over Seven Years Number of Developers, MM 0 3 6 2005 2007 2009 2011 2013 2015

. . . . . . . . . . . . . . . . . . . . . . 455 36.21 Don't write a trivial anonymous function x->f(x) for a named function f . . . . . . . 455 36.22 Avoid using floats for numeric literals in generic . . . . . . . . . . . . . . . . . . . . 1068 56 C Interface 1069 57 LLVM Interface 1083CONTENTS x 58 C Standard Library 1084 59 StackTraces 1088 60 SIMD Support 1090 III Standard Library 1091 61 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1662 103.6 For 64 bit Julia, install x86_64 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1662 103.7 For 32 bit Julia, install

. . . . . . . . . . . . . . . . . . . . . . 455 36.21 Don't write a trivial anonymous function x->f(x) for a named function f . . . . . . . 455 36.22 Avoid using floats for numeric literals in generic . . . . . . . . . . . . . . . . . . . . 1068 56 C Interface 1069 57 LLVM Interface 1083CONTENTS x 58 C Standard Library 1084 59 StackTraces 1088 60 SIMD Support 1090 III Standard Library 1091 61 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1662 103.6 For 64 bit Julia, install x86_64 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1662 103.7 For 32 bit Julia, install

. . . . . . . . . . . . . . . . . . . . . . 489 37.22 Don't write a trivial anonymous function x->f(x) for a named function f . . . . . . . 489 37.23 Avoid using floats for numeric literals in generic . . . . . . . . . . . . . . . . . . . . . . . 1136 51 Constants 1137 52 Filesystem 1142CONTENTS x 53 I/O and Network 1166 53.1 General I/O . . . . . . . . . . . . . . . . . . . . . . . . . . . . variable x julia> x = 10 10 # Doing math with x's value julia> x + 1 11 # Reassign x's value julia> x = 1 + 1 2 # You can assign values of other types, like strings of text julia> x = "Hello

. . . . . . . . . . . . . . . . . . . . . . 489 37.22 Don't write a trivial anonymous function x->f(x) for a named function f . . . . . . . 489 37.23 Avoid using floats for numeric literals in generic . . . . . . . . . . . . . . . . . . . . . . . 1136 51 Constants 1137 52 Filesystem 1142CONTENTS x 53 I/O and Network 1166 53.1 General I/O . . . . . . . . . . . . . . . . . . . . . . . . . . . . variable x julia> x = 10 10 # Doing math with x's value julia> x + 1 11 # Reassign x's value julia> x = 1 + 1 2 # You can assign values of other types, like strings of text julia> x = "Hello

. . . . . . . . . . . . . . . . . . . . . . 489 37.22 Don't write a trivial anonymous function x->f(x) for a named function f . . . . . . . 489 37.23 Avoid using floats for numeric literals in generic . . . . . . . . . . . . . . . . . . . . . . . 1136 51 Constants 1137 52 Filesystem 1142CONTENTS x 53 I/O and Network 1166 53.1 General I/O . . . . . . . . . . . . . . . . . . . . . . . . . . . . variable x julia> x = 10 10 # Doing math with x's value julia> x + 1 11 # Reassign x's value julia> x = 1 + 1 2 # You can assign values of other types, like strings of text julia> x = "Hello

. . . . . . . . . . . . . . . . . . . . . . 500 37.23 Don't write a trivial anonymous function x->f(x) for a named function f . . . . . . . 500 37.24 Avoid using floats for numeric literals in generic DataType fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1300CONTENTS x 58.3 Subtypes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1301 variable x julia> x = 10 10 # Doing math with x's value julia> x + 1 11 # Reassign x's value julia> x = 1 + 1 2 # You can assign values of other types, like strings of text julia> x = "Hello

. . . . . . . . . . . . . . . . . . . . . . 500 37.22 Don't write a trivial anonymous function x->f(x) for a named function f . . . . . . . 500 37.23 Avoid using floats for numeric literals in generic DataType fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1297CONTENTS x 58.3 Subtypes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1298 variable x julia> x = 10 10 # Doing math with x's value julia> x + 1 11 # Reassign x's value julia> x = 1 + 1 2 # You can assign values of other types, like strings of text julia> x = "Hello

. . . . . . . . . . . . . . . . . . . . . . 501 37.22 Don't write a trivial anonymous function x->f(x) for a named function f . . . . . . . 501 37.23 Avoid using floats for numeric literals in generic DataType fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1298CONTENTS x 58.3 Subtypes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1299 variable x julia> x = 10 10 # Doing math with x's value julia> x + 1 11 # Reassign x's value julia> x = 1 + 1 2 # You can assign values of other types, like strings of text julia> x = "Hello

. . . . . . . . . . . . . . . . . . . . . . 500 37.22 Don't write a trivial anonymous function x->f(x) for a named function f . . . . . . . 500 37.23 Avoid using floats for numeric literals in generic DataType fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1297CONTENTS x 58.3 Subtypes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1298 variable x julia> x = 10 10 # Doing math with x's value julia> x + 1 11 # Reassign x's value julia> x = 1 + 1 2 # You can assign values of other types, like strings of text julia> x = "Hello