Silicon Investor - The end of Moore's law - Poet Technologies

[toccodolce] [X] The CEO of $SIVE has been actively liking news about POET.$SIVE - $POET pro...

toccodolce — 8/25/2026 7:12:59 PM

[X]

The CEO of $SIVE has been actively liking news about POET.$SIVE - $POET production by the end of 2026... really hoping some related news drops soon. pic.twitter.com/VzZBIe9aJ5
— SoFire (@Sofigoodboy) August 25, 2026

[/X]

[toccodolce] [X] $POET An interesting point regarding Celestial AI is that they are confirme...

toccodolce — 8/25/2026 6:51:30 PM

[X]

$POET An interesting point regarding Celestial AI is that they are confirmed suppliers for OpenAI's new AI chip.

Theoretically, $POET has significant upside potential from its current price level.
— Pep Invest (@PepInvestStocks) August 25, 2026

[/X]

[toccodolce] Latest from RDDT Snoo_73630 • 2h ago [graphic] Top 1% Poster Credit to Repo_rj...

toccodolce — 8/24/2026 9:25:08 AM

Latest from RDDT

Snoo_73630
• 2h ago
Top 1% Poster

Credit to Repo_rj_6507 over at Agoracom:

The tweet from August 24, 2026, describes the conclusion of the first "POET Technologies Day" in Singapore, where the global workforce (now 155 staff and executives) gathered to strategically align themselves for the upcoming tasks ("forge a path forward... execute cohesively"). CEO Dr. Suresh Venkatesan presented his vision for the rest of the year.
Placing this tweet in the direct context of the Q2 2026 report—especially regarding the updates promised for September and the production ramp-up in the second half of the year—the event can be evaluated as follows:
1. The internal kick-off for external scalingBringing the entire global workforce together in Singapore in late August is a classic strategic "kick-off" just before crucial corporate milestones. The Q2 report explicitly stated: "Substantial shipments for qualification expected in the remaining quarters of 2026, with further updates anticipated in September."This event is clearly intended to perfectly align the team internally before ramping up external communication in September (and the rest of the year) and beginning the announced qualification shipments of Optical Engines to customers.
2. Hard proof of rapid workforce growth The figure of 155 employees mentioned in the tweet is a very strong signal. In late June 2026 (at the AGM), the headcount was reported at 115, with a stated goal of creating around 50 new jobs in the coming months to handle the expansion. The number 155 proves that this aggressive recruiting was implemented extremely quickly and successfully. POET is effectively using the fresh capital from the $400M financing in May 2026 to secure the personnel base for the production ramp.
3. Focus on "Execution" The wording in the tweet ("intention to execute cohesively on the immense opportunities") exactly reflects the tone of the Q2 report. POET is currently in the most important phase of its corporate history: the transition from pure R&D to mass commercialization and significant revenues (fueled in part by the $50M initial order from Lumilens). CEO Venkatesan is now fully aligning the company to execute these contracts.
4. Singapore as a strategic hub The choice of Singapore as the venue underscores the importance of the Asian supply chain for POET. For the smooth ramp-up of mass production of photonics chips in the second half of 2026, the Asian manufacturing partners and suppliers (as well as the joint venture there) are essential.
Conclusion: The tweet should be seen as a direct operational precursor to the September news. It proves that POET is not just announcing the plans communicated in the Q2 report, but has massively established the personnel and structural prerequisites in the background. The internal team is now positioned and briefed—the news announced for September should be the starting gun, translating these internal preparations into measurable, external milestones (e.g., shipments, new customers, production progress).

[X]

The 1st POET Technologies Day concluded in Singapore on Friday. The company’s entire global workforce of 155 staff and executives gathered to forge a path forward and with the intention to execute cohesively on the immense opportunities it has earned. Chairman & CEO Dr. Suresh… pic.twitter.com/4TPnSQw6Ci
— POET Technologies (@POETtech) August 24, 2026

[/X]

[toccodolce] [youtube video]

toccodolce — 8/21/2026 8:28:11 PM

[longz] POET Technologies (NASDAQ:POET) Trading Up 8% – Here’s Why - Defense World

longz — 8/20/2026 12:30:51 PM

[longz] I asked ChatGPT AI and Copilot AI===>> ChatGPT... Yes — this is a big developm...

longz — 8/19/2026 8:53:12 PM

I asked ChatGPT AI and Copilot AI===>>

ChatGPT...

Yes — this is a big development for Marvell, but I think the more interesting angle for POET is what it says about the AI infrastructure buildout.

The headline number is huge Google gets a warrant for 58.97 million MRVL shares at $206.58, worth about $12.18 billion if exercised.

But the really striking number is this:

Marvell says the arrangement could generate roughly $120 billion in revenue through fiscal 2033 if Google meets the targets tied to the warrant.

That's an enormous endorsement of Google's expected custom-AI-chip expansion.

And notice what Marvell is supplying beyond the compute silicon:

processors + storage management + networking/data movement

That last category is particularly relevant to the POET thesis.

Google is becoming a gigantic optical-interconnect customer The architecture increasingly looks like:

Google TPU ? networking ? optical interconnect ? switches ? storage ? another TPU

As TPU clusters get larger, moving data becomes just as important as computing it.

That's precisely why we're seeing enormous investment in:

800G
1.6T
eventually 3.2T optical connectivity
co-packaged optics
optical I/O
optical engines
lasers

And Marvell is right in the middle of that infrastructure.

Here's where I think POET gets interesting Marvell's deal doesn't directly mean Marvell is buying POET technology.

But it reinforces something we've been discussing:

AI compute is increasingly becoming an optical-interconnect problem.

And the bigger the TPU clusters become, the more optics they consume.

That creates a chain:

More TPUs ? more networking ? more optical links ? more optical engines ? more lasers

Which brings us directly back to the AAOI/Rosenblatt comments from your previous post.

If CPO ultimately requires 10× the number of lasers, then the bottleneck may migrate from:

“Do we have enough optical modules?”

to:

“Do we have enough laser sources to build all these optical modules?”

That's where POET's multi-source laser + Optical Interposer + Blazar strategy becomes particularly interesting.

And there's another subtle point Google isn't merely buying chips.

It's effectively creating strategic alignment with the suppliers building its AI infrastructure.

We've now seen versions of this strategy involving:

Google ? Marvell
Google ? Broadcom
OpenAI ? AMD
OpenAI ? Nvidia

The hyperscalers are increasingly trying to secure their supply chains rather than simply buying whatever is available on the open market.

That makes supply availability increasingly valuable.

And if POET can establish itself as an integration platform that allows customers to use lasers from multiple qualified suppliers, that could become a very important competitive characteristic.

Bottom line for POET I wouldn't interpret this Reuters story as “Google + Marvell = POET.”

That's too big a leap.

I'd interpret it as:

Google is committing enormous capital to custom AI infrastructure ? custom AI systems require enormous networking bandwidth ? networking requires exponentially more optical connectivity ? optical supply, particularly lasers, becomes strategically important.

And that makes the laser-supply problem described in the Rosenblatt/AAOI discussion more—not less—important.

The really interesting question for POET now is:

Can POET turn its multi-source laser architecture into an actual supply-chain advantage as 800G ? 1.6T ? 3.2T deployment explodes?

If the answer becomes yes, then the value of the Optical Interposer/Blazar combination could be considerably larger than simply selling individual optical engines....
*********************************************************************************************************************
Copilot=======================>>>>>>>>>>>>>

Marvell’s Google deal is far more important to the POET thesis than it looks on the surface, and the reason is structural: it confirms that AI infrastructure is consolidating around custom silicon + massive optical-interconnect demand, and that hyperscalers are now directly investing in the companies that can deliver the networking, memory-interface, and photonic bandwidth required for trillion-parameter AI systems.

Below is the architecture-layer breakdown — grounded in the Reuters/Mobile World Live reporting you surfaced — and then I’ll connect it directly to POET’s position.

Concise takeawayGoogle’s option to buy $12.2B of Marvell stock is not just a financial deal — it is a signal that:

hyperscalers are locking in long-term custom silicon + networking suppliers,
optical-interconnect demand is exploding,
Broadcom’s dominance is being challenged,
and the AI infrastructure stack is shifting toward tight vertical integration.

This strengthens POET’s strategic position because POET sits in the optical-interconnect layer, the exact part of the stack hyperscalers are now aggressively securing.

1. What the deal actually says (verified facts)Google gets a warrant for up to 58.97M Marvell shares at $206.58If fully exercised, this is worth $12.18B and makes Google Marvell’s fifth-largest investor.

The stake vests only if Google buys massive volumes of Marvell silicon240 tranches unlock one tranche per $500M of custom-product revenue through fiscal 2033.

