Serenity@aleabitoreddit
Ex-WSB trader turned AI/semiconductor supply-chain analyst. Buys the chokepoints NVIDIA can't do without. 700k+ followers on X.
This page tracks X (Twitter) analyst @aleabitoreddit (aka Serenity) in real time — latest posts, calls and stock picks. aleabitoreddit hunts NVIDIA AI semiconductor supply-chain chokepoints across CPO, optical interconnect and Sivers Semiconductors (SIVE); below is a live roundup of aleabitoreddit's takes with each stock's today / 90-day move.
Latest from X — mentioned tickers
Auto-updated from the latest posts. Today's move shown per ticker.
This post recaps the combined 2026 capex guidance of AMZN, META, GOOGL, and MSFT reaching $720-745B, up from $695-725B. Despite short-term market deleveraging, the author is bullish on upstream semiconductors and neoclouds in the medium-long term, citing hyperscaler compute shortages and rising cloud demand. They highlight NVDA as a past beneficiary and expect similar rerating in memory, CPUs/MLCCs, CPO, glass substrates, and 800V technology as capex flows in.
Reddit reports strong Q2 earnings with revenue and EPS beats, and Q3 guidance above expectations. Despite a 22.6% drop, author attributes to false narratives.
Hyperscalers (GOOGL, MSFT, META, AMZN) see strong demand and capital expenditure, bullish. Korean memory stocks (SK Hynix, Samsung) rallied despite Jim Cramer's bearish call on data center stocks.
The author upgrades their view on POET's optical interposer collaboration with Sivers, noting that POET changed wording from 'readiness' to 'production' with Sivers lasers by end of 2026, and expects volume orders from Jabil to POET in H1 2027, supported by Sivers' institutional funding for mass production.
The author discusses $XFAB's ER, confirming its exposure to CPO and 800V shifts, with photonics volume production expected in 2028 and SiC design wins for 800V deployment. The thesis is validated despite current revenue not being impressive, and the stock may rerate as growth opportunities materialize.
Author highlights Amazon's raised 2026 capex to $220B and persistent compute shortage, bullish for AI semi trade. AXT reports record InP revenue, plans to double capacity twice by 2027, aiming to become the largest InP producer with 50%+ gross margins. Demand outstrips supply across optical chains. Also mentions MU as beneficiary of higher memory costs.
FormFactor's earnings call shows CPO (co-packaged optics) revenue accelerating: systems revenue nearly doubled, guidance raised from $10-20M to significantly exceed $20M for 2026. Management cites rapid growth and production infrastructure for CPO chips. The CPO piece flows through TSMC (likely via COUPE) and maps to Nvidia CPO products, suggesting early adoption by equipment/test players before volume ramps.
Microsoft and Meta earnings calls indicate sustained AI capex with compute scarcity across hyperscalers, reinforcing bullish signals for AI infrastructure plays.
AXT Inc (AXTI) enters long-term agreement with Lumentum (LITE) for InP substrates, with Lumentum paying a $43.5M deposit for capacity reservation.
The author bought 1 share of IREN to show solidarity with existing holders.
Featured theses
Highest-conviction calls from recent posts.
Author highlights Amazon's raised 2026 capex to $220B and persistent compute shortage, bul
X / @aleabitoreddit ↗Author highlights Amazon's raised 2026 capex to $220B and persistent compute shortage, bullish for AI semi trade. AXT reports record InP revenue, plans to double capacity twice by 2027, aiming to become the largest InP producer with 50%+ gross margins. Demand outstrips supply across optical chains. Also mentions MU as beneficiary of higher memory costs.
Why: Record InP revenue, capacity doubling plans, goal to become largest InP producer, 50%+ gross margin target, demand outstrips supply.
Why: Raised 2026 capex to $220B, compute shortage expected through 2027, most capacity already reserved, bullish for AI semiconductor trade.
Google's Gemini AI has surged from 750M to 950M monthly users in a few months, processing
X / @aleabitoreddit ↗Google's Gemini AI has surged from 750M to 950M monthly users in a few months, processing 22B API tokens/min. This growth caused compute shortages, leading Google to cut allocations to Meta and others. The data suggests AI capex will remain high.
Why: Rapid user growth (200M additions in months) and high token volume indicate strong AI demand, justifying continued high capex and competitive advantage.
