$OPTX — Syntec Optics:
The Precision Optics Sleeper—

While the market focuses on chips and transceivers, someone has to make the physical optical components that go inside them — the lenses, filters, coatings, and assemblies that make light bend correctly.
Syntec Optics is that company, and it’s just entering the AI data center market.

Syntec Optics is one of the largest custom and diverse end-market optics and photonics manufacturers in US, with product lines spanning defense, biomedical, space, and now AI data center optics.

4 Engines Driving OPTX Growth:
1.AI Data Center Optics — 2026 started with the beginning of the production stage for the AI data center product line — the company’s newest and highest-growth opportunity, transitioning from design-stage programs into revenue.
2.Gross Margin Inflection — Gross margin saw significant expansion, nearly doubling compared to both the prior quarter and prior year — from 12–13% to 24% in Q4 2025 — driven by improved manufacturing yields and cost discipline.

3.Defense Tech & Onshoring — Syntec sees positive momentum in defense tech, pointing to structural shifts toward domestic sourcing of optical systems driven by the 2026 Defense Authorization Act , which requires contractors to reduce reliance on adversary nations by 2030.

4.Space Optics — Syntec shipped over $2.6 million in space optics in 2025 and reached its 17,000th space optic manufactured , with sustained demand that could nearly 3x 2026 deliveries

$CIEN — Ciena Corporation:
The Coherent Optical Backbone—

Lumentum and $COHR connect GPUs inside the data center. Ciena connects the data centers themselves. It’s a different layer of the same photonics stack — and arguably the one with the deepest moat, given its decade-long lead in coherent optical technology.

WaveLogic 6 Extreme (WL6e), built on a 3nm process, is the industry’s first 1.6 Terabit-per-second coherent solution — and delivers a 50% reduction in power and space per bit versus prior technology.
4 Engines Driving CIEN Growth:

1.WaveLogic 6 Extreme (WL6e) — WL6e has reached 90 total customers, adding 18 new ones in Q1 FY2026 alone, with direct cloud provider revenue growing 76% year-over-year to represent 42% of total revenue.

2.WL6 Nano Pluggables — In FY2025, Ciena surpassed its target of more than doubling pluggable revenue, reaching over $168 million, with WL6 Nano 800G pluggables now shipping to multiple cloud providers for testing and certification.

3.Inside the Data Center (Nubis) — Following the Nubis acquisition, Ciena is now moving inside the data center, providing optical interconnects that allow thousands of GPUs to function as a single massive computer for scale-up and scale-out workloads.

4.Record Backlog — Ciena enters 2026 with a staggering $7.8 billion in orders. management has projected 2026 revenue between arnd $6 billion — roughly 24% growth at the midpoint📈

🔥 Lumentum had published investor presentation.

What drives $LITE ??

4 Engines Driving Lumentum Growth -

1. Cloud Transcievers : Optical transceivers are swappable hardware devices that convert electrical signals into optical (light) signals and vice-versa. Critical backbone of photonics in AI connect.

Lumentum indicated a Turning Point Reached in Cloud Transceivers. They have become “lead pack” of suppliers for 1.6T

2. OCS (Optical Circuit Switches): OCS are advanced, high-density network switching devices designed for AI and cloud data centers.
Ramping to >$1B 2027 Run Rate, OCS Order backlog of $400M+ to be fulfilled in 2H CY26

3. Scale-out CPO: In optics and AI data center networking, scale-out refers to the use of high-speed optical fiber links to connect multiple compute clusters, racks, or servers horizontally.
Scale-out creates vast, high-bandwidth fabrics essential for AI training and large-scale parallel processing
“Copper Wall” Requires Scale-Up CPO. Generational shift to optical as copper begins to reach limits

4. Scale-Up CPO:
Scale-Up means: electrical links connecting GPUs within and between a small number of racks

Opportunity for Lumentum is 3X - 4X larger than Scale-out CPO

The Optics Arms Race: >70% Growth CAGR

I like $LITE with multiple growth engines focused completely on AI Networking and huge available capacity to meet demand when compared to other Optic companies
(Coherent $COHR has more diversified range of services other than AI n/w)

$GOOG released TurboQuant -- a KV cache compression algorithm enabling AI inference with 6x less memory and 8x faster attention computation.

🚀Why is it a Bullish Case for $CRDO:

1. Efficient AI = More Interconnects:
- The 'Jevons Paradox' logic applies here: TurboQuant will likely accelerate AI adoption, which already requires a massive buildout of data center infrastructure.
- Every new rack, every new GPU cluster, every new inference node needs AECs and SerDes — Credo's core products.