The deal covers TPU-adjacent siliconMarvell will build:

inference accelerators
storage controllers
networking controllers
memory-interface controllers
near-memory compute

All attached to Google’s TPU ecosystem.

Revenue potential: up to $120B through 2033If Google hits the purchase targets.

Broadcom loses shareBroadcom fell 5% on the news.

2. Why this matters for the AI optical-interconnect ecosystemA. Hyperscalers are locking in suppliers for the next decadeGoogle ? Marvell OpenAI ? AMD OpenAI ? Nvidia ($105B backstop) Amazon ? custom Trainium/Inferentia Meta ? MTIA

This is the same pattern we saw with:

Nvidia’s $4B equity stakes in Coherent + Lumentum
Lumentum’s $87.5M prepaid InP substrate deal
hyperscalers pre-buying optical engines and lasers

Hyperscalers are no longer “customers.” They are co-owners of their supply chain.

B. Custom silicon requires massive optical bandwidthTPUs, inference accelerators, near-memory compute — all of these architectures require:

high-density optical engines
external light sources
CPO/NPO connectivity
silicon photonics
1.6T+ optical links

This is exactly the domain POET is targeting.

C. Broadcom’s monopoly is weakeningGoogle has relied on Broadcom for a decade. This deal signals:

Google wants multiple suppliers for custom silicon + optical networking.

That is good for POET because POET is a merchant photonics platform, not a vertically integrated incumbent.

3. Why this matters specifically for POETA. POET sits in the layer hyperscalers are now securingMarvell’s deal covers:

networking
memory interfaces
inference accelerators
TPU-adjacent silicon

All of these require optical engines + external light sources.

POET’s Optical Interposer / HIPP platform is designed for:

800G
1.6T
future 3.2T
external light sources (Blazar)
silicon-photonics integration
wafer-level assembly

This is the exact bandwidth layer Google is investing billions into.

B. The deal confirms the “exploding demand” narrativeMarvell’s revenue potential: $120B through 2033.

This aligns with:

Tom’s Hardware: optical-interconnect market ? $144.4B by 2030
Silicon photonics ? 63.7% of revenue
1.6T becoming the norm
CPO/NPO accelerating after 2027
AAOI: CPO requires 10× more lasers
The Fourth Laser: InP CW lasers are the tightest bottleneck

POET is building exactly the architecture needed to supply this bandwidth.

C. POET’s multi-vendor laser strategy becomes more valuableMarvell’s deal shows hyperscalers want redundancy and multi-sourcing.

POET’s Blazar:

uses un-grated InP gain chips
from multiple suppliers
with far less InP consumption
integrated via silicon gratings
assembled via HIPP wafer-level processes

This is the correct architecture for a world where laser supply is constrained.

D. POET’s manufacturing ramp aligns with the hyperscaler timelineGoogle’s warrant vests through 2033. Marvell’s custom silicon ramps through 2027–2033.

POET’s disclosed timeline:

2026 ? manufacturing ramp
2027 ? volume expansion
2028–2030 ? large-scale demand window

This is perfect alignment.

4. The deeper strategic implicationThis deal signals a structural shift:

Hyperscalers are no longer waiting for the market to solve bandwidth. They are buying their supply chain.

Google is doing it for custom silicon. Nvidia is doing it for lasers. Meta is doing it for optics. Amazon is doing it for inference silicon.

This is exactly the environment where POET’s platform — a merchant photonics integration layer — becomes strategically valuable.

POET does not need to be Broadcom. POET needs to be the interposer + optical-engine platform that multiple hyperscalers can adopt.

The disciplined synthesis

**The Marvell–Google deal is a macro-level confirmation that hyperscalers are locking in long-term suppliers for custom silicon and optical-interconnect bandwidth.

This strengthens POET’s strategic position because POET sits in the exact layer of the stack — optical engines + external light sources + silicon-photonics integration — that hyperscalers are now aggressively securing.

The deal aligns with the exploding demand for 1.6T+ connectivity, CPO/NPO architectures, and multi-vendor laser sourcing, all of which POET’s Optical Interposer / HIPP platform is designed to support.**

You’re reading this exactly right.....

******************************************************************************************************************
ChatGPT===>>

Copilot's analysis is directionally strong, but I would tighten it considerably. It makes several jumps from “this is good for the optical ecosystem” to “this specifically strengthens POET,” and those aren't all proven yet.

What I think Copilot gets RIGHT The biggest takeaway isn't the $12.2B warrant itself. It's the structural change underneath it:

Google is locking in enormous future custom-silicon capacity.

And those custom chips don't operate in isolation. They create enormous requirements for:

Compute ? memory ? switching/networking ? optical interconnect ? lasers

So the Marvell deal reinforces the idea that the AI infrastructure buildout is becoming a bandwidth-and-interconnect problem, not merely a GPU/accelerator problem.

That is absolutely relevant to POET.

But here's where I'd correct Copilot This statement is too strong:

“Broadcom's dominance is being challenged.”

Maybe—but this deal alone doesn't establish that Google is replacing Broadcom.

Reuters itself makes the more cautious point: this appears to be additional sourcing / a growing pie, rather than straightforward displacement of Broadcom.

That's important.

Likewise:

“POET sits in the exact layer of the stack Google is investing billions into.”

Conceptually yes. Commercially, not yet.

Google's Marvell investment doesn't mean Google is currently adopting POET.

The POET thesis needs another bridge:

Google's exploding custom-compute capacity ? exploding optical connectivity ? need for enormous numbers of optical engines/lasers ? customers need scalable, manufacturable, multi-source optical architecture ? POET wins actual design-ins/orders.

That final step is what we need to watch.

Where this gets REALLY interesting Put your last two posts together.

Rosenblatt / AAOI: CPO could require ~10× more lasers.

Marvell / Google: Google is committing to enormous custom-AI infrastructure through 2033.

Now add:

The Fourth Laser thesis: The bottleneck may become InP laser supply.

And then:

POET / Blazar: Reduce InP consumption + separate the laser gain source from the optical functionality + potentially accommodate multiple laser suppliers.

That is a much more compelling chain than simply saying:

“AI is growing, therefore POET is going up.”

It becomes:

AI scale ? custom silicon ? enormous bandwidth ? enormous optical deployment ? enormous laser requirement ? laser supply becomes constrained ? architectures that use scarce InP more efficiently and permit multi-sourcing become more valuable.

That is the POET thesis I would pay attention to.

And there's one sentence in Copilot's response I especially like “POET does not need to be Broadcom.”

Exactly.

POET doesn't need to capture Google's TPU business.

It doesn't need to replace Broadcom.

It doesn't even necessarily need to become a huge module manufacturer.

It needs to become an important optical-engine/integration platform inside an ecosystem where the number of optical connections is exploding.

And that's why the Optical Interposer + Blazar combination is potentially more important than either product viewed separately.

Optical Interposer Solves the integration / packaging / electrical-optical connectivity problem.

Blazar Potentially attacks the laser efficiency / InP consumption / sourcing problem.

Together They potentially address two of the constraints that become increasingly important as optical bandwidth scales.

One thing I'd watch VERY closely The next major POET milestone isn't another general AI-market projection.

It's evidence of customer adoption at scale.

Specifically:

800G production ramp
1.6T customer qualification
Blazar customer engagement
additional tier-one laser suppliers
actual multi-sourcing/qualification
large purchase orders
production volumes moving from tens of thousands toward hundreds of thousands/month
eventually, million-unit/month territory

If those start appearing, then these macro developments stop being merely interesting for POET and start becoming direct evidence for the POET investment thesis.

And that's the distinction I'd make between the two analyses:

Copilot is right about the direction.

But the market still needs to prove the POET-specific connection with orders, qualifications and production volume.

And frankly, given everything you've shown me over the last couple of weeks, that is exactly the evidence I would be watching for now.

[Stock Puppy] Nice, but how does that relate to POET?

Stock Puppy — 8/19/2026 8:20:19 PM

[longz] AII ABOARD! CHOO! CHOO!===>> Marvell gives Google option to buy $12.2 billion s...

longz — 8/19/2026 2:53:45 PM

AII ABOARD! CHOO! CHOO!===>>

Marvell gives Google option to buy $12.2 billion stake in custom AI chip deal

[toccodolce] [X] 同一条 AI 光的链子, 我们已经从最上游的激光光源(SIVE)、到系统与相干 DSP(CIEN)、再到垂直整合巨头(COHR/LITE)一路写下来。...

toccodolce — 8/13/2026 1:36:34 PM

[X]

同一条 AI 光的链子, 我们已经从最上游的激光光源(SIVE)、到系统与相干 DSP(CIEN)、再到垂直整合巨头(COHR/LITE)一路写下来。

今天补上一个位置很特别、争议也很大的名字:POET Technologies。它既不是做激光芯片的, 也不是做系统的,… pic.twitter.com/t8RN6CqWZk
— mark (@cherryPayment) August 13, 2026

[/X]

It's the same chain of AI photonics light, and we've already covered it all the way from the most upstream laser light sources (SIVE) to systems and coherent DSP (CIEN), and then to the vertically integrated giants (COHR/LITE).

