~/talk

Ramine Roane, AMD | theCUBE + NYSE Wired: Media Week - Cyber & AI Innovators Summit

Talkat Cyber & AI Innovators Summit 2024 · 2024-12-16..12-18· [machine-learning]
links:none on file
slides:none on file

AMD claims a pivot into AI with MI300X GPUs optimized for large models, reporting about $5 billion in sales this year and citing Frontier/MI250X as prior exascale proof points. Architecture uses chiplets and 3D stacking to pack 2.4X HBM memory advantage, hosting Llama 3-405B in a single 8‑GPU server (vs two for 80GB cards), supports HIP/CUDA compatibility, uses Infinity Fabric and Ethernet/Ultra Ethernet, and shows near‑linear scaling across hundreds to thousands of GPUs for training while targeting high‑volume inference.

AI-written summary · about

  1. 0:00Introduction and welcome
  2. 0:23AMD’s AI leadership position
  3. 1:29Supercomputers and MI300X GPU
  4. 3:59Market growth and TAM
  5. 4:43Chiplet architecture advantage
  6. 6:29Memory density and SRAM limitations
  7. 8:15NVIDIA’s Achilles heel power
  8. 9:19CUDA moat and compatibility
  9. 11:07Systems design and ZT Systems
  10. 12:38Networking and Infinity Fabric
  11. 14:30Training vs inference markets
  12. 17:02Scaling laws and AI economics
  13. 19:33Data sources and AGI potential
  14. 22:03AGI timeline and robotics
  15. 23:50Closing remarks
~/search
type to search 6,518 conferences, 143,918 speakers and 192,133 talks
~/help
cmdctrl k
search every conference, speaker and series
/
focus the filter search
h
home
c
conferences
s
speakers
e
series
t
light / dark — the status bar button does the same
← →
on a talk: 5 seconds back / forward
0–9
on a talk: jump to 0%–90% of the recording
?
this help
esc
close this, then clear every filter — or just the search, while you are typing in it

Ignored while you are typing. Pages without a filter rail have no search to focus.