Frontier-tech opportunities traced to research, across 10 topics

⚡ At a glance

Company (stock code)	How urgent	Next big event	Why it matters
COHR Coherent	⏳ Promising: start small	No big event expected soon	A light-based 'highway' between AI chip and memory
STM STMicroelectronics	⏳ Promising: start small	No big event expected soon	Silicon photonics reaches high-volume factories
NVDA Nvidia	⏳ Promising: start small	No big event expected soon	Silicon photonics reaches high-volume factories
MTARTECH.NS MTAR Technologies	👀 Too early: keep an eye on it	No big event expected soon	ISRO's GSLV-F18 puts NVS-03 in orbit
UPST Upstart	👀 Too early: keep an eye on it	No big event expected soon	Plaid launches a transformer AI credit score it says predicts…
SOFI SoFi Technologies	👀 Too early: keep an eye on it	No big event expected soon	Plaid launches a transformer AI credit score it says predicts…
BAJFINANCE.NS Bajaj Finance	👀 Too early: keep an eye on it	No big event expected soon	Plaid launches a transformer AI credit score it says predicts…
JHG Janus Henderson	👀 Too early: keep an eye on it	No big event expected soon	Janus Henderson puts a Nasdaq-100 fund on a blockchain (JQQQ)
NDAQ Nasdaq	👀 Too early: keep an eye on it	No big event expected soon	Janus Henderson puts a Nasdaq-100 fund on a blockchain (JQQQ)
CRSP CRISPR Therapeutics	👀 Too early: keep an eye on it	2026-11-27	FDA draft rule would let one approval cover many personalised…
NTLA Intellia Therapeutics	👀 Too early: keep an eye on it	2026-11-27	FDA draft rule would let one approval cover many personalised…
IQMX IQM Quantum Computers	👀 Too early: keep an eye on it	No big event expected soon	IQM entangles all 54 qubits on its cloud Emerald chip at once
🔒 Private Altera	👀 Too early: keep an eye on it	No big event expected soon	Altera launches Agilex 2X FPGAs with 1.8× better performance per watt…
MSFT Microsoft	👀 Too early: keep an eye on it	No big event expected soon	OpenAI's GPT-6.1 Terra
AMZN Amazon	👀 Too early: keep an eye on it	No big event expected soon	Anthropic's Claude Haiku 5.5 costs about 75% less to run than Haiku…
CEVA Ceva	👀 Too early: keep an eye on it	No big event expected soon	Ceva unveils next-gen NeuPro AI engines
PAYTM.NS One 97 Communications (Paytm)	👀 Too early: keep an eye on it	No big event expected soon	India delays a 0.4% UPI merchant fee to 'immediately after Diwali'
BESI.AS Besi	👀 Too early: keep an eye on it	No big event expected soon	Gluing chips together copper-to-copper (hybrid bonding)
6857.T Advantest	👀 Too early: keep an eye on it	No big event expected soon	Gluing chips together copper-to-copper (hybrid bonding)
CGPOWER.NS CG Power	👀 Too early: keep an eye on it	No big event expected soon	Gluing chips together copper-to-copper (hybrid bonding)
ADEA Adeia	👀 Too early: keep an eye on it	No big event expected soon	Gluing chips together copper-to-copper (hybrid bonding)
000660.KS SK hynix	👀 Too early: keep an eye on it	No big event expected soon	Gluing chips together copper-to-copper (hybrid bonding)
META Meta Platforms	👀 Too early: keep an eye on it	No big event expected soon	Meta's superintelligence lab ships its first frontier model
LYG Lloyds Banking Group	👀 Too early: keep an eye on it	No big event expected soon	Lloyds' agentic AI fraud assistant passes 90% decision accuracy on…
ICICIBANK.NS ICICI Bank	👀 Too early: keep an eye on it	No big event expected soon	RBI presses banks to prove cyber defences as AI-assisted API attacks…
QUICKHEAL.NS Quick Heal Technologies	👀 Too early: keep an eye on it	No big event expected soon	RBI presses banks to prove cyber defences as AI-assisted API attacks…
UFO Procure Space ETF	👀 Too early: keep an eye on it	No big event expected soon	SpaceX plans to catch both Starship stages on Flight 15
4062.T Ibiden	👀 Too early: keep an eye on it	No big event expected soon	Glass replacing plastic under AI chips (glass substrates)
Unlisted (parent: Tata group cos) Tata Electronics fab	👀 Too early: keep an eye on it	No big event expected soon	Glass replacing plastic under AI chips (glass substrates)
GLW Corning	👀 Too early: keep an eye on it	No big event expected soon	Glass replacing plastic under AI chips (glass substrates)
TSM TSMC	👀 Too early: keep an eye on it	No big event expected soon	Glass replacing plastic under AI chips (glass substrates)
GOOGL Alphabet	👀 Too early: keep an eye on it	No big event expected soon	Google shows 'magic state cultivation' on its Willow chip
QTUM Defiance Quantum ETF	👀 Too early: keep an eye on it	No big event expected soon	Google shows 'magic state cultivation' on its Willow chip
GFS GlobalFoundries	👀 Too early: keep an eye on it	No big event expected soon	PsiQuantum shows a light-based switch that rewires AI data-centre…
PRME Prime Medicine	👀 Too early: keep an eye on it	2026-11-27	First prime edit inside a human body
BRN.AX BrainChip	👀 Too early: keep an eye on it	2026-11-03	Brain-like chip that can learn on the device (neuromorphic)
AMD AMD	👀 Too early: keep an eye on it	2026-11-03	Brain-like chip that can learn on the device (neuromorphic)
RKLB Rocket Lab	👀 Too early: keep an eye on it	No big event expected soon	Rocket Lab pushes Neutron's first launch to 2027 after finding flaws…
QNT Quantinuum	👀 Too early: keep an eye on it	No big event expected soon	Quantinuum reports 99.9995% logical fidelity on Helios
MU Micron	👀 Too early: keep an eye on it	No big event expected soon	Making noisy analog AI chips accurate (weight splitting)
KAYNES.NS Kaynes Technology	👀 Too early: keep an eye on it	No big event expected soon	Making noisy analog AI chips accurate (weight splitting)
TWST Twist Bioscience	👀 Too early: keep an eye on it	No big event expected soon	A porous silica chip writes 11× more DNA per spot
TSEM Tower Semiconductor	👀 Too early: keep an eye on it	No big event expected soon	Computing with light instead of electricity (optical AI chip)
AIXA Aixtron	👀 Too early: keep an eye on it	No big event expected soon	Stacking atom-thin transistors on top of each other (2D CFETs)
LAES SEALSQ	👀 Too early: keep an eye on it	No big event expected soon	Alice & Bob: about 22,600 'cat qubits' could break RSA-2048…

Tap a card below for thesis, catalysts, risks and plan. Our judgment · Not financial advice · No return promises.

🎯 Opportunity cards

Tap a card to open thesis, catalysts, risks and plan. Readiness = average of linked research items. Our judgment · Not financial advice.

Illustration for opportunity stmicroelectronics

STM

STMicroelectronics · Europe

⏳ Promising: start small 💪 High conviction ⏳ 2026-2027

1Paper
2Prototype
3Factories & orders
4Revenue

💡 Thesis

🎯 PIC100 photonics in volume
🔗 Silicon photonics reaches high-volume factories

📅 Catalysts

✅ PIC100 photonics in volume
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Auto/industrial weakness
❌ Auto/industrial weakness
❌ Nvidia networking SVP: 'If you can use copper, use copper' — optics only where distance forces it.

✅ Plan

⏳ Promising: start small · conviction High (score 66)
Our judgment · Not financial advice · No return promises

🗓 Timeline

Step 1Capacity grows; more customers
Step 2Over 4x capacity planned by 2027 (company plan)
Step 3Next version (TSV) for co-packaged optics

Timeline steps are our judgment unless marked as company plan.

NVDA

Nvidia · USA

⏳ Promising: start small 💪 Medium conviction ⏳ Next results ~late Nov 2026 (est.)

1Paper
2Prototype
3Factories & orders
4Revenue

💡 Thesis

🎯 Vera Rubin ramp in full production
🔗 Silicon photonics reaches high-volume factories
🔗 A light switch sending 200 billion bits per second per colour
🔗 Mistral releases the weights of Large 4, a ~1-trillion-parameter open model aime

📅 Catalysts

✅ Vera Rubin ramp in full production
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Customer capex slowdown; China rules
❌ Customer capex slowdown; China rules
❌ Nvidia networking SVP: 'If you can use copper, use copper' — optics only where distance forces it.

✅ Plan

⏳ Promising: start small · conviction Medium (score 63)
Our judgment · Not financial advice · No return promises

🗓 Timeline

Step 1Capacity grows; more customers
Step 2Over 4x capacity planned by 2027 (company plan)
Step 3Next version (TSV) for co-packaged optics

Timeline steps are our judgment unless marked as company plan.

COHR

Coherent · USA

⏳ Promising: start small 💪 Medium conviction ⏳ 2026-2029

1Paper
2Prototype
3Factories & orders
4Revenue

💡 Thesis

🎯 Lasers/optics for photonic links & CPO
🔗 A light-based 'highway' between AI chip and memory
🔗 A light switch sending 200 billion bits per second per colour
🔗 Silicon photonics reaches high-volume factories

📅 Catalysts

✅ Lasers/optics for photonic links & CPO
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Optics price wars
❌ Optics price wars
❌ Nvidia networking SVP: 'If you can use copper, use copper' — optics only where distance forces it.

✅ Plan

⏳ Promising: start small · conviction Medium (score 64)
Our judgment · Not financial advice · No return promises

🗓 Timeline

Step 1Prototype shown to partners
Step 2Early customer trials
Step 3Possibly inside some AI servers

Timeline steps are our judgment unless marked as company plan.

AMZN

Amazon · USA

👀 Too early: keep an eye on it 💪 Low conviction (watchlist) ⏳ 2026-2028

1Paper / research
2Limited preview
3Generally available
4Widely adopted

💡 Thesis

🎯 Anthropic investor; Claude Haiku 5.5 on AWS
🔗 Anthropic's Claude Haiku 5.5 costs about 75% less to run than Haiku 4.5

📅 Catalysts

✅ Anthropic investor; Claude Haiku 5.5 on AWS
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ AI model price wars
❌ Diluted exposure (our judgment)

✅ Plan

👀 Too early: keep an eye on it · conviction low (watchlist) (score 58)
Our judgment · Not financial advice · No return promises

🗓 Timeline

1 yrSmall models replace big ones for routine tasks
3 yrAI cost per task keeps falling
5 yrAI in most software by default

Timeline steps are our judgment unless marked as company plan.