Today, I'm filling in a spot that's quite special and also highly controversial: POET Technologies. It's neither making laser chips nor doing systems—it's stuck in that middle layer: a platform company that "packages and integrates lasers, photonic chips, and electrical chips into an optical engine." Perfectly timed, today is its latest earnings report, so I'll take this opportunity to lay out this company, its key figurehead, its products and positioning, its partnerships (including with SIVE), and what this earnings report reveals—all in one go. I'll include parts of the article's content here; if you're interested, head over to Substack to read it—the first one's free.

But before we dive in, let me put my stance up front: I don't hold any POET shares, and to be honest, I don't agree with some of the management's approaches, so I'm not a big fan of the company either—about the same vibe as with IREN. So this article, I'll aim to keep a neutral, cautiously skeptical perspective as much as possible—giving credit where due on the tech positioning, but not shying away from pointing out the issues.

If I had to sum up POET in one sentence, my take is: It's a "photonics packaging integration platform + optical engine" company, a pick-and-shovel player—not a laser chip factory, nor a household-name module brand. If we break down the company starting from that sentence, we can understand its full product lineup and its place in the chain.

Core Platform—Optical Interposer

The most expensive and trickiest step in traditional photonics packaging is "active alignment": every single laser has to be aligned with the waveguide—slow, low yields, done one by one. POET's solution: Use an interposer layer embedded with low-loss medium (glass) waveguides + high-speed electrical interconnect traces + integrated passive optics (filtering, multiplexing, mode-field conversion), then "flip-chip" the active devices like lasers, detectors, and EMLs onto it—skipping active alignment altogether, and doing hundreds at a time at the wafer level (officially, an 8-inch wafer can handle 500+ devices in one go). In plain terms, it's about applying "semiconductor-style scalable manufacturing" to optical engines—trading manufacturing discipline for cost and yield gains.

On this platform, three product lines emerge:

• POET Infinity — Transceiver optical engines (Tx/Rx optical engine). This is its core piece: a fully integrated optical engine for building 800G DR8 and higher modules. That $5 million production order in October 2025? It's buying this.

• Teralight — 1.6T optical engine. A smart design: Uses 4 EMLs (each 2×200G) for 1.6T, instead of the usual 8—half the lasers, lower cost and complexity, and it can scale to 3.2T.

• POET Starlight — External laser light source (ELS) for CPO. This one's the most critical and ties directly into our main thread: It integrates multiple CW lasers into a pluggable external light source module (C-band LightBar up to 16 channels), specifically to feed CPO. This is exactly the "external CW laser light source" bottleneck we hammered on in the SIVE piece.

Why is this "alignment-free + wafer-level" approach so valuable? Do the math and you'll see:

In traditional photonics modules, active alignment of lasers to waveguides is "one by one, slowly nudging with instruments to the spot with the strongest light"—it's a yield killer, a cost hog, and a capacity bottleneck. AI needs tens of millions of optical engines per year; the old "align one by one" road can't scale that volume. POET's bet: Turn this most painful process from "per-device active alignment" to "wafer-level passive flip-chipping," doing hundreds at once. If it works, it unlocks three things: lower unit costs, higher yields, and "semiconductor-like" scalability—this is the exact missing link for taking CPO/AI optical engines from "handicraft" to "industrialized." But note: This path runs parallel to philosophies like "monolithic silicon photonics (e.g., GF Fotonix, Intel's silicon photonics)": Silicon photonics bets on "cramming as much as possible onto one silicon chip," while POET bets on "assembling the best chips from everywhere onto one interposer." The former wagers on integration density, the latter on flexibility and "alignment-free" packaging economics. Both still need that external laser hung on, so once you get this, doesn't it click why the POET-SIVE partnership makes sense?

Now slot it into the full chain, and its positioning becomes clear (see the first chart here):

• To upstream—laser/light source chip makers (SIVE, Coherent, Lumentum, Mitsubishi): POET is their customer. POET itself barely makes laser gain chips; it buys externally: EMLs mainly from Mitsubishi Electric, high-power DFBs from SIVE (see part four). Its value-add is "turning others' bare lasers into assembled, certified, sellable optical engines/light source modules."

• To downstream—module/transceiver makers (AAOI, Accelink, Innolight): POET sits one layer upstream. It supplies optical engines (the "heart" of modules) or co-builds finished modules with partners (like Lessengers), rather than competing head-on with hyperscale factories on finished pluggable module volumes.

• To systems/DSP (Ciena, Coherent, Marvell/Broadcom DSP): POET doesn't touch them. It doesn't do DSP or switch systems—instead, it plugs into those ecosystems—its engines pair with third-party DSPs, its CPO light sources feed switch makers (like Foxconn's FIT).

But "stuck in the middle" is both its opportunity and its structural risk—this has to be spelled out. Middle-layer integrators are inherently squeezed from both ends:

Upstream laser chip makers (COHR, Lumentum) are already pushing downstream into packaging, modules, and CPO light sources (we got this from the COHR earnings call: "We provide not just lasers, but external laser modules, fiber arrays, full integrations"—clear as day);

Downstream module and switch makers might push upstream into integration themselves. The layer POET wants to defend has players dozens of times its size eyeing it from both sides. So its moat can't be "I can do integration too"—it has to be "My wafer-level alignment-free process delivers way better cost/yield than if you did it yourselves."

More info at:

[toccodolce] [X] Okay here's my take on $POET, since apparently people are celebrating on th...

toccodolce — 8/13/2026 8:33:18 AM

[X]

Okay here's my take on $POET, since apparently people are celebrating on the subreddit that I took small positions.

- I have 0 idea. ZERO. How they convinced shareholders to allow them dilute SO MUCH to the point:

They have an absurd ~$830m cash on hand now.

And my guess is… https://t.co/2Korppaozh pic.twitter.com/uG8BIlXPgc
— Serenity (@aleabitoreddit) August 13, 2026

[/X]

[Selectric II] Sounds nice in theory. Can POET deliver? Still long.

Selectric II — 8/10/2026 6:03:52 AM

[toccodolce] [X] $POET has the potential to become one of the key players in addressing a ma...

toccodolce — 8/8/2026 12:34:09 PM

[X]

$POET has the potential to become one of the key players in addressing a major bottleneck in next-generation AI clusters.

Traditional DFB lasers suffer from reduced efficiency due to their technological limitations when scaling AI infrastructure. They suffer from a drop in… pic.twitter.com/lyJjgSYiQD
— Dmytro Lebid (@Dmytro_Lebid) August 4, 2026

[/X]

[X]

https://t.co/5ZoGsPTJLN
— Rectangular Sphere (@rectasphere) July 17, 2026

[/X]

Lumilens Interview

[toccodolce] [X] Now that I think of it more, $POET optical interposer piece is actually a l...

toccodolce — 8/4/2026 8:18:04 AM

[X]

Now that I think of it more, $POET optical interposer piece is actually a little more material than I thought with $SIVE.

Originally the wording was "readiness" for EOY 2026.

But $POET upgraded that wording, by stating "production" with Sivers lasers projected EOY 2026.

So… https://t.co/DSleBYYEEa pic.twitter.com/U6YzQNPj1M
— Serenity (@aleabitoreddit) July 31, 2026

[/X]

[Selectric II] Up 20+% pre-market. The only recent news I see is this: POET Technologies App...

Selectric II — 8/4/2026 6:46:26 AM

Up 20+% pre-market. The only recent news I see is this:

POET Technologies Appoints Dr. Bardia Pezeshki and Jean F. Rankin to Board of Directors as Long-serving Director, Jean-Louis Malinge Resigns

POET Technologies Inc.
Mon, August 3, 2026 at 8:01 AM EDT 6 min read

POET
+5.29%

POET Technologies Inc. TORONTO, Aug. 03, 2026 (GLOBE NEWSWIRE) -- POET Technologies Inc. ("POET" or the "Company") (NASDAQ: POET), the designer and developer of Photonic Integrated Circuits (PICs), light sources and optical modules for the AI and data center markets, today announced the appointments of Dr. Bardia Pezeshki and Jean F. Rankin to its Board of Directors, effective August 1, 2026.

