TickerTain
TickerTain
NewsroomShortsPortfolioConvergence
NewsroomShortsPortfolioConvergence
←
▶ 26:51 · $ETCHED-AI · Etched.ai CEO touts rack-scale inference chips with cluster memory
episode briefing
Bloomberg Tech

Chip Stocks on Track for Best Quarter Ever | Bloomberg Tech 6/30/2026

2026-06-30 · 4 company · 5 thematic
sentiment
3 bull0 bear1 neu
speakers
gavin uberti

Gavin Uberti co-founded Etched.ai, a startup building rack-scale inference chips with cluster-scale memory architecture, backed by Jane Street and Venture Tech Alliance.

rob thummel

Rob Thummel manages Tortoise Capital's AI Infrastructure Active ETF, focusing on energy and infrastructure investments benefiting from AI data center buildout.

quote
“WE CAN GET MORE OUT OF THE SAME VOLUME OF HBM. WE WANT TO BE ABLE TO USE THE SAME HBM AND THE SAME SRAM TO RUN MORE TOKENS FOR USERS. A GPU THROTTLES. YOU CANNOT FIT MORE CHIPS UNDER THAT. WE LOWER THE VOLTAGE A LOT, RUN IT UNDER THE HALF…”
—gavin uberti
now playing · $ETCHED-AI
$MU···bullish· high· posrob thummel
Tortoise Capital bullish on Micron: 85% gross margin beats Nvidia
AI data center demand has created a structural memory undersupply, transforming Micron from a cyclical player into a secular grower with 85% gross margins — 10 points above Nvidia…
$SKHY···bullish· mediumrob thummel
Tortoise Capital eyes SK Hynix ADR for AI infrastructure fund
SK Hynix's 57% HBM market share makes it a core AI infrastructure holding; Tortoise welcomes the planned US ADR listing as it would simplify access for their AI Infrastructure Act…
$SPCX···neutral· mediumunknown
Bloomberg Intelligence models 800% SpaceX revenue growth by 2030
SpaceX could grow revenue 9x to $160B by 2030 driven by xAI data center agreements (Colossus 1/2), but its rich valuation requires successfully pivoting the Starlink business from…
$ETCHED-AIbullish· high· posgavin uberti
Etched.ai CEO touts rack-scale inference chips with cluster memory
Etched's rack-scale inference system uses cluster-scale memory sharing and low-voltage operation to deliver superior tokens-per-dollar and power efficiency versus GPUs, with a tea…