This post recaps the combined 2026 capex guidance of AMZN, META, GOOGL, and MSFT reaching
X / @aleabitoreddit ↗This post recaps the combined 2026 capex guidance of AMZN, META, GOOGL, and MSFT reaching $720-745B, up from $695-725B. Despite short-term market deleveraging, the author is bullish on upstream semiconductors and neoclouds in the medium-long term, citing hyperscaler compute shortages and rising cloud demand. They highlight NVDA as a past beneficiary and expect similar rerating in memory, CPUs/MLCCs, CPO, glass substrates, and 800V technology as capex flows in.
Why: The author uses NVDA as a prime example of a company that has benefited from massive capex flows into AI compute, implying continued strength in its bottleneck position as GPU supplier amid hyperscaler compute shortages.
Reddit reports strong Q2 earnings with revenue and EPS beats, and Q3 guidance above expect
X / @aleabitoreddit ↗Reddit reports strong Q2 earnings with revenue and EPS beats, and Q3 guidance above expectations. Despite a 22.6% drop, author attributes to false narratives.
Hyperscalers (GOOGL, MSFT, META, AMZN) see strong demand and capital expenditure, bullish.
X / @aleabitoreddit ↗Hyperscalers (GOOGL, MSFT, META, AMZN) see strong demand and capital expenditure, bullish. Korean memory stocks (SK Hynix, Samsung) rallied despite Jim Cramer's bearish call on data center stocks.
Why: Demand for compute far exceeds capacity; hyperscalers continue heavy capex.
Why: Demand for compute far exceeds capacity; hyperscalers continue heavy capex.
Why: Demand for compute far exceeds capacity; hyperscalers continue heavy capex.
Why: Demand for compute far exceeds capacity; hyperscalers continue heavy capex.
Why: Stock gained 30% despite Jim Cramer's bearish call on leveraged DC stocks, suggesting strong underlying demand.
The author upgrades their view on POET's optical interposer collaboration with Sivers, not
X / @aleabitoreddit ↗The author upgrades their view on POET's optical interposer collaboration with Sivers, noting that POET changed wording from 'readiness' to 'production' with Sivers lasers by end of 2026, and expects volume orders from Jabil to POET in H1 2027, supported by Sivers' institutional funding for mass production.
Why: POET upgraded guidance for optical interposer with Sivers lasers from 'readiness' to 'production' by end of 2026, and volume orders from Jabil are expected in H1 2027, indicating strong demand and progress.
Why: Sivers is the laser supplier for POET's interposer, and its ability to achieve production by end of 2026 is supported by a recent oversubscribed institutional funding round for mass production.
The author discusses $XFAB's ER, confirming its exposure to CPO and 800V shifts, with phot
X / @aleabitoreddit ↗The author discusses $XFAB's ER, confirming its exposure to CPO and 800V shifts, with photonics volume production expected in 2028 and SiC design wins for 800V deployment. The thesis is validated despite current revenue not being impressive, and the stock may rerate as growth opportunities materialize.
Why: Company explicitly confirmed exposure to co-packaged optics (CPO) and 800V transition, with multiple SiC design wins and photonics volume production planned for 2028. The author's thesis on growth opportunities is validated.
FormFactor's earnings call shows CPO (co-packaged optics) revenue accelerating: systems re
X / @aleabitoreddit ↗FormFactor's earnings call shows CPO (co-packaged optics) revenue accelerating: systems revenue nearly doubled, guidance raised from $10-20M to significantly exceed $20M for 2026. Management cites rapid growth and production infrastructure for CPO chips. The CPO piece flows through TSMC (likely via COUPE) and maps to Nvidia CPO products, suggesting early adoption by equipment/test players before volume ramps.
Why: CPO revenue nearly doubled QoQ, guidance raised to significantly exceed $20M for 2026, and management cites accelerating growth and production ramp for CPO chips. Test equipment demand from CPO is an early indicator of volume adoption.
Why: CPO flows through a 10% customer (likely TSMC via COUPE), implying TSMC is a key supplier of CPO chips for Nvidia and others. Indirect beneficiary of CPO adoption.
Why: The CPO chips flowing through TSMC are likely mapped to Nvidia's CPO products. Rapid CPO growth suggests early adoption in Nvidia's infrastructure, but no direct revenue mention.