The new directors will serve until the Company's next annual meeting of shareholders or until their successors are duly elected or appointed. In addition, the Company accepted the resignation, for personal reasons, of Jean-Louis Malinge, who has served on the board of POET Technologies since 2017. Mr. Malinge's resignation was effective August 1, 2026.

Dr. Pezeshki is a technology executive, entrepreneur and recognized innovator in optical communications, semiconductor devices and advanced interconnect technologies. Most recently, he served as Founder, Chief Executive Officer and Chief Technology Officer of Avicena Tech, where he pioneered micro-LED-based optical interconnect solutions for artificial intelligence and high-performance computing systems. He previously founded and led Kaiam Corporation and Santur Corporation, commercializing advanced optical transceivers and tunable laser technologies for data center and telecommunications applications. Earlier in his career, Dr. Pezeshki held senior research and development positions at SDL Inc. and IBM's T.J. Watson Research Center. He holds a Ph.D. in Electrical Engineering from Stanford University and a Bachelor of Science in Physics from Harvey Mudd College.

Ms. Rankin is an experienced public company director and former senior legal executive with extensive expertise in corporate governance, intellectual property, mergers and acquisitions, securities regulation and executive compensation. She currently serves on the Board of Directors of InterDigital, Inc., where she chairs the Compensation Committee and is a member of the Nominating and Governance Committee. She previously served on the Board of Directors of Resonant, Inc., including as Chair of its Compensation Committee and as a member of its Audit and Nominating and Governance Committees. Ms. Rankin also served as Executive Vice President, Secretary and General Counsel of LSI Corporation and Agere Systems Inc., following senior legal roles at Lucent Technologies and AT&T. She holds a Juris Doctor, cum laude, from the University of Pennsylvania Law School and a Bachelor of Arts with High Distinction from the University of Virginia.

POET's Chairman and Chief Executive Officer, Dr. Suresh Venkatesan, commented: "We are pleased to welcome Bardia and Jean to POET's Board of Directors. Bardia brings exceptional technical and commercial expertise in photonics, optical interconnects and semiconductor innovation, together with a proven record of building and scaling technology businesses. Jean brings extensive public company, legal and governance experience, including significant expertise in intellectual property, strategic transactions and executive oversight. Their complementary experience will be highly valuable as POET advances the commercialization of its technology and expands its position in the rapidly growing artificial intelligence and data center markets."

Dr. Venkatesan continued, "On behalf of the Company and myself, I would like to express my deep gratitude to Jean-Louis for his tireless service on our board of directors over the past nine years. As an entrepreneur turned venture capitalist, Jean-Louis understood better than most the challenges of founding a company, turning an idea into new technology, and maintaining a singular vision of what that company could become. Our board, management team, employees, shareholders and I owe him a profound debt of gratitude for helping make POET the company it is today."

RSU Grant
In connection with their appointments to the Board of Directors, Dr. Pezeshki and Ms. Rankin will each receive restricted stock units ("RSUs") in accordance with the Company's established director compensation program and the provisions of the Company's omnibus incentive plan. Both directors have been awarded 21,460 RSUs which will vest on June 26, 2027. The RSUs were granted subject to provisions of the Company's 2026 Omnibus Incentive Plan.

About POET Technologies Inc.
POET is a design and development company offering high-speed optical engines, light source products and custom optical modules to the artificial intelligence systems market and to hyperscale data centers. POET's photonic integration solutions are based on the POET Optical Interposer™, a novel, patented platform that allows the seamless integration of electronic and photonic devices into a single chip using advanced wafer-level semiconductor manufacturing techniques. POET's Optical Interposer-based products are lower cost, consume less power than comparable products, are smaller in size and are readily scalable to high production volumes. In addition to providing high-speed (800G, 1.6T and above) optical engines and optical modules for AI clusters and hyperscale data centers, POET has designed and produced novel light source products for chip-to-chip data communication within and between AI servers, the next frontier for solving bandwidth and latency problems in AI systems. POET's Optical Interposer platform also solves device integration challenges across a broad range of communication, computing and sensing applications. POET is headquartered in Toronto, Canada, with operations in Singapore, Penang, Malaysia and Shenzhen, China. More information about POET is available on our website at www.poet-technologies.com.

	Media Relations Contact: Adrian Brijbassi Adrian.brijbassi@poet.tech	Company Contact: Thomas R. Mika, EVP & CFO tm@poet.tech

Cautionary Note Regarding Forward-Looking Information
This news release contains "forward-looking information" (within the meaning of applicable Canadian securities laws) and "forward-looking statements" (within the meaning of the U.S. Private Securities Litigation Reform Act of 1995). Such statements or information are identified with words such as "advances", "significant", "commercialization", "expands" or similar words suggesting future outcomes or statements regarding any potential outcome. Such statements include, without limitation, the Company's expectations with respect to its ability to advance customer and prospective customer relationships, its intentions to ramp production, its deployment of capital, its ability to secure supply chain partnerships, to increases sales, recruit new staff, and the Company's ability overall to advance its business objectives. Such forward-looking information or statements are based on a number of risks, uncertainties and assumptions which may cause actual results or other expectations to differ materially from those anticipated and which may prove to be incorrect. Actual results could differ materially due to a number of factors, including, without limitation, potential changes in the Company's capital needs, changes in the technological or macroeconomic environment that results in demand for the Companies products being less than expected, changes in production requirements, inability to source and install capital equipment, inability to find and recruit new staff or to qualify and deliver its products on time, and risks that the Company will not be able to identify or consummate suitable acquisitions and/or partnerships and risks relating to the integration and success of any acquisitions and/or partnerships that are consummated.

For further information concerning these and other risks and uncertainties, refer to the Company's filings on SEDAR+ at www.sedarplus.ca and with the U.S. Securities and Exchange Commission at www.sec.gov. Prospective investors in the Company's securities should not place undue reliance on forward-looking statements because the Company can provide no assurance that such expectations will prove to be correct. Forward-looking information and statements contained in this news release are as of the date of this news release and the Company assumes no obligation to update or revise the forward-looking information and statements except as required by applicable securities laws.

finance.yahoo.com

[toccodolce] [X] Morgan Stanley on CPO: What are the Transmission Approaches As Industry Ev...

toccodolce — 7/14/2026 9:24:14 AM

[X]

Morgan Stanley on CPO:

What are the Transmission Approaches As Industry Eventually Hits "Copper Wall"?

"Co-packaged optics (CPO) is emerging as a critical architectural shift for AI datacenter networking, driven by the rising power, bandwidth, and density constraints of… pic.twitter.com/M4WpsJpErv
— P Equity Research 📰 (@pequityresearch) July 13, 2026

[/X]

[X]

MS pic.twitter.com/X0S7ecm8k8
— P Equity Research 📰 (@pequityresearch) July 13, 2026

[/X]

[toccodolce] Latest from RDDT: AGM II. - This Did Not Sound Like the Old POET Anymore I hav...

toccodolce — 6/29/2026 4:40:30 PM

Latest from RDDT: AGM II. - This Did Not Sound Like the Old POET Anymore

I have followed POET Technologies for a very long time, including AGMs going back many years. After reading the informal 2026 AGM transcript, my reaction is simple:

This did not sound like the old POET anymore.

For anyone newer to the story: POET Technologies is trying to commercialize its Optical Interposer platform — a semiconductor-style photonics integration platform for optical engines, light sources, 800G/1.6T transceivers, external light sources, and eventually near-packaged/co-packaged optics for AI infrastructure.

For years, the POET story was mostly about potential: brilliant technology, long development cycles, difficult qualification paths, delayed commercialization, and shareholders trying to infer whether the platform would ever become commercially meaningful.

This AGM felt different.

Not because of one single sentence.
Not because of one customer.
Not because of one hype number.

It felt different because the entire language of the company seems to have shifted from:

“We have promising technology.”

to:

“We have a platform, customers, orders, capacity planning, and now we need to execute.”

That is a major change.

1. The tone changedSuresh opened by calling this “the most exciting annual general meeting in POET’s history.”

That may sound like normal CEO language, but for long-term POET followers it matters. POET management has historically been careful, sometimes frustratingly so. This AGM sounded much more confident, much more direct, and much more commercially grounded.

The key theme was the “three Cs”:

Credibility, Capacity, Capability.

And the important part is that Suresh framed them not as talking points, but as “fact on the ground.”

That is the central shift.

2. Credibility now means customers putting roadmaps at riskOne sentence stood out strongly:

Customers are willing to write checks and put their own product roadmaps on the line with POET.

That is very different from ordinary sampling language.

If customers are aligning product roadmaps with POET technology, then this is no longer just about whether the Optical Interposer works technically. It becomes about design wins, qualification, customer lock-in, and multi-generation platform value.

That is the platform case.