MSFT

Microsoft · USA

👀 Too early: keep an eye on it 💪 Low conviction (watchlist) ⏳ 2026-2028

1Paper / research
2Limited preview
3Generally available
4Widely adopted

💡 Thesis

🎯 Major OpenAI partner; GPT-6.1 Terra launch
🔗 OpenAI's GPT-6.1 Terra: computer-use AI at half the price, with errors cut from
🔗 Mistral releases the weights of Large 4, a ~1-trillion-parameter open model aime

📅 Catalysts

✅ Major OpenAI partner; GPT-6.1 Terra launch
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ AI is one part of a huge business
❌ Diluted exposure (our judgment)

✅ Plan

👀 Too early: keep an eye on it · conviction low (watchlist) (score 57)
Our judgment · Not financial advice · No return promises

🗓 Timeline

1 yrCommon in office automation
3 yrAI agents handle routine desk tasks
5 yrMost form-filling work automated

Timeline steps are our judgment unless marked as company plan.

6857.T

Advantest · Japan

👀 Too early: keep an eye on it 💪 Medium conviction ⏳ Now-2027

1Paper
2Prototype
3Factories & orders
4Revenue

💡 Thesis

🎯 AI chip testing demand
🔗 Gluing chips together copper-to-copper (hybrid bonding)

📅 Catalysts

✅ AI chip testing demand
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Cyclical; valuation
❌ Cyclical; valuation
❌ SK hynix: hybrid bonding not ready for HBM4E.

✅ Plan

👀 Too early: keep an eye on it · conviction Medium (score 64)
Our judgment · Not financial advice · No return promises

🗓 Timeline

Step 1More orders for bonding machines
Step 2Used in some memory and logic stacks
Step 3Standard for top AI chips

Timeline steps are our judgment unless marked as company plan.

CGPOWER.NS

CG Power · India

👀 Too early: keep an eye on it 💪 Low conviction (watchlist) ⏳ 2026-2027

1Paper
2Prototype
3Factories & orders
4Revenue

💡 Thesis

🎯 Chip packaging JV with Renesas & Stars
🔗 Gluing chips together copper-to-copper (hybrid bonding)

📅 Catalysts

✅ Chip packaging JV with Renesas & Stars
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Execution; ramp delays
❌ Execution; ramp delays
❌ SK hynix: hybrid bonding not ready for HBM4E.

✅ Plan

👀 Too early: keep an eye on it · conviction low (watchlist) (score 52)
Our judgment · Not financial advice · No return promises

🗓 Timeline

Step 1More orders for bonding machines
Step 2Used in some memory and logic stacks
Step 3Standard for top AI chips

Timeline steps are our judgment unless marked as company plan.

000660.KS

SK hynix · Korea

👀 Too early: keep an eye on it 💪 High conviction ⏳ Now-2027

1Paper
2Prototype
3Factories & orders
4Revenue

💡 Thesis

🎯 Q3 (Oct 28): revenue ₩104.6T (+328% YoY), DRAM/HBM booked through 2029, long-term deals with 14 customers cover >50% of 2027-28 output
🔗 Gluing chips together copper-to-copper (hybrid bonding)

📅 Catalysts

✅ Q3 (Oct 28): revenue ₩104.6T (+328% YoY), DRAM/HBM booked through 2029, long-term deals with 14 customers cover >50% of 2027-28 output
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Memory price cycle reversal
❌ Memory price cycle reversal
❌ SK hynix: hybrid bonding not ready for HBM4E.

✅ Plan

👀 Too early: keep an eye on it · conviction High (score 80)
Our judgment · Not financial advice · No return promises

🗓 Timeline

Step 1More orders for bonding machines
Step 2Used in some memory and logic stacks
Step 3Standard for top AI chips

Timeline steps are our judgment unless marked as company plan.

Illustration for opportunity meta platforms

Meta Platforms · USA

👀 Too early: keep an eye on it 💪 High conviction ⏳ Now-2027

1Research
2Preview
3Shipping
4Widely adopted

💡 Thesis

🎯 Muse Wonder frontier model and Muse Spark Open
🔗 Meta's superintelligence lab ships its first frontier model, Muse Wonder
🔗 Meta releases Muse Spark Open, which it calls the most capable US open-weight mo

📅 Catalysts

✅ Muse Wonder frontier model and Muse Spark Open
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ AI spending is huge; returns unproven
❌ AI is a small direct revenue line vs ads (our judgment)

✅ Plan

👀 Too early: keep an eye on it · conviction High (score 71)
Our judgment · Not financial advice · No return promises

🗓 Timeline

1 yrRapid model updates
3 yrAgent products built on it
5 yrCompetitive position vs other labs unclear

Timeline steps are our judgment unless marked as company plan.

Illustration for opportunity mtar technologies

MTARTECH.NS

MTAR Technologies · India

👀 Too early: keep an eye on it 💪 Medium conviction ⏳ 2026-2028

1Design
2Testing
3Flight
4Operations

💡 Thesis

🎯 ISRO engine-parts supplier; GSLV-F18 success (Oct 30) and more ISRO launches planned this financial year
🔗 ISRO's GSLV-F18 puts NVS-03 in orbit, set to restore India's own GPS (NavIC)

📅 Catalysts

✅ ISRO engine-parts supplier; GSLV-F18 success (Oct 30) and more ISRO launches planned this financial year
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Space is one segment alongside clean energy and defence
❌ Not named in this launch's coverage; read-across (our judgment)

✅ Plan

👀 Too early: keep an eye on it · conviction Medium (score 59)
Our judgment · Not financial advice · No return promises

🗓 Timeline

1 yrNVS-03 commissioned; standalone NavIC positioning back
3 yrNVS-04/05 complete the new series
5 yrNavIC common in Indian phones and vehicles

Timeline steps are our judgment unless marked as company plan.

UPST

Upstart · USA

👀 Too early: keep an eye on it 💪 High conviction ⏳ 2026-2027

1Research / sandbox
2Pilot
3Live launch
4Scaled & regulated

💡 Thesis

🎯 Cash-flow and AI credit models gaining ground (sector signal from Plaid launch)
🔗 Plaid launches a transformer AI credit score it says predicts repayment 42% bett

📅 Catalysts

✅ Cash-flow and AI credit models gaining ground (sector signal from Plaid launch)
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Credit cycle and funding risk
❌ Not directly tied to Plaid; sector read-across only (our judgment)

✅ Plan

👀 Too early: keep an eye on it · conviction High (score 66)
Our judgment · Not financial advice · No return promises

🗓 Timeline

1 yrMore lenders test transaction-based scores
3 yrCommon in fintech lending
5 yrCould feed mainstream bank underwriting

Timeline steps are our judgment unless marked as company plan.

SOFI

SoFi Technologies · USA

👀 Too early: keep an eye on it 💪 Medium conviction ⏳ 2026-2028

1Research / sandbox
2Pilot
3Live launch
4Scaled & regulated

💡 Thesis

🎯 Fintech lender that could adopt cash-flow underwriting
🔗 Plaid launches a transformer AI credit score it says predicts repayment 42% bett

📅 Catalysts

✅ Fintech lender that could adopt cash-flow underwriting
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Credit losses in a downturn
❌ Read-across only (our judgment)

✅ Plan

👀 Too early: keep an eye on it · conviction Medium (score 58)
Our judgment · Not financial advice · No return promises

🗓 Timeline

1 yrMore lenders test transaction-based scores
3 yrCommon in fintech lending
5 yrCould feed mainstream bank underwriting

Timeline steps are our judgment unless marked as company plan.

Illustration for opportunity bajaj finance

BAJFINANCE.NS

Bajaj Finance · India

👀 Too early: keep an eye on it 💪 Low conviction (watchlist) ⏳ 2026-2028

1Research / sandbox
2Pilot
3Live launch
4Scaled & regulated

💡 Thesis

🎯 Large Indian consumer lender using data-driven credit (read-across)
🔗 Plaid launches a transformer AI credit score it says predicts repayment 42% bett

📅 Catalysts

✅ Large Indian consumer lender using data-driven credit (read-across)
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Indian credit-cycle risk
❌ Read-across only (our judgment)

✅ Plan

👀 Too early: keep an eye on it · conviction low (watchlist) (score 54)
Our judgment · Not financial advice · No return promises

🗓 Timeline

1 yrMore lenders test transaction-based scores
3 yrCommon in fintech lending
5 yrCould feed mainstream bank underwriting

Timeline steps are our judgment unless marked as company plan.

Illustration for opportunity janus henderson

JHG

Janus Henderson · USA

👀 Too early: keep an eye on it 💪 Medium conviction ⏳ 2026-2028

1Research / sandbox
2Pilot
3Live launch
4Scaled & regulated

💡 Thesis

🎯 Launched JQQQ tokenized Nasdaq-100 fund (Oct 27)
🔗 Janus Henderson puts a Nasdaq-100 fund on a blockchain (JQQQ)

📅 Catalysts

✅ Launched JQQQ tokenized Nasdaq-100 fund (Oct 27)
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Tokenized assets are a tiny share of its funds
❌ Small revenue impact near term (our judgment)

✅ Plan

👀 Too early: keep an eye on it · conviction Medium (score 58)
Our judgment · Not financial advice · No return promises

🗓 Timeline

1 yrMore index funds go on-chain
3 yrTokenized funds used as collateral in trading
5 yrPart of mainstream fund distribution if rules allow

Timeline steps are our judgment unless marked as company plan.

NDAQ

Nasdaq · USA

👀 Too early: keep an eye on it 💪 Low conviction (watchlist) ⏳ 2026-2028

1Research / sandbox
2Pilot
3Live launch
4Scaled & regulated

💡 Thesis

🎯 Licensed Nasdaq-100 to the first issuer-sponsored tokenized fund
🔗 Janus Henderson puts a Nasdaq-100 fund on a blockchain (JQQQ)

📅 Catalysts

✅ Licensed Nasdaq-100 to the first issuer-sponsored tokenized fund
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Licensing fees from tokenized funds are small
❌ Diluted exposure (our judgment)

✅ Plan

👀 Too early: keep an eye on it · conviction low (watchlist) (score 54)
Our judgment · Not financial advice · No return promises

🗓 Timeline

1 yrMore index funds go on-chain
3 yrTokenized funds used as collateral in trading
5 yrPart of mainstream fund distribution if rules allow

Timeline steps are our judgment unless marked as company plan.

