Situational Awareness LP Q1 2026 just filed.
The fund that built $5.5B going long AI infrastructure just deployed $6.6B in semiconductor puts.
β $2.05B SMH put β $1.57B NVDA put β $1.0B AVGO put β $989M AMD put β $1.07B ORACLE put
AUM: $254M at inception. $13.7B today. 54Γ in 6 quarters. Full inception-to-present breakdown β https://t.co/PWa9jKd7CI
Watching Jensen in Taiwan. $NVDA body language reads confident. Dancing on stage isn't pre-negotiation. Whatever capacity matters for Rubin, the read is he already got what he came for. The real question: who didn't? Looking at you $AMD $AVGO $GOOG.
Watching Jensen in Taiwan. $NVDA body language reads confident. Dancing on stage isn't pre-negotiation. Whatever capacity matters for Rubin, the read is he already got what he came for. The real question: who didn't? Looking at you $AMD $AVGO $GOOG.
Watching Jensen in Taiwan. $NVDA body language reads confident. Dancing on stage isn't pre-negotiation. Whatever capacity matters for Rubin, the read is he already got what he came for. The real question: who didn't? Looking at you $AMD $AVGO $GOOG.
@dee_bosa The 95% on frontier models is the tell. Enterprises priced AI as structural cost when it's unoptimized routing. Model-orchestration gateways, matching task complexity to model tier become the next infrastructure category. Hidden in plain sight.
@StockSavvyShay $NOW is the cleanest live example of cross-sectional dispersion in software. Down big over 12 months while AI-infra names ripped. Workflow SaaS got priced as if AI replaces it. AI-infra got priced as if AI feeds it. Both readings can't be right.
@Beth_Kindig The $7B to $125B move on AI CPUs is the line to focus on. That's an 18x in five years, and it's driven by the attach ratio CPUs per accelerator shifting from 1:4 to 1:2 or 1:1 as agentic workloads need orchestration per accelerator.
@GavinSBaker@EquityBrian@InvestLikeBest Power, cooling, and optical aren't one bucket. Power and cooling scale with total buildout linear, substitutable. Optical scales with interconnect density, growing faster than GPU count. The most defensible part of the dispersion, not the weakest.
@jiahanjimliu This fits the thesis we mapped in Situational Awareness's Q1 book. The CRWV tranches were the first leasable-compute bet. NBIS at 5.6% is the same trade scaled a multi-billion-dollar call on neocloud capacity outside hyperscalers.
https://t.co/7l2ubTVJYS
Situational Awareness LP Q1 2026 just filed.
The fund that built $5.5B going long AI infrastructure just deployed $6.6B in semiconductor puts.
β $2.05B SMH put β $1.57B NVDA put β $1.0B AVGO put β $989M AMD put β $1.07B ORACLE put
AUM: $254M at inception. $13.7B today. 54Γ in 6 quarters. Full inception-to-present breakdown β https://t.co/PWa9jKd7CI
@HyperAICapital This fits the thesis we mapped in Situational Awareness's Q1 book. The $CRWV tranches were the first leasable-compute bet. $NBIS at 5.6% is the same trade scaled a multi-billion-dollar call on neocloud capacity outside hyperscalers.
https://t.co/7l2ubTVJYS
Situational Awareness LP Q1 2026 just filed.
The fund that built $5.5B going long AI infrastructure just deployed $6.6B in semiconductor puts.
β $2.05B SMH put β $1.57B NVDA put β $1.0B AVGO put β $989M AMD put β $1.07B ORACLE put
AUM: $254M at inception. $13.7B today. 54Γ in 6 quarters. Full inception-to-present breakdown β https://t.co/PWa9jKd7CI
$NVDA Jensen is in Taiwan meeting TSMC's chairman. He flies there almost every month.
12 elite AI/tech hedge funds agree with him $TSM just became their most widely held stock, overtaking $AMZN.
The full Q1 2026 institutional positioning:
https://t.co/gqFN0CahP9
@Nominalszn@NotA_Bull π―agree, short-term uncertainty until they can build a frontier model, but the massive amount of data they have to work with is an advantage.
$NVDA Jensen is in Taiwan meeting TSMC's chairman. He flies there almost every month.
12 elite AI/tech hedge funds agree with him $TSM just became their most widely held stock, overtaking $AMZN.
The full Q1 2026 institutional positioning:
https://t.co/gqFN0CahP9
@Beth_Kindig The institutional money is aligned. $TSM just became the most widely held stock across 12 elite AI/tech funds we track overtaking $AMZN.
https://t.co/sKpl7Lk6GZ
$NVDA Jensen is in Taiwan meeting TSMC's chairman. He flies there almost every month.
12 elite AI/tech hedge funds agree with him $TSM just became their most widely held stock, overtaking $AMZN.
The full Q1 2026 institutional positioning:
https://t.co/gqFN0CahP9
Meanwhile, the smart money already made their call:
$TSM is now the most widely held stock across the 12 elite AI/tech funds we track overtaking $AMZN. 10 of 12 funds own it.
$NVDA remains #1 conviction (score: 91.4). Altimeter has 28.6% in it. $MSFT biggest sell (net -5).
$NVDA Jensen is in Taiwan meeting TSMC's chairman. He flies there almost every month.
12 elite AI/tech hedge funds agree with him $TSM just became their most widely held stock, overtaking $AMZN.
The full Q1 2026 institutional positioning:
https://t.co/gqFN0CahP9
@Beth_Kindig The real read-through for CRWV: rising rents on older GPUs means the market is repricing useful life upward, the direct rebuttal to the depreciation bear case. Longer economic life justifies extending the D&A schedule, which lifts margins and shrinks terminal-value risk.
@Sam_Badawi Look at the composition, not just the AUM. ALAB, CIEN, COHR, LITE, that's a concentrated bet on the interconnect and optical layer, not the accelerators. Baker isn't betting on AI broadly. He's betting the bottleneck moves to moving data between chips, not the chips themselves.