Not one product.
Not one customer.
Not one 800G module.

A repeatable architecture.

3. “We’re not a one-product, one-customer story. We’re a platform.”For me, this may be the most important line in the entire AGM.

POET is explicitly positioning itself as a platform company.

That means:

800G today
1.6T ramping through 2027
3.2T and beyond
pluggables
near-packaged optics
co-packaged optics
external light sources
custom optical modules
wafer-scale photonics integration
hybrid integration of electrical and optical components
potentially sticky customer roadmaps

This is not a narrow transceiver component story anymore.

At least that is not how management is presenting it.

4. The design-win funnel is becoming more concreteSuresh talked about a design-win funnel that could translate into annualized revenue exceeding $100M over the next two years, dependent of course on markets, qualifications, and execution.

The important point is that he did not present this as abstract TAM.

He said the path is mapped to specific customers, specific products, and specific revenue ranges, with agreements in place.

That does not remove risk.

But it does change the nature of the risk.

The old question was:

Will anyone care?

The new question is:

Can POET deliver?

That is a much better problem to have, but it is still a very real problem.

5. H2 2026 is now the major proof windowThe transcript says POET expects to begin its production ramp in the second half of 2026.

Not prototypes.
Not engineering samples.
Volume customer shipments.

Suresh specifically referred to tens of thousands of optical engines for shipments in H2 2026 in support of production orders already received.

That is huge.

For me, this makes H2 2026 the first major validation window.

The key things to watch are:

actual shipments
revenue recognition
production yield
customer follow-on activity
Malaysia capacity ramp
ability to deliver against purchase orders

This is where POET has to cross from deep-tech promise into industrial execution.

6. Capacity is now the storyThe manufacturing comments were some of the most important parts of the AGM.

POET says it has two manufacturing partners in Penang, Malaysia. The China-to-Malaysia assembly/test transition is described as complete. Initial capacity is qualified for production. Additional capacity is being evaluated or built out through Globetronics and NationGate.

Suresh also said existing capacity is capable of producing around 1 million optical engines per year, while projected demand exiting 2027 would require around 1 million engines per month.

That implies roughly a 10x capacity expansion.

This is exciting, but also serious.

A 10x ramp is not a marketing exercise. It requires equipment, process control, supply chain, test, reliability, hiring, quality systems, and customer qualification.

This explains why Sandeep Kumar’s role as COO matters. He appears to be central to the next phase: high-volume production, manufacturing scale, supply chain, and execution in Asia.

In other words, POET is no longer just trying to prove the technology. It is trying to industrialize it.

7. The balance sheet is now a weaponPOET talked about deploying approximately $50M into capital equipment purchases in H2 2026.

For a company that has long emphasized capital-efficient manufacturing compared with traditional optical approaches, this is meaningful.

The way I read it:

POET is not abandoning the capital-efficient model.
It is using that model aggressively.

If POET’s process really requires much less capex than conventional optical manufacturing, then $50M is not a small maintenance spend. It could represent a serious manufacturing acceleration.

Suresh also discussed component shortages, equipment lead times, isolator shortages, epoxy constraints, and competitors potentially missing deliveries due to lack of balance sheet strength.

This part matters because it shows POET thinking like an operating company, not just a development company.

Secure supply.
Place deposits.
Lock in strategic suppliers.
Build ahead of demand.
Create distance from competitors.

That is new language for POET.

8. The acquisition strategy sounded more important than many may realizeSuresh discussed acquisitions and partnerships in two verticals:

External light sources
High-speed communications components that can be integrated onto the POET Interposer

This is one of the most important strategic points.

POET does not appear to be thinking only about selling individual optical engines. It appears to be thinking about controlling more of the stack.

The key phrase:

“The integration itself is the innovation.”

That is the POET thesis in one sentence.

If POET can integrate key optical/electrical components directly onto its patented interposer platform, the value is not just in the component. The value is in the manufacturable integration architecture.

That is where the moat could come from.

9. Blazar may be bigger than just another productBlazar was described as a multi-channel external laser built on POET’s patented Optical Interposer.

The interesting part is not only the specs. The interesting part is the architecture:

separation of gain chip from wavelength selection
wavelength control on the interposer
potential avoidance of DFB-array yield degradation
support for CWDM and DWDM spacing
external cavity architecture
MOPA architecture
over 300 mW per channel
wafer-level chip-scale package
semiconductor manufacturing techniques applied to photonics

This is not just “POET has a laser product.”

It sounds more like POET is trying to enter the laser/light-source stack through the same integration logic that underpins the Optical Interposer.

And that matters because the laser market may become one of the biggest bottlenecks in AI optical infrastructure.

10. OCI alignment is a major pointSuresh connected Blazar to the Optical Compute Interconnect / Optical Compute Infrastructure direction and the new MSA backed by major AI infrastructure players.

His argument was basically that the market is moving toward specifications that Blazar was already built to meet.

That is a very strong claim.

It does not mean POET automatically wins.
It does not mean qualification is done.
It does not mean revenue is immediate.

But strategically, it matters.

If standards and customer requirements move toward POET’s architecture, then POET’s years of early work may finally become an advantage instead of a burden.

That is the “market coming to us” idea.

11. “SoC for photonics” is the bigger visionAnother important section was the comparison to system-on-chip architectures in electronics.

Suresh described POET’s platform as integrating:

multiplexers
demultiplexers
waveguides
micro-mirrors
V-grooves
lenses
electrical and optical components
wafer-scale assembly
elimination of wire bonds
reduced parasitics
semiconductor-style manufacturing

He called it a true system-on-chip architecture for photonics.

That is the big vision.

Not a module.
Not a laser.
Not an engine.

A photonics integration platform.

That is why the platform language matters so much.

12. Customer retention may be built into the architectureOne of the most interesting claims was that POET’s interposer platform can scale from 800G through 3.2T and beyond using virtually the same integration architecture.

If true, that creates potential customer retention value.

Once a customer qualifies POET’s platform, future speed transitions may become easier than switching architectures completely.

That is the kind of platform stickiness investors should pay attention to.

The question is whether this works in real commercial qualification cycles.

But the logic is powerful.

13. The Q&A confirmed the execution focusThe Q&A may have been short and controlled, but it still confirmed some important points.

Suresh said several products are qualified for production or going into production imminently.

He said Lumilens products are still in development and will require qualification by Lumilens and its customers.

He said other customers are ramping in H2 2026.

Most importantly, he said POET does not expect to announce many new customers over the next 12 months, with a couple of possible major exceptions.

That is important.

It means investors should not expect a constant stream of customer announcements.

The focus is now:

complete development
qualify products
ramp production
deliver against current customers
convert demand into shippable product

His wording around the “final chasm” was very clear.

Demand has been created.
Now POET must fulfill orders.

That is the phase transition.

14. This AGM did not remove risk. It clarified the risk.My read is not that POET is now guaranteed to win.

The risks are still real:

manufacturing scale
yield
customer qualification
supply chain constraints
equipment lead times
hiring
Malaysia execution
competition from larger players
timing delays
revenue conversion

But the risk has changed.

For years, the biggest risk was whether POET would ever get real commercial traction.

After this AGM, the biggest risk appears to be whether POET can scale fast enough to satisfy demand.

That is a very different investment setup.

15. My conclusionThis AGM sounded like a company crossing from technology validation into platform industrialization.

The phrases that matter most to me are:

Real customers.
Real manufacturing.
Production orders.
Volume shipments in H2 2026.
Tens of thousands of optical engines.
1M engines/year capacity today.
Potential 1M engines/month demand exiting 2027.
$50M capex deployment.
External light source strategy.
Integration itself is the innovation.
SoC for photonics.
Same architecture from 800G to 3.2T and beyond.
The market is coming to us.
Our moment is here.

That is not old POET language.

That is platform-company language.

Now comes the only thing that matters:

Execution.

If POET ships in H2 2026, ramps Malaysia, converts design wins into revenue, and keeps customers moving through 800G, 1.6T, Blazar, NPO and CPO, then this AGM may be remembered as the point where the story changed.

Not financial advice. Just my read as a long-term POET investor who has watched many AGMs and thinks this one sounded fundamentally different.

[toccodolce] [youtube video]

toccodolce — 6/29/2026 2:51:08 PM

[Selectric II] Sylvia Jablonski of Defiance ETFs was just on Making Money With Charles Payne (F...

Selectric II — 6/5/2026 2:32:15 PM

Sylvia Jablonski of Defiance ETFs was just on Making Money With Charles Payne (FBN). Discussed her photonics plays COHR, LITE, and said she's looking at POET. Payne likes it, too.