Illustration for opportunity iqm quantum computers

IQMX

IQM Quantum Computers · Finland

👀 Too early: keep an eye on it 💪 High conviction ⏳ 2027+

1Lab physics
2Hardware demo
3Cloud access / customers
4Commercial advantage

💡 Thesis

🎯 Full 54-qubit entanglement on cloud Emerald
🔗 IQM entangles all 54 qubits on its cloud Emerald chip at once

📅 Catalysts

✅ Full 54-qubit entanglement on cloud Emerald
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Fidelity only just above threshold
❌ Small company in a crowded field (our judgment)

✅ Plan

👀 Too early: keep an eye on it · conviction High (score 75)
Our judgment · Not financial advice · No return promises

🗓 Timeline

1 yrLarger IQM chips
3 yrCustomer algorithms using full chips
5 yrCompetitive cloud offering if error rates fall

Timeline steps are our judgment unless marked as company plan.

🔒 Private

Altera · USA

👀 Too early: keep an eye on it 💪 Low conviction (watchlist) ⏳ 2027

1Paper
2Prototype
3Factories & orders
4Revenue

💡 Thesis

🎯 Agilex 2X FPGA launch (Oct 28)
🔗 Altera launches Agilex 2X FPGAs with 1.8× better performance per watt for AI inf

📅 Catalysts

✅ Agilex 2X FPGA launch (Oct 28)
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Not listed; FPGAs lose to custom chips for pure AI maths
❌ Performance figure is vendor-claimed

✅ Plan

👀 Too early: keep an eye on it · conviction low (watchlist) (score 71)
Our judgment · Not financial advice · No return promises

🗓 Timeline

Step 1Early customer designs.
Step 2In network gear and AI servers.
Step 3Replaced by the next generation.

Timeline steps are our judgment unless marked as company plan.

CEVA

Ceva · USA

👀 Too early: keep an eye on it 💪 Medium conviction ⏳ 2027-2029 (royalties lag)

1Paper
2Prototype
3Factories & orders
4Revenue

💡 Thesis

🎯 Next-gen NeuPro NPU IP (Oct 28); early access now, general availability Q2 2027
🔗 Ceva unveils next-gen NeuPro AI engines (2× throughput per mm²) and enters AI in

📅 Catalysts

✅ Next-gen NeuPro NPU IP (Oct 28); early access now, general availability Q2 2027
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Crowded IP market; vendor-claimed gains
❌ General availability not until Q2 2027

✅ Plan

👀 Too early: keep an eye on it · conviction Medium (score 62)
Our judgment · Not financial advice · No return promises

🗓 Timeline

Step 1Early-access licensees.
Step 2First chips using it ship.
Step 3Royalties if widely adopted.

Timeline steps are our judgment unless marked as company plan.

BESI.AS

Besi · Europe

👀 Too early: keep an eye on it 💪 High conviction ⏳ 2026-2028

1Paper
2Prototype
3Factories & orders
4Revenue

💡 Thesis

🎯 Hybrid bonding tied to Applied Materials partnership
🔗 Gluing chips together copper-to-copper (hybrid bonding)
🔗 China's DF1000 AI chip stacks memory directly on its logic, reaching 6.4 TB/s wi

📅 Catalysts

✅ Hybrid bonding tied to Applied Materials partnership
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Adoption timing slips
❌ Adoption timing slips
❌ SK hynix: hybrid bonding not ready for HBM4E.

✅ Plan

👀 Too early: keep an eye on it · conviction High (score 82)
Our judgment · Not financial advice · No return promises

🗓 Timeline

Step 1More orders for bonding machines
Step 2Used in some memory and logic stacks
Step 3Standard for top AI chips

Timeline steps are our judgment unless marked as company plan.

ADEA

Adeia · USA

👀 Too early: keep an eye on it 💪 Medium conviction ⏳ 2026-2028

1Paper
2Prototype
3Factories & orders
4Revenue

💡 Thesis

🎯 Licenses hybrid-bonding IP
🔗 Gluing chips together copper-to-copper (hybrid bonding)
🔗 China's DF1000 AI chip stacks memory directly on its logic, reaching 6.4 TB/s wi

📅 Catalysts

✅ Licenses hybrid-bonding IP
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Litigation outcomes
❌ Litigation outcomes
❌ SK hynix: hybrid bonding not ready for HBM4E.

✅ Plan

👀 Too early: keep an eye on it · conviction Medium (score 58)
Our judgment · Not financial advice · No return promises

🗓 Timeline

Step 1More orders for bonding machines
Step 2Used in some memory and logic stacks
Step 3Standard for top AI chips

Timeline steps are our judgment unless marked as company plan.

Illustration for opportunity one 97 communications paytm

PAYTM.NS

One 97 Communications (Paytm) · India

👀 Too early: keep an eye on it 💪 Medium conviction ⏳ 2026-2027

1Research / sandbox
2Pilot
3Live launch
4Scaled & regulated

💡 Thesis

🎯 UPI MDR 0.4% above ₹2,000 to start after Diwali, date to be notified
🔗 India delays a 0.4% UPI merchant fee to 'immediately after Diwali'

📅 Catalysts

✅ UPI MDR 0.4% above ₹2,000 to start after Diwali, date to be notified
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Fee could be cut or dropped after Supreme Court review
❌ Policy may change (our judgment)

✅ Plan

👀 Too early: keep an eye on it · conviction Medium (score 62)
Our judgment · Not financial advice · No return promises

🗓 Timeline

1 yrFee starts or is changed after court review
3 yrPayment apps earn more per transaction
5 yre-rupee used for programmable payments alongside UPI

Timeline steps are our judgment unless marked as company plan.

Illustration for opportunity procure space etf

UFO

Procure Space ETF · USA

👀 Too early: keep an eye on it 💪 Medium conviction ⏳ 2027+

1Design
2Testing
3Flight
4Operations

💡 Thesis

🎯 Basket exposure to space companies (our judgment)
🔗 SpaceX plans to catch both Starship stages on Flight 15
🔗 Rocket Lab pushes Neutron's first launch to 2027 after finding flaws in fuel-tan
🔗 ISRO's GSLV-F18 puts NVS-03 in orbit, set to restore India's own GPS (NavIC)

📅 Catalysts

✅ Basket exposure to space companies (our judgment)
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Many holdings are satellite/telecom, not rockets
❌ SpaceX and Blue Origin are private, so ETFs miss them (our judgment)

✅ Plan

👀 Too early: keep an eye on it · conviction Medium (score 54)
Our judgment · Not financial advice · No return promises

🗓 Timeline

1 yrCatch attempt result
3 yrFaster launch cadence
5 yrCheaper launches if reuse is routine

Timeline steps are our judgment unless marked as company plan.

Illustration for opportunity lloyds banking group

LYG

Lloyds Banking Group · UK

👀 Too early: keep an eye on it 💪 Low conviction (watchlist) ⏳ 2026-2027

1Research / sandbox
2Pilot
3Live launch
4Scaled & regulated

💡 Thesis

🎯 Agentic AI fraud platform >90% accuracy (Oct 26)
🔗 Lloyds' agentic AI fraud assistant passes 90% decision accuracy on 10,600 live c

📅 Catalysts

✅ Agentic AI fraud platform >90% accuracy (Oct 26)
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Fraud AI is a small cost lever for a large bank
❌ Diluted exposure (our judgment)

✅ Plan

👀 Too early: keep an eye on it · conviction low (watchlist) (score 48)
Our judgment · Not financial advice · No return promises

🗓 Timeline

1 yrRolled to more fraud types
3 yrAgentic AI common in bank investigations
5 yrLinked with anti-money-laundering checks

Timeline steps are our judgment unless marked as company plan.

GOOGL

Alphabet · USA

👀 Too early: keep an eye on it 💪 Medium conviction ⏳ 2027+

1Lab physics
2Hardware demo
3Cloud access / customers
4Commercial advantage

💡 Thesis

🎯 Magic-state cultivation on Willow (Oct 15) and Genie 4 world model
🔗 Google shows 'magic state cultivation' on its Willow chip, cutting errors about
🔗 Google DeepMind's Genie 4 generates interactive worlds with sound and dialogue
🔗 Google's Gemini 4 Carbon ranks #1 on a new independent AI index, first given to

📅 Catalysts

✅ Magic-state cultivation on Willow (Oct 15) and Genie 4 world model
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Quantum and world models are tiny vs. its ad and cloud business
❌ Exposure is diluted by a huge core business (our judgment)

✅ Plan

👀 Too early: keep an eye on it · conviction Medium (score 56)
Our judgment · Not financial advice · No return promises

🗓 Timeline

1 yrRepeated on bigger chips
3 yrCombined with logical qubits in algorithms
5 yrPart of an error-corrected machine if scaling works

Timeline steps are our judgment unless marked as company plan.

ICICIBANK.NS

ICICI Bank · India

👀 Too early: keep an eye on it 💪 Low conviction (watchlist) ⏳ 2026-2027

1Research / sandbox
2Pilot
3Live launch
4Scaled & regulated

💡 Thesis

🎯 Named in Reuters report: API link used in about $2M fraud at two non-bank lenders (Oct 8-9)
🔗 RBI presses banks to prove cyber defences as AI-assisted API attacks hit Indian

📅 Catalysts

✅ Named in Reuters report: API link used in about $2M fraud at two non-bank lenders (Oct 8-9)
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Reputation and regulatory risk; money impact small for its size
❌ At-risk read-across only (our judgment)

✅ Plan

👀 Too early: keep an eye on it · conviction low (watchlist) (score 48)
Our judgment · Not financial advice · No return promises

🗓 Timeline

1 yrTighter API security checks
3 yrHigher bank cybersecurity spending
5 yrAI-versus-AI fraud defence becomes standard

Timeline steps are our judgment unless marked as company plan.

Illustration for opportunity quick heal technologies

QUICKHEAL.NS

Quick Heal Technologies · India

👀 Too early: keep an eye on it 💪 Medium conviction ⏳ 2026-2028

1Research / sandbox
2Pilot
3Live launch
4Scaled & regulated

💡 Thesis

🎯 Indian cybersecurity vendor; RBI push may lift bank security spending (read-across)
🔗 RBI presses banks to prove cyber defences as AI-assisted API attacks hit Indian

📅 Catalysts

✅ Indian cybersecurity vendor; RBI push may lift bank security spending (read-across)
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Small company; banks often buy from global vendors
❌ Read-across only (our judgment)

✅ Plan

👀 Too early: keep an eye on it · conviction Medium (score 48)
Our judgment · Not financial advice · No return promises

🗓 Timeline

1 yrTighter API security checks
3 yrHigher bank cybersecurity spending
5 yrAI-versus-AI fraud defence becomes standard

Timeline steps are our judgment unless marked as company plan.

