Tag: SpaceX

"Basically every point so far, we've underestimated required amount of compute to execute on this inflection." -@danielnewmanUV

GPUs, ASICs, memory, CPUs, networking, power. Every layer has run ahead of the model.

$TSM locked in a decade-long supply agreement with $AMKR to process and package chips at Amkor's Arizona facility.

Both companies are expanding US capacity and the partnership creates a fully domestic path from silicon wafer to finished semiconductor.

$AMAT rolled out two chipmaking systems for 3D structure production as the industry pushes density gains.

Applied Materials positioned the tools to solve manufacturing constraints tied to stacked transistor designs.

Starship is what gives $SPCX a credible path into the ~$26.5T AI infrastructure TAM.

If AI compute moves to orbit then SpaceX is the only vertically integrated player across the full stack, launch, manufacturing, power and now compute.

$AMD has acquired MEXT, whose flash-to-DRAM optimization software targets the memory constraints facing large-scale AI workloads.

The deal gives AMD a direct play in memory efficiency as data center operators face rising costs and physical space limits.

@danielnewmanUV on the networking names he's tracking at the moment.

$NVDA, $AVGO, $MRVL, and one dark horse name he and @StockSavvyShay have butted heads on in recent months.

$GOOGL is committing $1.5B through 2027 to scale its Jackson County, Alabama data center campus for AI infrastructure.

Google says the expansion supports growing demand for computing resources as enterprise AI adoption accelerates.

"Doesn't matter how much money you can throw into space, space is hard." -@StockSavvyShay

$SPCX has roughly 90% of launch share and a moat Blue Origin still can't crack, on a business @danielnewmanUV notes is still loss-making at 8% growth.

Select Date

WHY IS SPACEX SPENDING $60B ON CURSOR?

Podcast

sentiment

powered by

DAILY RIP

powered by

INSIGHTS

powered by

Rate your experience