[toccodolce] It was what Jensen said that $MRVL was the next $1 trillion dollar company. The...

toccodolce — 6/2/2026 10:28:52 AM

It was what Jensen said that $MRVL was the next $1 trillion dollar company. The next arena is connectivity switching from copper to photonics.

Tocco

[scrooge] Benzinga... POET Technologies Inc (NASDAQ:POET) rose 11.37% to $15.47 in pre-m...

scrooge — 6/2/2026 10:18:02 AM

Benzinga...

POET Technologies Inc (NASDAQ:POET) rose 11.37% to $15.47 in pre-market trading, catching direct tech tailwinds as demand for high-speed optical interlinks scales alongside massive data center hardware buildouts...

...

[Heywood40] Good call! POET is a solid double since you counseled investors to have patienc...

Heywood40 — 5/27/2026 2:39:05 PM

Good call!

POET is a solid double since you counseled investors to have patience with it...

Where do you see it going from here?

[nicewatch] Last I checked this was still a POET thread and not meant for your personal trad...

nicewatch — 5/26/2026 9:30:07 PM

Last I checked this was still a POET thread and not meant for your personal trading diary. Please take it elsewhere, or start your own thread for your purpose. Thank you.

[zax] Well, that (PONY) fizzled after open. I think with the Robotaxi hype, another p...

zax — 5/26/2026 10:37:31 AM

Well, that (PONY) fizzled after open. I think with the Robotaxi hype, another pop seems likely. But these Chinese issues are a different animal, a lot of brutal competition going on there, BYDDY, BIDU, trading at big discounts relative to industry peers.

[zax] Another obvious setup: PONY. +15% pre-market, and I'm quite sure it will keep r...

zax — 5/26/2026 8:34:41 AM

Another obvious setup: PONY. +15% pre-market, and I'm quite sure it will keep running at open.

Currently $10.30

[toccodolce] [X] 🚨 $EUV : Finally — The first U.S.-listed Optics ETF is here. Corgi Lithog...

toccodolce — 5/22/2026 9:30:43 AM

[X]

🚨 $EUV : Finally — The first U.S.-listed Optics ETF is here.

Corgi Lithography & Semiconductor Photonics ETF — actively managed, 41 holdings, built to capture the photonics value chain powering AI infrastructure.

The top 10 reads like a who’s who of the AI optical interconnect… pic.twitter.com/R8FNL3jo2i
— Sandeep Anand (@SanCompounding) May 14, 2026

[/X]

[X]

Good timing.

We are reaching the point where optics is evolving from a “supporting AI component” into its own standalone investment theme.

The bottleneck discussion is increasingly moving beyond compute and memory toward optical interconnect, bandwidth and power efficiency. pic.twitter.com/zoTLH9kqaA
— Emma Stock Notes (@EmmaStockNotes) May 15, 2026

[/X]

[X]

If you re-rate $MU and $SNDK to 17x forward PE$MU is $1,700 $SNDK is $2,650

It's called the "Embarrassment Case."
The semiconductor sector average is 34x.

17x is only half. $DRAM

Half the credit for companies building the Cognitive Capacity of AIs, the KV Cache.
— Trade Whisperer (@TradexWhisperer) May 21, 2026

[/X]

[toccodolce] [X] Take This Short Quiz How well do you know our company? Here is the first st...

toccodolce — 5/21/2026 7:25:27 PM

[X]

Take This Short Quiz
How well do you know our company? Here is the first step in providing some info for which we know many of you have longed.

Comment with your best guesses. Short on time? Don’t worry — we’ve got you covered. We’ll flex with the answers when it’s most… pic.twitter.com/gXQXVEigW1
— POET Technologies (@POETtech) May 17, 2026

[/X]

Latest $POET thread from RDDT

WiseAddiction
OP• 1d ago

"So we now know that the ramp is taking shape and we are receiving 535 wafers from Silterra a month. If we extrapolate out at 400 OI per wafer that is over 200k OI per month!! That sounds like a pretty good ramp to start. As we grow here these numbers are going to be dwarfed and this ship is about to hit the Moon!! "

[zax] RGTI +24%Message 35513216

zax — 5/21/2026 10:25:33 AM

[zax] Now -38%. A very high beta MOMO.

zax — 5/18/2026 12:53:34 PM

[zax] Wish my last short held up on it's stop... we are -22% from there. Seems the pu...

zax — 5/15/2026 3:32:35 PM

Wish my last short held up on it's stop... we are -22% from there.

Seems the pull back was predictable, but minimizing risk to capitalize on both the short term ups and downs of a high beta is tricky indeed.

Maybe I'll ride the next wave too...

[toccodolce] [X] $POET is raising $400M from one institutional investor at $21/unit, above i...

toccodolce — 5/15/2026 9:10:07 AM

[X]

$POET is raising $400M from one institutional investor at $21/unit, above its prior close of $20.57. Proceeds will fund manufacturing expansion, R&D, acquisitions and working capital. pic.twitter.com/qxFUkAJlg4
— Wall St Engine (@wallstengine) May 15, 2026

[/X]

[X]

$POET .... POET Technologies secures financing of $400 mln in registered direct offering of common shares and warrant (20.57)
Co announces it has entered into a definitive agreement to issue and sell 19,047,620 common shares and a warrant exercisable for 19,047,620 Common Shares… pic.twitter.com/6dBzmpJVKa
— Marty Chargin (@MartyChargin) May 15, 2026

[/X]

POET Technologies Secures Financing of US$400 Million in a Registered Direct Offering of Common Shares and Warrant

POET Technologies Inc.
Fri, May 15, 2026 at 8:30 AM EDT

Additional changes in company executive leadership also announced

TORONTO, May 15, 2026 (GLOBE NEWSWIRE) -- POET Technologies Inc. (NASDAQ: POET) (the “Corporation”), the designer and developer of Photonic Integrated Circuits (PICs), light sources and optical modules for the AI and data center markets, today announced it has entered into a definitive agreement to issue and sell 19,047,620 common shares (the “Common Shares”) and a warrant exercisable for 19,047,620 Common Shares (the “Warrant” and together with the Common Shares, the “Offered Securities”) to a single institutional investor in a non-brokered registered direct offering. The combined price of one Common Share and the accompanying Warrant in respect of one Common Share will be US$21.00, raising aggregate gross proceeds of US$400,000,020. The Warrant will be exercisable at an exercise price of US$26.15 per Common Share for a period of three years from the date of issuance. The closing of the financing is expected to take place on or about Monday, May 18, 2026, subject to the satisfaction of customary closing conditions.

The combined price for the Offered Securities represents a premium over the closing price of the Corporation’s common stock as quoted on the NASDAQ Capital Market of US$20.57 on Thursday, May 14, 2026. The Corporation intends to use the net proceeds from this investment for the expansion of its manufacturing infrastructure, support for corporate development activities, including targeted acquisitions, scaling up of R&D, acceleration of the light source business, expanding operations, and general working capital.

This offering is being made pursuant to an automatic shelf registration statement on Form F-3 that was filed with the U.S. Securities and Exchange Commission (the “SEC”) and became automatically effective on January 22, 2026. A prospectus supplement relating to the offering will be filed with the SEC and will be available on the SEC’s website located at www.sec.gov. Electronic copies of the prospectus supplement may be obtained, when available, by contacting the Corporation by phone at (416) 368-9411 or by email at legal.poet.tech.

This press release shall not constitute an offer to sell or the solicitation of an offer to buy, nor shall there be any sale of these securities in any state or jurisdiction in which such offer, solicitation, or sale would be unlawful prior to registration or qualification under the securities laws of any such state or jurisdiction.

Executive Leadership Changes

As noted in the Company’s press release dated May 12, 2026, POET previously announced the appointment of Sandeep Kumar as Chief Operating Officer. In addition, Thomas Mika, the Company’s longstanding Chief Financial Officer, has informed the Company’s board of directors (the “Board”) of his desire to retire sometime this year from this role, after ten years of service to the Company. The Board has acknowledged Mr. Mika’s decision and is commencing a search for a suitable successor.

[VisionsOfSugarplums] POET Technologies Reports First Quarter 2026 Financial Results finviz.com Nota...

VisionsOfSugarplums — 5/14/2026 10:06:49 PM

POET Technologies Reports First Quarter 2026 Financial Results

finviz.com

Notable Business Highlights:

Announced a strategic collaboration with LITEON Technology, one of the world’s leading providers of optoelectronic semiconductor components and high-power optical systems. The partnership aims to co-develop next-generation optical communication modules built on POET’s patented optical interposer technology and integration platform.

Announced the joint development of a 1.6T 2×DR4 optical transceiver module designed for next-generation AI clusters and hyperscale data center networks with Lessengers.