Illustration for opportunity crispr therapeutics

CRSP

CRISPR Therapeutics · Switzerland/USA

👀 Too early: keep an eye on it 💪 Medium conviction ⏳ 2027+

1Preclinical
2Early trial (Ph 1/2)
3Late trial (Ph 3)
4Approved

💡 Thesis

🎯 FDA draft path for individualized gene editing (Oct 27); comments due Nov 27
🔗 FDA draft rule would let one approval cover many personalised gene-editing treat

📅 Catalysts

✅ FDA draft path for individualized gene editing (Oct 27); comments due Nov 27
📅 2026-11-27 FDA comment deadline: individualized gene-editing draft guidance

⚠️ Risks

❌ Its approved drug targets common mutations; the new path matters most for future platforms
❌ Benefit is indirect and years away (our judgment)

✅ Plan

👀 Too early: keep an eye on it · conviction Medium (score 55)
Our judgment · Not financial advice · No return promises

🗓 Timeline

1 yrGuidance finalised
3 yrFirst platform-style approvals
5 yrCustom edits become a routine path for rare diseases

Timeline steps are our judgment unless marked as company plan.

Illustration for opportunity intellia therapeutics

NTLA

Intellia Therapeutics · USA

👀 Too early: keep an eye on it 💪 Medium conviction ⏳ 2027+

1Preclinical
2Early trial (Ph 1/2)
3Late trial (Ph 3)
4Approved

💡 Thesis

🎯 FDA draft path for individualized gene editing (Oct 27)
🔗 FDA draft rule would let one approval cover many personalised gene-editing treat

📅 Catalysts

✅ FDA draft path for individualized gene editing (Oct 27)
📅 2026-11-27 FDA comment deadline: individualized gene-editing draft guidance

⚠️ Risks

❌ Clinical-stage; trial setbacks can hit the stock hard
❌ Read-across only (our judgment)

✅ Plan

👀 Too early: keep an eye on it · conviction Medium (score 51)
Our judgment · Not financial advice · No return promises

🗓 Timeline

1 yrGuidance finalised
3 yrFirst platform-style approvals
5 yrCustom edits become a routine path for rare diseases

Timeline steps are our judgment unless marked as company plan.

QNT

Quantinuum · USA

👀 Too early: keep an eye on it 💪 High conviction ⏳ 2027-2029

1Lab physics
2Hardware demo
3Cloud access / customers
4Commercial advantage

💡 Thesis

🎯 99.9995% logical fidelity on Helios; Apollo targeted for 2029
🔗 Quantinuum reports 99.9995% logical fidelity on Helios, five errors per million

📅 Catalysts

✅ 99.9995% logical fidelity on Helios; Apollo targeted for 2029
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Company-reported records; long path to profit
❌ Pure-play quantum stocks are volatile (our judgment)

✅ Plan

👀 Too early: keep an eye on it · conviction High (score 74)
Our judgment · Not financial advice · No return promises

🗓 Timeline

1 yrMore records on Helios
3 yrApollo prototypes
5 yrFault-tolerant machine if the 2029 plan holds

Timeline steps are our judgment unless marked as company plan.

4062.T

Ibiden · Japan

👀 Too early: keep an eye on it 💪 Medium conviction ⏳ 2027-2028

1Paper
2Prototype
3Factories & orders
4Revenue

💡 Thesis

🎯 Glass substrate validation with TSMC (reported)
🔗 Glass replacing plastic under AI chips (glass substrates)

📅 Catalysts

✅ Glass substrate validation with TSMC (reported)
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Glass may arrive late
❌ Glass may arrive late
❌ Glass timelines keep slipping (Tom's Hardware).

✅ Plan

👀 Too early: keep an eye on it · conviction Medium (score 61)
Our judgment · Not financial advice · No return promises

🗓 Timeline

Step 1Pilot lines and samples
Step 2First AI products on glass, if yields are good
Step 3Common in big AI chips

Timeline steps are our judgment unless marked as company plan.

Illustration for opportunity tata electronics fab

Unlisted (parent: Tata group cos)

Tata Electronics fab · India

👀 Too early: keep an eye on it 💪 Low conviction (watchlist) ⏳ 2027+ (est.)

1Paper
2Prototype
3Factories & orders
4Revenue

💡 Thesis

🎯 ₹91,526 cr Gujarat fab with PSMC
🔗 Glass replacing plastic under AI chips (glass substrates)

📅 Catalysts

✅ ₹91,526 cr Gujarat fab with PSMC
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Not directly investable
❌ Not directly investable
❌ Glass timelines keep slipping (Tom's Hardware).

✅ Plan

👀 Too early: keep an eye on it · conviction low (watchlist) (score 45)
Our judgment · Not financial advice · No return promises

🗓 Timeline

Step 1Pilot lines and samples
Step 2First AI products on glass, if yields are good
Step 3Common in big AI chips

Timeline steps are our judgment unless marked as company plan.

GLW

Corning · USA

👀 Too early: keep an eye on it 💪 Medium conviction ⏳ 2027-2029

1Paper
2Prototype
3Factories & orders
4Revenue

💡 Thesis

🎯 Glass core substrates
🔗 Glass replacing plastic under AI chips (glass substrates)

📅 Catalysts

✅ Glass core substrates
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Glass adoption slips
❌ Glass adoption slips
❌ Glass timelines keep slipping (Tom's Hardware).

✅ Plan

👀 Too early: keep an eye on it · conviction Medium (score 53)
Our judgment · Not financial advice · No return promises

🗓 Timeline

Step 1Pilot lines and samples
Step 2First AI products on glass, if yields are good
Step 3Common in big AI chips

Timeline steps are our judgment unless marked as company plan.

TSM

TSMC · Taiwan

👀 Too early: keep an eye on it 💪 Medium conviction ⏳ Q4 2026 (Jan 2027 results)

1Paper
2Prototype
3Factories & orders
4Revenue

💡 Thesis

🎯 2nm steep ramp; Q4 guide $52.3-54.3B
🔗 Glass replacing plastic under AI chips (glass substrates)

📅 Catalysts

✅ 2nm steep ramp; Q4 guide $52.3-54.3B
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Taiwan risk; overseas fab margin drag
❌ Taiwan risk; overseas fab margin drag
❌ Glass timelines keep slipping (Tom's Hardware).

✅ Plan

👀 Too early: keep an eye on it · conviction Medium (score 57)
Our judgment · Not financial advice · No return promises

🗓 Timeline

Step 1Pilot lines and samples
Step 2First AI products on glass, if yields are good
Step 3Common in big AI chips

Timeline steps are our judgment unless marked as company plan.

Illustration for opportunity prime medicine

PRME

Prime Medicine · USA

👀 Too early: keep an eye on it 💪 High conviction ⏳ 2027

1Preclinical
2Early trial (Ph 1/2)
3Late trial (Ph 3)
4Approved

💡 Thesis

🎯 First in-body prime edit (Wilson's disease); data expected next year
🔗 First prime edit inside a human body: Prime Medicine treats a Wilson's disease p
🔗 FDA draft rule would let one approval cover many personalised gene-editing treat

📅 Catalysts

✅ First in-body prime edit (Wilson's disease); data expected next year
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Single patient, no efficacy data yet
❌ Clinical-stage biotech can fall sharply on trial news (our judgment)

✅ Plan

👀 Too early: keep an eye on it · conviction High (score 68)
Our judgment · Not financial advice · No return promises

🗓 Timeline

1 yrFirst safety and copper data
3 yrMore patients dosed if safe
5 yrPossible approval path if efficacy holds

Timeline steps are our judgment unless marked as company plan.

Illustration for opportunity defiance quantum etf

QTUM

Defiance Quantum ETF · USA

👀 Too early: keep an eye on it 💪 Medium conviction ⏳ 2027+

1Lab physics
2Hardware demo
3Cloud access / customers
4Commercial advantage

💡 Thesis

🎯 Basket exposure to quantum and related computing firms (our judgment)
🔗 Google shows 'magic state cultivation' on its Willow chip, cutting errors about
🔗 Quantinuum reports 99.9995% logical fidelity on Helios, five errors per million
🔗 IQM entangles all 54 qubits on its cloud Emerald chip at once

📅 Catalysts

✅ Basket exposure to quantum and related computing firms (our judgment)
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Holdings are mostly large tech, not pure quantum
❌ Check holdings before assuming quantum exposure

✅ Plan

👀 Too early: keep an eye on it · conviction Medium (score 47)
Our judgment · Not financial advice · No return promises

🗓 Timeline

1 yrRepeated on bigger chips
3 yrCombined with logical qubits in algorithms
5 yrPart of an error-corrected machine if scaling works

Timeline steps are our judgment unless marked as company plan.

Illustration for opportunity globalfoundries

GFS

GlobalFoundries · USA

👀 Too early: keep an eye on it 💪 Medium conviction ⏳ 2027+

1Paper
2Prototype
3Factories & orders
4Revenue

💡 Thesis

🎯 Makes PsiQuantum's silicon-photonics chips, including the new optical switch (Oct 28)
🔗 PsiQuantum shows a light-based switch that rewires AI data-centre networks in 3

📅 Catalysts

✅ Makes PsiQuantum's silicon-photonics chips, including the new optical switch (Oct 28)
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Photonics is a small part of revenue
❌ Switch results are company-reported; only an 8×8 port count

✅ Plan

👀 Too early: keep an eye on it · conviction Medium (score 46)
Our judgment · Not financial advice · No return promises

🗓 Timeline

Step 1Samples tested by a few big cloud firms.
Step 2First switches in real AI clusters if tests go well.
Step 3Could be a standard part of AI networks, or lose to cheaper electrical switches.

Timeline steps are our judgment unless marked as company plan.

RKLB

Rocket Lab · USA

👀 Too early: keep an eye on it 💪 Medium conviction ⏳ 2027

1Design
2Testing
3Flight
4Operations

💡 Thesis

🎯 Neutron debut slipped to 2027
🔗 Rocket Lab pushes Neutron's first launch to 2027 after finding flaws in fuel-tan

📅 Catalysts

✅ Neutron debut slipped to 2027
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Further schedule slips
❌ Delay already announced; execution risk remains

✅ Plan

👀 Too early: keep an eye on it · conviction Medium (score 56)
Our judgment · Not financial advice · No return promises

🗓 Timeline

1 yrStage tests finish
3 yrFirst launch
5 yrRegular flights if debut succeeds

Timeline steps are our judgment unless marked as company plan.