2. TurboQuant could shift workloads toward more distributed inference:
- One scenario is that AI inference market share shifts from expensive GPUs with HBM to CPUs running on traditional server memory like DDR5 — since TurboQuant's 8x increase in KV cache speed makes this more feasible.
- CPU-based inference nodes means more interconnects, not fewer — a direct tailwind for Credo's HiWire AEC business

3 Key Products of Credo Tech:

a. Weaver (OmniConnect): TurboQuant's is making AI inference more efficient — compressing the data an AI needs to "think" through a long conversation or document. This data still needs to travel fast between the AI processor and memory. That's exactly the gap Weaver fills.
Weaver is engineered specifically to overcome memory bottlenecks in AI inference workloads, boosting I/O density by up to 10x

b. MicroLED ALCs (Active Laser Cable): These are special cables that use optics to send data over longer distances (up to 30 meters). This is cheaper alternative to traditional copper based fibers (Hyperlume acquisition)

c. ZeroFlap Optics: Smart connectors that monitor the health of the network in real-time to prevent crashes or slowdowns. Credo's Zero-Flap AECs deliver up to 1,000 times better reliability than traditional laser-based optical modules while consuming roughly half the power — and they achieve this through monitoring health in real time.

Growth shows.. Numbers speak..

$CRDO/USD H2 #Indicator #Signal #Trading #GlobalOneAI

Jefferies upgrades Kratos $KTOS to Buy with an $85 price target, citing a $14B pipeline and a Valkyrie production ramp to 40 units by 2028.

With hypersonic revenue set to double in 2026, Kratos secures a dominant lead in affordable, high growth defense tech.

Morgan Stanley raises Western Digital $WDC price target to $380, maintaining an Overweight rating.

With HDD production sold out through 2026 and shortages projected through CY28, Western Digital secures record pricing power and margin upside.

$WDC/USD H2 #Indicator #Signal #Trading #GlobalOneAI

Red Cat $RCAT partners with HADDY to double its Blue Ops USV production using AI powered robotic 3D printing.

By adopting a microfactory model, Red Cat secures a scalable, on demand manufacturing lead for autonomous maritime defense.

$FLY $RDW $VOYG $RKLB | Firefly Aerospace and Seagate Space signed an MOU for the Gateway S offshore launch platform.

With ABS approval and modular architecture, this partnership secures Firefly’s lead in mobile, responsive space infrastructure for the Alpha rocket.

CrowdStrike $CRWD increased its share buyback program to $1.5B, adding $500M in new capacity.

Citing a "valuation disconnect" and record AI momentum, the move underscores a path toward $20B ARR by FY2036 and enhanced value for long term holders.

$CRWD/USD H2 #Indicator #Signal #Trading #GlobalOneAI

Planet Labs $PL shipped 3 Pelican satellites to Vandenberg for a SpaceX launch, accelerating its lead in high resolution Earth intelligence.

With onboard NVIDIA $NVDA AI and 50cm imagery, the fleet delivers near real time insights for global security.

A new era of on-orbit inspection begins.

#astroscale_JP has unveiled the client satellites and concept of operations for ISSA-J1, the world’s first commercial multi-orbit satellite inspection mission, set to launch in 2027.

$PL: Planet today announced the shipment of three Pelican satellites to Vandenberg Space Force Base in California ahead of their launch aboard the upcoming CAS500/2 rideshare mission with SpaceX.

By launching these satellites to orbit, Planet will further expand the Pelican fleet’s industry-leading capabilities and rapid revisit rate, providing high-resolution, AI-enabled solutions to its customers around the world.

"Kicking off our first Pelican launch of 2026 demonstrates the pace at which we’re scaling our most advanced constellation to date, with higher resolution, faster repeat rates, and lower latency," said Will Marshall, Co-Founder and CEO of Planet.

$ASTS: BlueBird 7 is scheduled to launch on April 14 at 6:45 AM.

MARK MY WORDS: $XNDU is going to revolutionize the entire quantum sector - especially photonics 🔥

This is the first pure-play photonic quantum computing company to hit Nasdaq, and the current price is nowhere near the future market cap it’s heading for. We’re talking massive, multi-billion-dollar upside as they scale fault-tolerant, room-temperature quantum systems that actually deliver real-world utility.

Just look at the partnerships alone:

- Tower Semiconductor powering high-volume silicon photonics manufacturing

- Lockheed Martin on quantum machine learning

- TELUS building sovereign quantum data centers in Canada

- Thorlabs, Applied Materials, AMD, Rolls-Royce, Mitsubishi Chemical, Volkswagen, Toyota Research Institute, and the U.S. Air Force Research Lab

- Plus DARPA advancing them to Stage B of the Quantum Benchmarking Initiative with up to $15M in funding

Xanadu isn’t hype - they already have systems and software in use across defense, aerospace, pharma, semiconductors, and automotive.

Their light-based approach solves the scalability nightmare that’s holding back the whole industry.

$AEHR - the asymmetry here is absolutely massive 🔥

CEO already spelled it out: the AI chip testing TAM is 3-5x larger than the SiC/EV boom that sent Aehr parabolic in 2023.

The market is finally waking up (+100% YTD)…

…but you can still buy the business today at a lower market cap than the old EV hype peak.

Every single high-performance AI GPU and accelerator needs full-wafer burn-in and rigorous testing before it ships. Aehr is the undisputed king of this critical bottleneck - with sticky, high-margin, recurring revenue that scales directly with AI buildout.

This isn’t cyclical. This is structural demand exploding while the valuation hasn’t caught up to the new reality.

Earnings next week could be the spark that sends this thing into orbit.