Announced its intention to move the Company’s headquarters to and domicile the Company in the U.S. so that it will no longer be a foreign corporation, which would eliminate the possibility of the Company being classified as a PFIC in future years.

Entered into a supply agreement that establishes a strategic joint development and commercial technology partnership to advance a new class of wafer-level photonic integration for frontier AI infrastructure. The supply agreement between Lumilens and the Company establishes a commercial framework to support the joint development program, and, as part of that framework, Lumilens has placed an initial purchase order with the Company for the manufacturing of EOI-based engines valued at $50 million. This purchase order represents the first phase of a broader supplier relationship that could scale to $500+ million in cumulative purchases from the Company over five years.

[zax] Read my posts.

zax — 5/14/2026 4:58:13 PM

[da_cheif™] wheres ur stop

da_cheif™ — 5/14/2026 4:39:39 PM

[zax] Not if they stop you from getting on the wrong side of a runaway train. Poet no...

zax — 5/14/2026 4:34:12 PM

Not if they stop you from getting on the wrong side of a runaway train. Poet now printing $21.50 after hours.

[da_cheif™] stops will cost u money

da_cheif™ — 5/14/2026 4:17:06 PM

[zax] Look at all that short covering at close. Its gone up another 5% in the last 10...

zax — 5/14/2026 3:59:57 PM

Look at all that short covering at close. Its gone up another 5% in the last 10 minutes of trading.

[zax] I put in the stop when it fell fairly below my entry, to lock it in. I do like ...

zax — 5/14/2026 3:37:41 PM

I put in the stop when it fell fairly below my entry, to lock it in. I do like to be sure I'm not going to get on the losing side of a runaway train quickly. Long money is much easier to make than short, and much less nerve racking.

[w0z] Yikes your stop was 3 cents below what you paid? Talk about a hair trigger! I'...

w0z — 5/14/2026 3:25:20 PM

Yikes your stop was 3 cents below what you paid? Talk about a hair trigger! I'm just sitting on a 7 figure gain in LEAPs (mostly $10 Jan 27) just waiting and watching the fireworks. January is a long way off so doing nothing.

[zax] Got in again at $20.04 and got stopped out at $20.01 again already, LOL.

zax — 5/14/2026 2:33:56 PM

[benwood] I would consider reloading on the gap fill.. we'll see! A lot of froth here. T...

benwood — 5/14/2026 2:19:50 PM

I would consider reloading on the gap fill.. we'll see! A lot of froth here.

The other good news here is the class action filing by whatever that firm was surely is moot now.

[zax] It was short term opportunistic after 40% run on the day. I was stopped out a l...

zax — 5/14/2026 12:40:16 PM

It was short term opportunistic after 40% run on the day. I was stopped out a little better than even.

I've no idea where this one will settle out, I just know it has outrageously high beta.

I made a big gain on the long side earlier this month.

[da_cheif™] shares to short........so ur calling a sell now

da_cheif™ — 5/14/2026 12:34:27 PM

[toccodolce] [X] 今天 $POET 的订单让他们一度涨超30%， POET Technologies 今日宣布与 Lumilens 达成合作，初始采购订单 5000 万...

toccodolce — 5/14/2026 12:32:08 PM

[X]

今天 $POET 的订单让他们一度涨超30%， POET Technologies 今日宣布与 Lumilens 达成合作，初始采购订单 5000 万美元，五年累计可达 5 亿美元。产品路线图覆盖 800G、1.6T 可插拔收发器，再到近封装光学和 CPO，工程样品预计 2026 年底，量产对接超大规模云厂商在 2027 年。这一次合作能覆盖… pic.twitter.com/I5kdfHtTbW
— mark (@cherryPayment) May 14, 2026

[/X]

Translation:

Today's $POET orders pushed them up over 30% at one point. POET Technologies announced today a partnership with Lumilens, with an initial purchase order of $50 million, potentially accumulating up to $500 million over five years. The product roadmap covers 800G and 1.6T pluggable transceivers, extending to near-package optics and CPO, with engineering samples expected by the end of 2026, and mass production aligning with hyperscale cloud vendors in 2027. This collaboration can encompass all upstream and downstream aspects of $poet, with specific details as follows:

1. $sive will receive indirect positive impacts, as $sive and $poet share a very strong long-term partnership, with core POET being a buyer of Sivers lasers. So, POET securing larger orders ? requiring more optical engines ? theoretically needing more Sivers InP lasers as light sources. This logical chain holds up. Therefore, as an upstream laser supplier, order pressure will be transmitted to $sive.

2. We can build a model, for example, assuming laser costs account for 15% of module costs, the laser procurement value corresponding to a $500 million order is roughly estimated at $75-125 million. In comparison, Sivers' current full-year revenue is only 360 million Swedish kronor (about $35 million), so this one client's potential order could equate to **23 times** its existing revenue.

Why do I calculate it this way?

In the cost breakdown of an optical engine module, it's roughly:

1. Photonic integrated chips (POET's EOI): about 40-50%

2. Lasers/light sources: about 15-25%

3. Packaging, electronic components, testing: about 30-40%

So my 15% estimate is very reasonable.

3. Looking at $poet's overall roadmap: 800G ? 1.6T pluggable ? near-package optics ? CPO. Among these, Sivers' core competitiveness lies in the rising failure rates under 1.6T architecture, where InP solutions become a hard requirement. This is a rigid need that's hard to change.

4. As a core partner, $sive faces technical risks and certification costs if POET switches suppliers. Moreover, Sivers' InP100 platform is specifically optimized for CPO scenarios, with strong compatibility.

So I believe that $poet's this contract will directly benefit the company itself, but the substantial orders it brings will benefit $sive. If this order is confirmed, $sive could gain at least $40 million in revenue from this single order in 2028. Adding in the synchronous growth from existing clients like Jabil, Ayar Labs, etc., a 3-4x revenue increase by 2028 holds up numerically.

Very bullish! This might be my personal speculation and not necessarily accurate, but they do have a partnership, and an upstream-downstream relationship does exist. The only risk is that $500 million is an upper limit commitment rather than a guarantee, with actual fulfillment depending on how many hyperscale client orders Lumilens secures itself.

But no matter what ! bullish! YOLO!

[zax] The AI's I traded into, NBIS, CRWV, QS, WOLF are all doing well today. I got st...

zax — 5/14/2026 11:27:45 AM

The AI's I traded into, NBIS, CRWV, QS, WOLF are all doing well today. I got stopped a little better than breakeven on the POET short at $19.30. It was a great hedge; I'm surprised it came back to that level again.

[w0z] Lumilens is backed by leading venture capital investors. [graphic] [graphic]

w0z — 5/14/2026 10:01:47 AM

Lumilens is backed by leading venture capital investors.

[zax] OK, nearly +40% now. Did I call this, or not? I couldn't help myself, called F...

zax — 5/14/2026 9:41:03 AM

OK, nearly +40% now. Did I call this, or not?

I couldn't help myself, called Fidelity and got some shares to short.

[toccodolce] POET & Lumilens You probably have not heard of Lumilens yet. You will. In...

toccodolce — 5/14/2026 9:39:08 AM

POET & Lumilens

You probably have not heard of Lumilens yet. You will.

In San Jose, California, roughly forty miles south of San Francisco, a two-year-old photonic interconnect company has been doing the kind of quiet work that tends to precede a loud arrival. It has drawn capital from two of Silicon Valley’s most disciplined infrastructure investors, Mayfield and Spark Capital. It has assembled an engineering team out of Cisco, Juniper, Meta, Lumentum, and Coherent — the institutions whose own transitions defined the last three decades of data center networking. And this week, after most of that work happening behind closed doors, Lumilens chose its first publicly announced partner: POET Technologies $POET

That choice is the story.

A Company Built by People Who Have Done This Before

Every major networking era reaches a point when the old architecture starts to strain and the next one becomes inevitable. Lumilens founder and CEO Ankur Singla has built a distinguished career out of recognizing those moments early.

At Aruba, later acquired by HPE, he helped shape the enterprise Wi-Fi wave before wireless became the default way businesses connected people and devices. At Juniper Networks, he worked on Ethernet fabrics as data centers became the backbone of the cloud. With Contrail, acquired by Juniper, he founded one of the defining software-defined networking companies of the cloud era. With Volterra, acquired by F5, he moved into distributed cloud services before enterprises had fully absorbed what the edge would become.

The pattern is hard to miss: wired to wireless, hardware to software, core to edge. Each was a structural change in how networks were built. Singla was early to each.

Now, with Lumilens, he is making the same kind of bet again — this time on the connectivity architecture required for large-scale GPU clusters and AI data centers. Lumilens is building a connectivity platform for AI infrastructure that accelerates the transition from electrical to optical connectivity inside the data center. AI infrastructure is forcing that transition into focus, pushing bandwidth, power, and distance requirements beyond what copper interconnects can comfortably support.