Illustration for opportunity kaynes technology

KAYNES.NS

Kaynes Technology · India

👀 Too early: keep an eye on it 💪 Low conviction (watchlist) ⏳ 2026-2027

1Paper
2Prototype
3Factories & orders
4Revenue

💡 Thesis

🎯 OSAT plant in commercial production (ISM)
🔗 Making noisy analog AI chips accurate (weight splitting)

📅 Catalysts

✅ OSAT plant in commercial production (ISM)
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Small scale; high expectations priced in
❌ Small scale; high expectations priced in
❌ System-level review: specialised analog/physical accelerators can show 'little or no advantage' once conversion, control and memory access are counted.

✅ Plan

👀 Too early: keep an eye on it · conviction low (watchlist) (score 34)
Our judgment · Not financial advice · No return promises

🗓 Timeline

Step 1Follow-up papers
Step 2Shown on real Micron test silicon
Step 3Could ship inside analog-AI memory products

Timeline steps are our judgment unless marked as company plan.

Illustration for opportunity tower semiconductor

TSEM

Tower Semiconductor · Israel

👀 Too early: keep an eye on it 💪 Medium conviction ⏳ 2026-2029

1Paper
2Prototype
3Factories & orders
4Revenue

💡 Thesis

🎯 Photonics foundry for startups
🔗 Computing with light instead of electricity (optical AI chip)
🔗 A light switch sending 200 billion bits per second per colour

📅 Catalysts

✅ Photonics foundry for startups
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Customer concentration
❌ Customer concentration
❌ Nature Photonics review (2026): in the cloud, losses and light-to-electric conversions 'dominate power' and have so far prevented commercial photonic tensor cores.

✅ Plan

👀 Too early: keep an eye on it · conviction Medium (score 49)
Our judgment · Not financial advice · No return promises

🗓 Timeline

Step 1Lab demos and first test chips
Step 2Pilot units in a few data centres, if tests go well
Step 3Could be a niche accelerator for specific AI jobs

Timeline steps are our judgment unless marked as company plan.

Illustration for opportunity twist bioscience

TWST

Twist Bioscience · USA

👀 Too early: keep an eye on it 💪 Low conviction (watchlist) ⏳ 2029+

1Lab proof
2Prototype system
3Pilot archives
4Commercial service

💡 Thesis

🎯 DNA synthesis maker; DNA storage would need cheap writing (our judgment)
🔗 A porous silica chip writes 11× more DNA per spot, a step to cheaper 'DNA drives
🔗 HKU's 'Gungnir' code recovers data from DNA damaged like 400-year-old archives
🔗 AtlasBase unveils 'Thalia', a DNA-writing chip with 5.6 billion sites, 700× earl

📅 Catalysts

✅ DNA synthesis maker; DNA storage would need cheap writing (our judgment)
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ DNA storage is not a material revenue line today (our judgment)
❌ Link to these papers is our judgment, not company news

✅ Plan

👀 Too early: keep an eye on it · conviction low (watchlist) (score 33)
Our judgment · Not financial advice · No return promises

🗓 Timeline

1 yrFollow-up with denser arrays
3 yrPrototype DNA-writing chips
5 yrCould cut DNA-write cost if paired with fab partners

Timeline steps are our judgment unless marked as company plan.

BRN.AX

BrainChip · Australia

👀 Too early: keep an eye on it 💪 Speculative conviction ⏳ 2027+

1Paper
2Prototype
3Factories & orders
4Revenue

💡 Thesis

🎯 Edge neuromorphic chips
🔗 Brain-like chip that can learn on the device (neuromorphic)

📅 Catalysts

✅ Edge neuromorphic chips
📅 2026-11-03 AMD Q3 2026 results (after US market close)

⚠️ Risks

❌ Tiny revenue; dilution
❌ Tiny revenue; dilution
❌ BrainChip, a listed neuromorphic pioneer, reported only US$1.22M revenue and a US$12.0M loss for H1 2026 — commercial demand for brain-like chips is still tiny.

✅ Plan

👀 Too early: keep an eye on it · conviction Speculative (score 55)
Our judgment · Not financial advice · No return promises

🗓 Timeline

Step 1More FPGA demos
Step 2First real chips (tape-out)
Step 3Possible use in robots and cameras

Timeline steps are our judgment unless marked as company plan.

AMD

AMD · USA

👀 Too early: keep an eye on it 💪 Low conviction (watchlist) ⏳ Nov 3 2026

1Paper
2Prototype
3Factories & orders
4Revenue

💡 Thesis

🎯 Q3 results Nov 3 2026; Q3 revenue guide ~$13.0B
🔗 Brain-like chip that can learn on the device (neuromorphic)

📅 Catalysts

✅ Q3 results Nov 3 2026; Q3 revenue guide ~$13.0B
📅 2026-11-03 AMD Q3 2026 results (after US market close)

⚠️ Risks

❌ Execution vs Nvidia
❌ Execution vs Nvidia
❌ BrainChip, a listed neuromorphic pioneer, reported only US$1.22M revenue and a US$12.0M loss for H1 2026 — commercial demand for brain-like chips is still tiny.

✅ Plan

👀 Too early: keep an eye on it · conviction low (watchlist) (score 39)
Our judgment · Not financial advice · No return promises

🗓 Timeline

Step 1More FPGA demos
Step 2First real chips (tape-out)
Step 3Possible use in robots and cameras

Timeline steps are our judgment unless marked as company plan.

Micron · USA/India

👀 Too early: keep an eye on it 💪 Medium conviction ⏳ 2026-2029

1Paper
2Prototype
3Factories & orders
4Revenue

💡 Thesis

🎯 Analog in-memory & HBM; Gujarat plant
🔗 Making noisy analog AI chips accurate (weight splitting)
🔗 Storing many values in one memory spot (superposed analog memory)

📅 Catalysts

✅ Analog in-memory & HBM; Gujarat plant
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Memory cycle
❌ Memory cycle
❌ System-level review: specialised analog/physical accelerators can show 'little or no advantage' once conversion, control and memory access are counted.

✅ Plan

👀 Too early: keep an eye on it · conviction Medium (score 51)
Our judgment · Not financial advice · No return promises

🗓 Timeline

Step 1Follow-up papers
Step 2Shown on real Micron test silicon
Step 3Could ship inside analog-AI memory products

Timeline steps are our judgment unless marked as company plan.

LAES

SEALSQ · Switzerland

👀 Too early: keep an eye on it 💪 Medium conviction ⏳ 2026-2030

1Lab physics
2Hardware demo
3Cloud access / customers
4Commercial advantage

💡 Thesis

🎯 Post-quantum security chips; Alice & Bob estimate (Oct 30) adds urgency to upgrades
🔗 Alice & Bob: about 22,600 'cat qubits' could break RSA-2048 encryption in about

📅 Catalysts

✅ Post-quantum security chips; Alice & Bob estimate (Oct 30) adds urgency to upgrades
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Small, volatile company; post-quantum revenue still early
❌ Speculative; not linked to Alice & Bob (our judgment)

✅ Plan

👀 Too early: keep an eye on it · conviction Medium (score 49)
Our judgment · Not financial advice · No return promises

🗓 Timeline

1 yrBanks and governments speed up post-quantum upgrades
3 yrFirst small error-corrected cat-qubit demos
5 yr+Machines of this size remain uncertain

Timeline steps are our judgment unless marked as company plan.

AIXA

Aixtron · Europe

👀 Too early: keep an eye on it 💪 Low conviction (watchlist) ⏳ 2029+

1Paper
2Prototype
3Factories & orders
4Revenue

💡 Thesis

🎯 Tools for 2D-material transistors
🔗 Stacking atom-thin transistors on top of each other (2D CFETs)

📅 Catalysts

✅ Tools for 2D-material transistors
No big event expected in the next 90 days in our calendar

⚠️ Risks

❌ Very long timeline
❌ Very long timeline
❌ Linked research is 6-10 years from market (our formula)

✅ Plan

👀 Too early: keep an eye on it · conviction low (watchlist) (score 26)
Our judgment · Not financial advice · No return promises

🗓 Timeline

Step 1More lab tests on contacts
Step 2Research pilot lines
Step 3Still mostly research; mass use likely in the 2030s

Timeline steps are our judgment unless marked as company plan.

📋 Full table view (filters, bull vs bear, scores)

⏱ 21 min read

🎯 Opportunities, earliest first

Every row links back to the research that could drive it. Earlier stage = more upside, more risk.

Stage:

Region:

Conviction:

No stage moves since 2026-10-30.

Added:

