[BidClub_]

PERSON DIRECTORY

Molly O'Shea

Host of Sourcery. Molly O'Shea appears in 33 indexed conversations across Sourcery. This directory brings every appearance, source, TL;DR, digest, and transcript into one searchable feed.

33 EPISODES1 SHOW
4 episodes2 active
Language
SourceryEN · 62 min

We’re Reaching the Physical Limits of Chips

Molly O'SheaAnnie Lamont

AI’s investable bottlenecks are shifting toward power, memory, cooling, and data transfer as Moore’s Law nears a physical plateau.Micron, SK hynix, and Samsung control 95% of memory, whose price rose 700% this year as hyperscalers absorbed supply.Photonics could relieve several constraints, with co-packaged optics expected beside GPUs within five years as memory pressure pushes specialized hardware beyond ever-larger models.

SourceryEN · 33 min

The $10T AI Buildout Has a Photonics Problem

Molly O'SheaHerwig Van HoveYannick De Koninck

AI’s next bottleneck is the interconnect fabric: models no longer fit on one GPU, while agentic calls make latency critical, putting photonics alongside compute as core infrastructure.Optical bandwidth addresses copper and power constraints, but lasers, tools, substrates, throughput, and yield remain scarce as NVIDIA’s demand shock runs ahead of supply; Themaa targets 2027 production and 2028 full ramp.

SourceryEN · 69 min

BlackRock's Tony Kim on AI's Next Winners?

Tony KimMolly O'Shea

AI has shifted value toward chips and hardware, with Kim estimating $30T+ there versus roughly $10T across software/services/internet.Base compute has jumped 10,000×, while the rebuild may require a trillion dollars this year and ten trillion over five years.Token factories may compress SaaS margins; memory and foundation-lab IPOs are watchpoints as Kim says his thesis may change.

SourceryEN · 69 min

BlackRock's Tony Kim on AI's Next Winners? Chips, Memory, Robotics & Quantum

Tony KimMolly O'Shea

AI has reset the market toward compute hardware, with Kim estimating more than $30T in chips and hardware versus roughly $10T in software, services, and internet.The resulting $1T of CapEx this year and $10T over five years is moving data centimeters and millimeters, but three-to-four-year fab cycles face memory shortages, making 2030 milestones such as quantum and orbital data centers key watchpoints.