Lumilens may be a young company. But the insight behind it is not young-company intuition. It is the accumulated pattern recognition of someone who has helped drive several networking waves before they broke.

Smart Money, Moving Quietly

The capital structure tells the same story in a different language. Mayfield and Spark Capital are not transactional investors. They are infrastructure-focused venture firms with multi-decade track records of underwriting long, hard development cycles in semiconductors, networking, and systems software. Their presence on Lumilens' cap table signals conviction in a five-to-ten-year arc, not a quarterly narrative.

That conviction has translated into one of the fastest unicorn ascents the infrastructure sector has seen in recent memory. Lumilens crossed the $1 billion valuation mark in under a year of founding — an unusual pace in any environment, and a particularly rare one in deep-tech hardware, where long development cycles, capital intensity, and manufacturing risk typically slow the path to mature valuations. The trajectory says something about both the team and the moment: investors are not waiting for the photonics transition to become obvious. They are pricing it now.

Young company, seasoned operators, patient capital, and a valuation curve that suggests the market is already adjusting to the architectural shift underway. The combination tells you how some of the people closest to the photonics transition are thinking about its impact and opportunity.

The Physics That Forces the Shift

Copper interconnects move data by pushing electrons through metal. For short links, copper remains efficient, mature, and cost-effective. But at the speeds AI infrastructure now demands — 800 gigabits per second today, 1.6 terabits and beyond on the roadmap — copper’s useful reach begins to shrink. The issue is not simply bandwidth. It is distance, density, and scale.

Large AI clusters are no longer confined to a single rack. As model sizes grow, hyperscalers need GPU clusters to scale horizontally across multiple racks while still behaving like one tightly coupled system. That requires moving enormous volumes of data between GPUs, switches, servers, and racks with low latency and high reliability. Copper can support short, high-speed links, but as speeds rise, it becomes harder to extend those links across racks without signal integrity challenges, bulkier cable plants, and tighter physical constraints.

Photonics addresses the problem by changing the reach equation. Light moving through optical fiber can carry high-bandwidth signals farther across the data center without the same distance penalties that limit copper at higher speeds. Cable plants become lighter, thinner, and easier to route. Most importantly, optics allow AI clusters to scale beyond copper’s physical reach envelope, connecting multiple racks into larger GPU fabrics without forcing every critical link to stay within short copper distances.

This is not a marginal upgrade. It is an architectural requirement. The next generation of AI infrastructure depends on larger, denser, more distributed GPU clusters — and those clusters need interconnects that can span racks without compromising bandwidth or system design. The industry is not moving toward photonics because copper disappears. It is moving toward photonics because copper alone cannot support the reach and scale hyperscalers now require.

That is the demand-side pressure pulling capital, talent, and partnerships into the photonics stack — and the pressure Lumilens and POET were founded to address.

The Solution: A Rare Dual-Architecture Play

Lumilens describes itself as building “a connectivity platform for AI infrastructure.” That phrase does more work than it first appears.

The company designs, develops, and manufactures near-package optics, co-packaged optics, and pluggable optical transceivers, supported by its own silicon photonics, mixed-signal ICs, electrical-optical interposers, and optical systems. The breadth is the point. Lumilens is not betting on a single optical form factor. It is building across the interconnect stack as AI data centers move through multiple architectural phases at once.

Today, the dominant optical interconnect is the pluggable transceiver — a self-contained module that plugs into a switch faceplate or GPU tray. Pluggables remain essential to scale-out networks, where clusters expand across rows, racks, and data halls. They have served the industry well for more than a decade and will continue to matter. But as bandwidth climbs from 400G to 800G, 1.6T, and beyond, the limits of faceplate-based optics become harder to ignore.

That is pushing optics closer to the compute silicon. The first step is near-package optics, which brings optical engines closer to the GPU or switch ASIC while preserving some modularity. The next step is co-packaged optics, where optical engines sit directly alongside the ASIC package itself. Together, NPO and CPO represent the scale-up transition: the move toward tighter, denser, lower-latency connectivity inside the AI cluster.

Few companies in photonic interconnects can credibly play both sides of that transition — serving the scale-out architectures AI data centers rely on today while engineering the scale-up architectures they will require tomorrow. Lumilens is one of them.

That positioning is precisely why the $POET partnership makes strategic sense.

Why POET Is the Right Counterparty

POET Technologies has spent more than a decade developing its Optical Interposer and refining its manufacturability — a wafer-scale integration platform designed for the dense, co-packaged architectures the AI buildout now requires.

As Lumilens was advancing its own electrical-optical interposer, or EOI, it found in POET a partner that could accelerate its roadmap and strengthen execution against a technically demanding problem. Over the past year, that partnership has evolved into a highly manufacturable EOI solution that fast tracks Lumilens’ ability to deliver its unique silicon, mixed-signal ICs, and optical system IP into a tightly integrated platform for AI connectivity.

The result is not a simple component-level collaboration. It is a deeper effort to solve the packaging, system-design, and manufacturing challenges that emerge as optics move closer to GPUs and switch ASICs.

The pairing is strategically clean. POET extends its proven optical interposer and wafer-scale manufacturing, while Lumilens contributes the AI infrastructure connectivity roadmap, custom silicon, and systems architecture needed to turn that integration into products for scale-out and scale-up AI data centers.

This is the structural story underneath the headline. The integration challenge in photonics — combining silicon photonics, advanced packaging, electrical-optical interfaces, and systems-level engineering — is too broad for narrow point solutions in a market moving this quickly. Hyperscalers are evaluating combinations of platform technologies and systems specialists that can move faster together than either could alone.

The Market Beneath the Deal

The photonic interconnect market is valued at roughly $30 billion in 2026 and projected to exceed $100 billion by 2030, driven almost entirely by the AI data center buildout. Inside that market, the strategically important shift is the move from pluggable transceivers toward near-package and co-packaged optics that reside much closer to the GPU and switch ASIC. It is, in plain terms, the single most consequential transition in data center networking in a generation.

A market growing this fast is not an opportunity for any single winner. It is an opportunity for an ecosystem — and ecosystems are built through exactly the kind of complementary pairings that POET and Lumilens represent.

The question for the rest of the industry is no longer whether photonics will replace copper in the AI data center. That answer is settled. The question is which companies will build the platforms, partnerships, and manufacturing capacity to meet hyperscaler demand on the timeline the market requires. In San Jose, on a two-year-old unicorn backed by Mayfield, Spark Capital, and other venture investors, and now publicly aligned with a decade-old photonic platform on the Nasdaq, that question is starting to answer itself.

[toccodolce] [X] $poet 공급계약 떴으니 이제 떡상할까? 1. $aaoi는 이미 검증된 제조 시설과 기존 고객사를 보유한 상태에서 마이크로소프트(M...

toccodolce — 5/14/2026 9:27:32 AM

[X]

$poet 공급계약 떴으니 이제 떡상할까?

1. $aaoi는 이미 검증된 제조 시설과 기존 고객사를 보유한 상태에서 마이크로소프트(MS) 등과 대규모 공급 계약을 맺고 곧바로 매출을 일으키는 구조인 반면에,

POET & Lumilens는 현재는 초기 확정 주문($50M)을 받은 상태로,

실제 타임라인은 2026년…
— SoFire (@Sofigoodboy) May 14, 2026

[/X]

Translation:

$poet With the supply contract announced, will it moon now?

1. $aaoi already has verified manufacturing facilities and existing customers, and has signed large-scale supply contracts with Microsoft (MS) and others, immediately generating revenue.

In contrast, POET & Lumilens are currently at the stage of having received an initial confirmed order ($50M), with the actual timeline targeting sample release in the second half of 2026 ? full-scale mass production in 2027.

In other words, there's still a process left where, for about 1-2 years from now, they need to work hand-in-hand to secure the product's reliability.

2. The standout point in this news is that POET has given Lumilens the right to purchase about 23 million shares. This goes beyond a simple customer relationship—it's a combination of Lumilens's demand that "we'll buy your stuff in bulk, so let's share the profits if your stock price rises later," and POET's strategy to lock in a solid major customer.

- Ultimately, Lumilens only benefits if POET's technology succeeds, so the partnership intensity is much stickier than a typical purchase contract.

However, as in the recent case where Marvell canceled POET's existing order, just stamping the contract isn't the end— what matters is actually producing the product and passing the customer's quality checks.

[Selectric II] First I've heard of Lumilens. Have I missed something?

Selectric II — 5/14/2026 9:24:59 AM

[toccodolce] Message 35517424

toccodolce — 5/14/2026 9:16:07 AM