Stage	Company (date added)	Region	Stock code	Catalyst	Timeline	Risk	Conviction	Driven by research
Stage 2 · Prototype S2 (10-30) → S2 (10-31)	Alphabet Added 2026-10-30	USA	GOOGL	Magic-state cultivation on Willow (Oct 15) and Genie 4 world model	2027+	Quantum and world models are tiny vs. its ad and cloud business 🐂 vs 🐻 + score 🐂 Magic-state cultivation on Willow (Oct 15) and Genie 4 world model. Exposure is diluted by a huge core business (our judgment) (judgment) Conviction score 56 = 0.4×linked readiness 60 + 0.4×20×exposure 2/5 + 0.2×20×execution 4/5 → Medium. Our judgment.	Medium	#12 Google shows 'magic state cultivation'… #22 Google DeepMind's Genie 4 generates in… #51 Google's Gemini 4 Carbon ranks #1 on a… #53 Anthropic's Claude Haiku 5.5 costs abo…
Stage 2 · Limited preview S2 (10-30) → S2 (10-31)	Quantinuum Added 2026-10-30	USA	QNT	99.9995% logical fidelity on Helios; Apollo targeted for 2029	2027-2029	Company-reported records; long path to profit 🐂 vs 🐻 + score 🐂 99.9995% logical fidelity on Helios; Apollo targeted for 2029. Pure-play quantum stocks are volatile (our judgment) (judgment) Conviction score 74 = 0.4×linked readiness 54 + 0.4×20×exposure 5/5 + 0.2×20×execution 3/5 → High. Our judgment.	High	#13 Quantinuum reports 99.9995% logical fi…
Stage 2 · Hardware demo S2 (10-30) → S2 (10-31)	IQM Quantum Computers Added 2026-10-30	Finland	IQMX	Full 54-qubit entanglement on cloud Emerald	2027+	Fidelity only just above threshold 🐂 vs 🐻 + score 🐂 Full 54-qubit entanglement on cloud Emerald. Small company in a crowded field (our judgment) (judgment) Conviction score 75 = 0.4×linked readiness 67 + 0.4×20×exposure 5/5 + 0.2×20×execution 2/5 → High. Our judgment.	High	#14 IQM entangles all 54 qubits on its clo…
Stage 2 · Hardware demo S2 (10-30) → S2 (10-31)	Defiance Quantum ETF Added 2026-10-30	USA	QTUM	Basket exposure to quantum and related computing firms (our judgment)	2027+	Holdings are mostly large tech, not pure quantum 🐂 vs 🐻 + score 🐂 Basket exposure to quantum and related computing firms (our judgment). Check holdings before assuming quantum exposure (judgment) Conviction score 47 = 0.4×linked readiness 47 + 0.4×20×exposure 2/5 + 0.2×20×execution 3/5 → Medium. Our judgment.	Medium	#12 Google shows 'magic state cultivation'… #13 Quantinuum reports 99.9995% logical fi… #14 IQM entangles all 54 qubits on its clo… #1 Alice & Bob: about 22,600 'cat qubits'…
Stage 2 · Hardware demo S2 (10-30) → S2 (10-31)	Prime Medicine Added 2026-10-30	USA	PRME	First in-body prime edit (Wilson's disease); data expected next year	2027	Single patient, no efficacy data yet 🐂 vs 🐻 + score 🐂 First in-body prime edit (Wilson's disease); data expected next year. Clinical-stage biotech can fall sharply on trial news (our judgment) (judgment) Conviction score 68 = 0.4×linked readiness 50 + 0.4×20×exposure 5/5 + 0.2×20×execution 2/5 → High. Our judgment.	High	#15 First prime edit inside a human body: … #3 FDA draft rule would let one approval …
Stage 2 · Early trial (Ph 1/2) S2 (10-30) → S2 (10-31)	Meta Platforms Added 2026-10-30	USA	META	Muse Wonder frontier model and Muse Spark Open	Now-2027	AI spending is huge; returns unproven 🐂 vs 🐻 + score 🐂 Muse Wonder frontier model and Muse Spark Open. AI is a small direct revenue line vs ads (our judgment) (judgment) Conviction score 71 = 0.4×linked readiness 78 + 0.4×20×exposure 3/5 + 0.2×20×execution 4/5 → High. Our judgment.	High	#21 Meta's superintelligence lab ships its… #23 Meta releases Muse Spark Open, which i…
Stage 2 · Preview S2 (10-30) → S2 (10-31)	Rocket Lab Added 2026-10-30	USA	RKLB	Neutron debut slipped to 2027	2027	Further schedule slips 🐂 vs 🐻 + score 🐂 Neutron debut slipped to 2027. Delay already announced; execution risk remains (judgment) Conviction score 56 = 0.4×linked readiness 41 + 0.4×20×exposure 4/5 + 0.2×20×execution 2/5 → Medium. Our judgment.	Medium	#28 Rocket Lab pushes Neutron's first laun…
Stage 2 · Testing S2 (10-30) → S2 (10-31)	Procure Space ETF Added 2026-10-30	USA	UFO	Basket exposure to space companies (our judgment)	2027+	Many holdings are satellite/telecom, not rockets 🐂 vs 🐻 + score 🐂 Basket exposure to space companies (our judgment). SpaceX and Blue Origin are private, so ETFs miss them (our judgment) (judgment) Conviction score 54 = 0.4×linked readiness 64 + 0.4×20×exposure 2/5 + 0.2×20×execution 3/5 → Medium. Our judgment.	Medium	#27 SpaceX plans to catch both Starship st… #28 Rocket Lab pushes Neutron's first laun… #2 ISRO's GSLV-F18 puts NVS-03 in orbit, …
Stage 2 · Testing S2 (10-30) → S2 (10-31)	Twist Bioscience Added 2026-10-30	USA	TWST	DNA synthesis maker; DNA storage would need cheap writing (our judgment)	2029+	DNA storage is not a material revenue line today (our judgment) 🐂 vs 🐻 + score 🐂 DNA synthesis maker; DNA storage would need cheap writing (our judgment). Link to these papers is our judgment, not company news (judgment) Conviction score 33 = 0.4×linked readiness 33 + 0.4×20×exposure 1/5 + 0.2×20×execution 3/5 → 👀 too early. Our judgment.	Watch	#11 A porous silica chip writes 11× more D… #9 HKU's 'Gungnir' code recovers data fro… #8 AtlasBase unveils 'Thalia', a DNA-writ…
Stage 1 · Lab proof S1 (10-28) → S1 (10-29) → S1 (10-30) → S1 (10-31)	Aixtron Added 2026-10-28	Europe	AIXA	Tools for 2D-material transistors	2029+	Very long timeline 🐂 vs 🐻 + score 🐂 Tools for 2D-material transistors. Very long timeline (judgment) Linked research is 6-10 years from market (our formula) (judgment) Conviction score 26 = 0.4×linked readiness 16 + 0.4×20×exposure 1/5 + 0.2×20×execution 3/5 → 👀 too early. Our judgment.	Watch	#40 Stacking atom-thin transistors on top …
Stage 1 · Paper S1 (10-28) → S1 (10-29) → S1 (10-30) → S1 (10-31)	Micron Added 2026-10-28	USA/India	MU	Analog in-memory & HBM; Gujarat plant	2026-2029	Memory cycle 🐂 vs 🐻 + score 🐂 Analog in-memory & HBM; Gujarat plant. Memory cycle (judgment) System-level review: specialised analog/physical accelerators can show 'little or no advantage' once conversion, control and memory access are counted. source Conviction score 51 = 0.4×linked readiness 27 + 0.4×20×exposure 3/5 + 0.2×20×execution 4/5 → Medium. Our judgment.	Medium	#38 Making noisy analog AI chips accurate … #37 Storing many values in one memory spot…
Stage 2 · Prototype S2 (10-28) → S2 (10-29) → S2 (10-30) → S2 (10-31)	Ibiden Added 2026-10-28 ^{[source 14]}	Japan	4062.T	Glass substrate validation with TSMC (reported)	2027-2028	Glass may arrive late 🐂 vs 🐻 + score 🐂 Glass substrate validation with TSMC (reported). Glass may arrive late (judgment) Glass timelines keep slipping (Tom's Hardware). source Conviction score 61 = 0.4×linked readiness 52 + 0.4×20×exposure 3/5 + 0.2×20×execution 4/5 → Medium. Our judgment.	Medium	#42 Glass replacing plastic under AI chips…
Stage 2 · Prototype S2 (10-28) → S2 (10-29) → S2 (10-30) → S2 (10-31)	Tata Electronics fab Added 2026-10-28 ^{[source 8]}	India	Unlisted (parent: Tata group cos)	₹91,526 cr Gujarat fab with PSMC	2027+ (est.)	Not directly investable 🐂 vs 🐻 + score 🐂 ₹91,526 cr Gujarat fab with PSMC. Not directly investable (judgment) Glass timelines keep slipping (Tom's Hardware). source Conviction score 45 = 0.4×linked readiness 52 + 0.4×20×exposure 2/5 + 0.2×20×execution 2/5 → 👀 too early. Not directly investable → 👀 too early. Our judgment.	Watch	#42 Glass replacing plastic under AI chips…
Stage 2 · Prototype S2 (10-28) → S2 (10-29) → S2 (10-30) → S2 (10-31)	Coherent Added 2026-10-28	USA	COHR	Lasers/optics for photonic links & CPO	2026-2029	Optics price wars 🐂 vs 🐻 + score 🐂 Lasers/optics for photonic links & CPO. Optics price wars (judgment) Nvidia networking SVP: 'If you can use copper, use copper' — optics only where distance forces it. source Conviction score 64 = 0.4×linked readiness 51 + 0.4×20×exposure 4/5 + 0.2×20×execution 3/5 → Medium. Our judgment.	Medium	#36 A light-based 'highway' between AI chi… #39 A light switch sending 200 billion bit… #46 Silicon photonics reaches high-volume …
Stage 2 · Prototype S2 (10-28) → S2 (10-29) → S2 (10-30) → S2 (10-31)	Tower Semiconductor Added 2026-10-28	Israel	TSEM	Photonics foundry for startups	2026-2029	Customer concentration 🐂 vs 🐻 + score 🐂 Photonics foundry for startups. Customer concentration (judgment) Nature Photonics review (2026): in the cloud, losses and light-to-electric conversions 'dominate power' and have so far prevented commercial photonic tensor cores. source Nvidia networking SVP: 'If you can use copper, use copper' — optics only where distance forces it. source Conviction score 49 = 0.4×linked readiness 34 + 0.4×20×exposure 3/5 + 0.2×20×execution 3/5 → Medium. Our judgment.	Medium	#35 Computing with light instead of electr… #39 A light switch sending 200 billion bit…
Stage 2 · Prototype S2 (10-28) → S2 (10-29) → S2 (10-30) → S2 (10-31)	Corning Added 2026-10-28	USA	GLW	Glass core substrates	2027-2029	Glass adoption slips 🐂 vs 🐻 + score 🐂 Glass core substrates. Glass adoption slips (judgment) Glass timelines keep slipping (Tom's Hardware). source Conviction score 53 = 0.4×linked readiness 52 + 0.4×20×exposure 2/5 + 0.2×20×execution 4/5 → Medium. Our judgment.	Medium	#42 Glass replacing plastic under AI chips…
Stage 2 · Prototype S2 (10-28) → S2 (10-29) → S2 (10-30) → S2 (10-31)	BrainChip Added 2026-10-28	Australia	BRN.AX	Edge neuromorphic chips	2027+	Tiny revenue; dilution 🐂 vs 🐻 + score 🐂 Edge neuromorphic chips. Tiny revenue; dilution (judgment) BrainChip, a listed neuromorphic pioneer, reported only US$1.22M revenue and a US$12.0M loss for H1 2026 — commercial demand for brain-like chips is still tiny. source Conviction score 55 = 0.4×linked readiness 28 + 0.4×20×exposure 5/5 + 0.2×20×execution 1/5 → Speculative. Our judgment.	Speculative	#43 Brain-like chip that can learn on the …
Stage 2 · Prototype S2 (10-29) → S2 (10-30) → S2 (10-31)	Ceva Added 2026-10-29	USA	CEVA	Next-gen NeuPro NPU IP (Oct 28); early access now, general availability Q2 2027	2027-2029 (royalties lag)	Crowded IP market; vendor-claimed gains 🐂 vs 🐻 + score 🐂 Next-gen NeuPro NPU IP (Oct 28); early access now, general availability Q2 2027. General availability not until Q2 2027 source Conviction score 62 = 0.4×linked readiness 66 + 0.4×20×exposure 3/5 + 0.2×20×execution 3/5 → Medium. Our judgment.	Medium	#34 Ceva unveils next-gen NeuPro AI engine…
Stage 2 · Prototype S2 (10-29) → S2 (10-30) → S2 (10-31)	GlobalFoundries Added 2026-10-29	USA	GFS	Makes PsiQuantum's silicon-photonics chips, including the new optical switch (Oct 28)	2027+	Photonics is a small part of revenue 🐂 vs 🐻 + score 🐂 Makes PsiQuantum's silicon-photonics chips, including the new optical switch (Oct 28). Switch results are company-reported; only an 8×8 port count source Conviction score 46 = 0.4×linked readiness 46 + 0.4×20×exposure 2/5 + 0.2×20×execution 3/5 → Medium. Our judgment.	Medium	#30 PsiQuantum shows a light-based switch …
Stage 2 · Prototype S2 (10-29) → S2 (10-30) → S2 (10-31)	Altera Added 2026-10-29	USA	Private (Silver Lake majority; Intel minority)	Agilex 2X FPGA launch (Oct 28)	2027	Not listed; FPGAs lose to custom chips for pure AI maths 🐂 vs 🐻 + score 🐂 Agilex 2X FPGA launch (Oct 28). Performance figure is vendor-claimed source Conviction score 71 = 0.4×linked readiness 67 + 0.4×20×exposure 4/5 + 0.2×20×execution 3/5 → 👀 too early. Not directly investable → 👀 too early. Our judgment.	Watch	#33 Altera launches Agilex 2X FPGAs with 1…
Stage 3 · Factories & orders S3 (10-28) → S3 (10-29) → S3 (10-30) → S3 (10-31)	Besi Added 2026-10-28 ^{[source 14]}	Europe	BESI.AS	Hybrid bonding tied to Applied Materials partnership	2026-2028	Adoption timing slips 🐂 vs 🐻 + score 🐂 Hybrid bonding tied to Applied Materials partnership. Adoption timing slips (judgment) SK hynix: hybrid bonding not ready for HBM4E. source Conviction score 82 = 0.4×linked readiness 66 + 0.4×20×exposure 5/5 + 0.2×20×execution 4/5 → High. Our judgment.	High	#41 Gluing chips together copper-to-copper… #4 China's DF1000 AI chip stacks memory d…
Stage 3 · Factories & orders S3 (10-28) → S3 (10-29) → S3 (10-30) → S3 (10-31)	Advantest Added 2026-10-28 ^{[source 14]}	Japan	6857.T	AI chip testing demand	Now-2027	Cyclical; valuation 🐂 vs 🐻 + score 🐂 AI chip testing demand. Cyclical; valuation (judgment) SK hynix: hybrid bonding not ready for HBM4E. source Conviction score 64 = 0.4×linked readiness 80 + 0.4×20×exposure 2/5 + 0.2×20×execution 4/5 → Medium. Our judgment.	Medium	#41 Gluing chips together copper-to-copper…
