Skip navigation
With NVIDIA extreme co-design, we’ve accelerated AI performance a millionfold over 10 years, and…
0:00 / 0:23

With NVIDIA extreme co-design, we’ve accelerated AI performance a millionfold over 10 years, and…

NVIDIA

2.7M followers

Visit on X
1,905 likesDownloadOriginal post

107,866 views

With NVIDIA extreme co-design, we’ve accelerated AI performance a millionfold over 10 years, and we’re not stopping there.

29 Comments

Comments from the original post on X (Twitter)

  • Jack Kopkin's profile picture

    @kiopmbd

    Millionfold but at what power draw per query

    3 likes

  • Mia's profile picture

    @NorvegiaMia

    The positive energy you carry in your speech and actions is so beautiful 🌺

    2 likes

  • ley's profile picture

    @leyqs97

    AI performance has come a long way. Do you think the next 10 years will bring an even bigger leap?😃

    2 likes

  • G's profile picture

    @tsuyopon24

    Wishing you continued evolution and expansion in your endeavors🔜success

    1 like

  • Piano Sheet Music's profile picture

    @pianosheet2024

    Nvidia likes and respects AMD! That’s why they share videos about AMD!😀

    1 like

  • Simo | ai-costguard's profile picture

    @SlimAssiliX

    A millionfold gain only means something once you define the benchmark, since raw compute, performance per watt, and cost per useful AI task can each tell a completely different story

    1 like

  • 社区's profile picture

    @laojiucai1234

    wow!great!keep moving!

    1 like

  • Bernard Oakwood's profile picture

    @duchenne

    A millionfold in 10 years is about 4x per year, every year. Moore's law gave 2x every two years. The cost of intelligence now falls faster than any input in history. By 2035 the scarce input is power, not chips.

  • NusoMaus's profile picture

    @Nusodinkelmi

    A millionfold faster in 10 years, and my build still takes exactly one coffee.

  • Mohammad Vazin's profile picture

    @mohammadvazin

    A millionfold in ten years, and most of it came from the stack, not the transistor: lower precision, sparsity, interconnect, kernels. That's also why it's hard to compete on chips alone. The open question is how much of that gain reaches the price per token builders pay.

  • Mayank Gulati's profile picture

    @mayankgulati99

    A millionfold in ten years and I'm still hitting rate limits 😭

  • Mehmet Akcin's profile picture

    @mhmdatx

    The part that does not get said enough: co-design now reaches the building. A millionfold in performance came with racks going from 10 kW to 250 kW and 800 VDC busbars in the open spec. The chip roadmap is public and on schedule. The facility roadmap is where most of the industry is behind.

  • The SI Maximalist's profile picture

    @thesimaximalist

    NVIDIA's hardware-software co-design is key to that millionfold gain!

  • Sachin Kumar's profile picture

    @sachintwtss

    a millionfold gain in ai performance over 10 years is hard to wrap your head around. co-design across the whole stack is clearly doing the heavy lifting.

  • 失重的雨's profile picture

    @szdyai666

    It gets harder and harder to improve as you go on

  • Macro Bombastic's profile picture

    @MacroBombastic

    yeah man that curve is wild, I remember when GPUs were just for gaming lol

  • Aapakari's profile picture

    @aapakari

    Does that millionfold count the software stack, or just the silicon?

  • Ky's profile picture

    @gimgiyeon13

    👏

  • 양승민 Andy Yang's profile picture

    @AndyYang0310

    Cloud Bigdata Neural Network Algorithm Deep Learning Divergence Convergence Nvidia 3D AI in collabo w/ Samsung HBM

  • AiTraceRoot's profile picture

    @AiTraceRoot

    Make AI even greater.