Stage 3 · Factories & orders S3 (10-28) → S3 (10-29) → S3 (10-30) → S3 (10-31)	Kaynes Technology Added 2026-10-28 ^{[source 8]}	India	KAYNES.NS	OSAT plant in commercial production (ISM)	2026-2027	Small scale; high expectations priced in 🐂 vs 🐻 + score 🐂 OSAT plant in commercial production (ISM). Small scale; high expectations priced in (judgment) System-level review: specialised analog/physical accelerators can show 'little or no advantage' once conversion, control and memory access are counted. source Conviction score 34 = 0.4×linked readiness 35 + 0.4×20×exposure 1/5 + 0.2×20×execution 3/5 → 👀 too early. Our judgment.	Watch	#38 Making noisy analog AI chips accurate …
Stage 3 · Factories & orders S3 (10-28) → S3 (10-29) → S3 (10-30) → S3 (10-31)	CG Power Added 2026-10-28 ^{[source 8]}	India	CGPOWER.NS	Chip packaging JV with Renesas & Stars	2026-2027	Execution; ramp delays 🐂 vs 🐻 + score 🐂 Chip packaging JV with Renesas & Stars. Execution; ramp delays (judgment) SK hynix: hybrid bonding not ready for HBM4E. source Conviction score 52 = 0.4×linked readiness 80 + 0.4×20×exposure 1/5 + 0.2×20×execution 3/5 → 👀 too early. Our judgment.	Watch	#41 Gluing chips together copper-to-copper…
Stage 3 · Factories & orders S3 (10-28) → S3 (10-29) → S3 (10-30) → S3 (10-31)	Adeia Added 2026-10-28	USA	ADEA	Licenses hybrid-bonding IP	2026-2028	Litigation outcomes 🐂 vs 🐻 + score 🐂 Licenses hybrid-bonding IP. Litigation outcomes (judgment) SK hynix: hybrid bonding not ready for HBM4E. source Conviction score 58 = 0.4×linked readiness 66 + 0.4×20×exposure 3/5 + 0.2×20×execution 2/5 → Medium. Our judgment.	Medium	#41 Gluing chips together copper-to-copper… #4 China's DF1000 AI chip stacks memory d…
Stage 3 · Factories & orders S3 (10-28) → S3 (10-29) → S3 (10-30) → S3 (10-31)	STMicroelectronics Added 2026-10-28	Europe	STM	PIC100 photonics in volume	2026-2027	Auto/industrial weakness 🐂 vs 🐻 + score 🐂 PIC100 photonics in volume. Auto/industrial weakness (judgment) Nvidia networking SVP: 'If you can use copper, use copper' — optics only where distance forces it. source Conviction score 66 = 0.4×linked readiness 84 + 0.4×20×exposure 2/5 + 0.2×20×execution 4/5 → High. Our judgment.	High	#46 Silicon photonics reaches high-volume …
Stage 4 · Revenue S4 (10-28) → S4 (10-29) → S4 (10-30) → S4 (10-31)	SK hynix Added 2026-10-28 ^{[source 21]}	Korea	000660.KS	Q3 (Oct 28): revenue ₩104.6T (+328% YoY), DRAM/HBM booked through 2029, long-term deals with 14 customers cover >50% of 2027-28 output	Now-2027	Memory price cycle reversal 🐂 vs 🐻 + score 🐂 Q3 (Oct 28): revenue ₩104.6T (+328% YoY), DRAM/HBM booked through 2029, long-term deals with 14 customers cover >50% of 2027-28 output. Memory price cycle reversal (judgment) SK hynix: hybrid bonding not ready for HBM4E. source Conviction score 80 = 0.4×linked readiness 80 + 0.4×20×exposure 4/5 + 0.2×20×execution 4/5 → High. Our judgment.	High	#41 Gluing chips together copper-to-copper…
Stage 4 · Revenue S4 (10-28) → S4 (10-29) → S4 (10-30) → S4 (10-31)	TSMC Added 2026-10-28 ^{[source 2]}	Taiwan	TSM	2nm steep ramp; Q4 guide $52.3-54.3B	Q4 2026 (Jan 2027 results)	Taiwan risk; overseas fab margin drag 🐂 vs 🐻 + score 🐂 2nm steep ramp; Q4 guide $52.3-54.3B. Taiwan risk; overseas fab margin drag (judgment) Glass timelines keep slipping (Tom's Hardware). source Conviction score 57 = 0.4×linked readiness 52 + 0.4×20×exposure 2/5 + 0.2×20×execution 5/5 → Medium. Our judgment.	Medium	#42 Glass replacing plastic under AI chips…
Stage 4 · Revenue S4 (10-28) → S4 (10-29) → S4 (10-30) → S4 (10-31)	Nvidia Added 2026-10-28 ^{[source 1]}	USA	NVDA	Vera Rubin ramp in full production	Next results ~late Nov 2026 (est.)	Customer capex slowdown; China rules 🐂 vs 🐻 + score 🐂 Vera Rubin ramp in full production. Customer capex slowdown; China rules (judgment) Nvidia networking SVP: 'If you can use copper, use copper' — optics only where distance forces it. source Conviction score 63 = 0.4×linked readiness 67 + 0.4×20×exposure 2/5 + 0.2×20×execution 5/5 → Medium. Our judgment.	Medium	#46 Silicon photonics reaches high-volume … #39 A light switch sending 200 billion bit… #5 Mistral releases the weights of Large …
Stage 4 · Widely adopted S4 (10-28) → S4 (10-29) → S4 (10-30) → S4 (10-31)	AMD Added 2026-10-28 ^{[source 3]}	USA	AMD	Q3 results Nov 3 2026; Q3 revenue guide ~$13.0B	Nov 3 2026	Execution vs Nvidia 🐂 vs 🐻 + score 🐂 Q3 results Nov 3 2026; Q3 revenue guide ~$13.0B. Execution vs Nvidia (judgment) BrainChip, a listed neuromorphic pioneer, reported only US$1.22M revenue and a US$12.0M loss for H1 2026 — commercial demand for brain-like chips is still tiny. source Conviction score 39 = 0.4×linked readiness 28 + 0.4×20×exposure 1/5 + 0.2×20×execution 5/5 → 👀 too early. Our judgment.	Watch	#43 Brain-like chip that can learn on the …
Stage 3 · Factories & orders S3 (10-30) → S3 (10-31)	Janus Henderson Added 2026-10-30	USA	JHG	Launched JQQQ tokenized Nasdaq-100 fund (Oct 27)	2026-2028	Tokenized assets are a tiny share of its funds 🐂 vs 🐻 + score 🐂 Launched JQQQ tokenized Nasdaq-100 fund (Oct 27). Small revenue impact near term (our judgment) (judgment) Conviction score 58 = 0.4×linked readiness 74 + 0.4×20×exposure 2/5 + 0.2×20×execution 3/5 → Medium. Our judgment.	Medium	#47 Janus Henderson puts a Nasdaq-100 fund…
Stage 3 · Live launch S3 (10-30) → S3 (10-31)	Nasdaq Added 2026-10-30	USA	NDAQ	Licensed Nasdaq-100 to the first issuer-sponsored tokenized fund	2026-2028	Licensing fees from tokenized funds are small 🐂 vs 🐻 + score 🐂 Licensed Nasdaq-100 to the first issuer-sponsored tokenized fund. Diluted exposure (our judgment) (judgment) Conviction score 54 = 0.4×linked readiness 74 + 0.4×20×exposure 1/5 + 0.2×20×execution 4/5 → 👀 too early. Our judgment.	Watch	#47 Janus Henderson puts a Nasdaq-100 fund…
Stage 3 · Live launch S3 (10-30) → S3 (10-31)	Upstart Added 2026-10-30	USA	UPST	Cash-flow and AI credit models gaining ground (sector signal from Plaid launch)	2026-2027	Credit cycle and funding risk 🐂 vs 🐻 + score 🐂 Cash-flow and AI credit models gaining ground (sector signal from Plaid launch). Not directly tied to Plaid; sector read-across only (our judgment) (judgment) Conviction score 66 = 0.4×linked readiness 75 + 0.4×20×exposure 3/5 + 0.2×20×execution 3/5 → High. Our judgment.	High	#48 Plaid launches a transformer AI credit…
Stage 3 · Live launch S3 (10-30) → S3 (10-31)	SoFi Technologies Added 2026-10-30	USA	SOFI	Fintech lender that could adopt cash-flow underwriting	2026-2028	Credit losses in a downturn 🐂 vs 🐻 + score 🐂 Fintech lender that could adopt cash-flow underwriting. Read-across only (our judgment) (judgment) Conviction score 58 = 0.4×linked readiness 75 + 0.4×20×exposure 2/5 + 0.2×20×execution 3/5 → Medium. Our judgment.	Medium	#48 Plaid launches a transformer AI credit…
Stage 3 · Live launch S3 (10-30) → S3 (10-31)	Bajaj Finance Added 2026-10-30	India	BAJFINANCE.NS	Large Indian consumer lender using data-driven credit (read-across)	2026-2028	Indian credit-cycle risk 🐂 vs 🐻 + score 🐂 Large Indian consumer lender using data-driven credit (read-across). Read-across only (our judgment) (judgment) Conviction score 54 = 0.4×linked readiness 75 + 0.4×20×exposure 1/5 + 0.2×20×execution 4/5 → 👀 too early. Our judgment.	Watch	#48 Plaid launches a transformer AI credit…
Stage 3 · Live launch S3 (10-30) → S3 (10-31)	Lloyds Banking Group Added 2026-10-30	UK	LYG	Agentic AI fraud platform >90% accuracy (Oct 26)	2026-2027	Fraud AI is a small cost lever for a large bank 🐂 vs 🐻 + score 🐂 Agentic AI fraud platform >90% accuracy (Oct 26). Diluted exposure (our judgment) (judgment) Conviction score 48 = 0.4×linked readiness 61 + 0.4×20×exposure 1/5 + 0.2×20×execution 4/5 → 👀 too early. Our judgment.	Watch	#49 Lloyds' agentic AI fraud assistant pas…
Stage 3 · Live launch S3 (10-30) → S3 (10-31)	One 97 Communications (Paytm) Added 2026-10-30	India	PAYTM.NS	UPI MDR 0.4% above ₹2,000 to start after Diwali, date to be notified	2026-2027	Fee could be cut or dropped after Supreme Court review 🐂 vs 🐻 + score 🐂 UPI MDR 0.4% above ₹2,000 to start after Diwali, date to be notified. Policy may change (our judgment) (judgment) Conviction score 62 = 0.4×linked readiness 66 + 0.4×20×exposure 3/5 + 0.2×20×execution 3/5 → Medium. Our judgment.	Medium	#50 India delays a 0.4% UPI merchant fee t…
Stage 3 · Live launch S3 (10-30) → S3 (10-31)	Microsoft Added 2026-10-30	USA	MSFT	Major OpenAI partner; GPT-6.1 Terra launch	2026-2028	AI is one part of a huge business 🐂 vs 🐻 + score 🐂 Major OpenAI partner; GPT-6.1 Terra launch. Diluted exposure (our judgment) (judgment) Conviction score 57 = 0.4×linked readiness 83 + 0.4×20×exposure 1/5 + 0.2×20×execution 4/5 → 👀 too early. Our judgment.	Watch	#52 OpenAI's GPT-6.1 Terra: computer-use A… #5 Mistral releases the weights of Large …
Stage 3 · Generally available S3 (10-30) → S3 (10-31)	Amazon Added 2026-10-30	USA	AMZN	Anthropic investor; Claude Haiku 5.5 on AWS	2026-2028	AI model price wars 🐂 vs 🐻 + score 🐂 Anthropic investor; Claude Haiku 5.5 on AWS. Diluted exposure (our judgment) (judgment) Conviction score 58 = 0.4×linked readiness 85 + 0.4×20×exposure 1/5 + 0.2×20×execution 4/5 → 👀 too early. Our judgment.	Watch	#53 Anthropic's Claude Haiku 5.5 costs abo…
Stage 3 · Generally available S3 (10-31)	CRISPR Therapeutics Added 2026-10-31 🆕	Switzerland/USA	CRSP	FDA draft path for individualized gene editing (Oct 27); comments due Nov 27	2027+	Its approved drug targets common mutations; the new path matters most for future platforms 🐂 vs 🐻 + score 🐂 FDA draft path for individualized gene editing (Oct 27); comments due Nov 27. Benefit is indirect and years away (our judgment) (judgment) Conviction score 55 = 0.4×linked readiness 57 + 0.4×20×exposure 2/5 + 0.2×20×execution 4/5 → Medium. Our judgment.	Medium	#3 FDA draft rule would let one approval …
Stage 2 · Early trial (Ph 1/2) S2 (10-31)	Intellia Therapeutics Added 2026-10-31 🆕	USA	NTLA	FDA draft path for individualized gene editing (Oct 27)	2027+	Clinical-stage; trial setbacks can hit the stock hard 🐂 vs 🐻 + score 🐂 FDA draft path for individualized gene editing (Oct 27). Read-across only (our judgment) (judgment) Conviction score 51 = 0.4×linked readiness 57 + 0.4×20×exposure 2/5 + 0.2×20×execution 3/5 → Medium. Our judgment.	Medium	#3 FDA draft rule would let one approval …
Stage 2 · Early trial (Ph 1/2) S2 (10-31)	SEALSQ Added 2026-10-31 🆕	Switzerland	LAES	Post-quantum security chips; Alice & Bob estimate (Oct 30) adds urgency to upgrades	2026-2030	Small, volatile company; post-quantum revenue still early 🐂 vs 🐻 + score 🐂 Post-quantum security chips; Alice & Bob estimate (Oct 30) adds urgency to upgrades. Speculative; not linked to Alice & Bob (our judgment) (judgment) Conviction score 49 = 0.4×linked readiness 22 + 0.4×20×exposure 4/5 + 0.2×20×execution 2/5 → Medium. Our judgment.	Medium	#1 Alice & Bob: about 22,600 'cat qubits'…
Stage 3 · Cloud access / customers S3 (10-31)	MTAR Technologies Added 2026-10-31 🆕	India	MTARTECH.NS	ISRO engine-parts supplier; GSLV-F18 success (Oct 30) and more ISRO launches planned this financial year	2026-2028	Space is one segment alongside clean energy and defence 🐂 vs 🐻 + score 🐂 ISRO engine-parts supplier; GSLV-F18 success (Oct 30) and more ISRO launches planned this financial year. Not named in this launch's coverage; read-across (our judgment) (judgment) Conviction score 59 = 0.4×linked readiness 78 + 0.4×20×exposure 2/5 + 0.2×20×execution 3/5 → Medium. Our judgment.	Medium	#2 ISRO's GSLV-F18 puts NVS-03 in orbit, …
Stage 3 · Flight S3 (10-31)	ICICI Bank Added 2026-10-31 🆕	India	ICICIBANK.NS	Named in Reuters report: API link used in about $2M fraud at two non-bank lenders (Oct 8-9)	2026-2027	Reputation and regulatory risk; money impact small for its size 🐂 vs 🐻 + score 🐂 Named in Reuters report: API link used in about $2M fraud at two non-bank lenders (Oct 8-9). At-risk read-across only (our judgment) (judgment) Conviction score 48 = 0.4×linked readiness 60 + 0.4×20×exposure 1/5 + 0.2×20×execution 4/5 → 👀 too early. Our judgment.	Watch	#6 RBI presses banks to prove cyber defen…
Stage 3 · Live launch S3 (10-31)	Quick Heal Technologies Added 2026-10-31 🆕	India	QUICKHEAL.NS	Indian cybersecurity vendor; RBI push may lift bank security spending (read-across)	2026-2028	Small company; banks often buy from global vendors 🐂 vs 🐻 + score 🐂 Indian cybersecurity vendor; RBI push may lift bank security spending (read-across). Read-across only (our judgment) (judgment) Conviction score 48 = 0.4×linked readiness 60 + 0.4×20×exposure 2/5 + 0.2×20×execution 2/5 → Medium. Our judgment.	Medium	#6 RBI presses banks to prove cyber defen…

⚠️ Not financial advice. Conviction reflects our editorial view of how strong the research-to-revenue story is, not a buy or sell recommendation. Prices and facts change; do your own research and consult a registered investment adviser.

Opportunities ahead, traced to the lab

⚡ At a glance

🎯 Opportunity cards

STMicroelectronics · Europe

💡 Thesis

📅 Catalysts

⚠️ Risks

✅ Plan

🗓 Timeline

Nvidia · USA

💡 Thesis

📅 Catalysts

⚠️ Risks

✅ Plan

🗓 Timeline

Coherent · USA

💡 Thesis

📅 Catalysts

⚠️ Risks

✅ Plan

🗓 Timeline

Amazon · USA

💡 Thesis

📅 Catalysts

⚠️ Risks

✅ Plan

🗓 Timeline

Microsoft · USA

💡 Thesis

📅 Catalysts

⚠️ Risks

✅ Plan

🗓 Timeline

Advantest · Japan

💡 Thesis

📅 Catalysts

⚠️ Risks

✅ Plan

🗓 Timeline

CG Power · India

💡 Thesis

📅 Catalysts

⚠️ Risks

✅ Plan

🗓 Timeline

SK hynix · Korea

💡 Thesis

📅 Catalysts

⚠️ Risks

✅ Plan

🗓 Timeline

Meta Platforms · USA

💡 Thesis

📅 Catalysts

⚠️ Risks

✅ Plan

🗓 Timeline

MTAR Technologies · India

💡 Thesis

📅 Catalysts

⚠️ Risks

✅ Plan

🗓 Timeline

Upstart · USA

💡 Thesis

📅 Catalysts

⚠️ Risks

✅ Plan

🗓 Timeline

SoFi Technologies · USA

💡 Thesis

📅 Catalysts

⚠️ Risks

✅ Plan

🗓 Timeline

Bajaj Finance · India

💡 Thesis

📅 Catalysts

⚠️ Risks

✅ Plan