The full-stack research platform for public equity investors. https://xitter.cf/t.co/0xRZegn2ep [email protected]
Joined November 2012
- Tweets2.2K
- Following1.2K
- Followers37.3K
- Likes1.6K
FUNDA retweeted
About 120-140 people attended our intimate @FundaAI happy hour. I was blown away by some of the feedback we received from clients and attendees. A special thanks to Funda for making this possible and to all attendees, and apologies if I wasn't able to chat with all of y'all.
FUNDA retweeted
You can stream a webinar. You can't stream walking into an arena with the entire industry in it.
November 17–18 at the Moody Center in Austin.
Register at the link in the replies.
FUNDA retweeted
FT’s ~$50B OpenAI run rate is a September snapshot. @FundaAI has it at $48.8B, close to that headline.
Dots launched Sept 29. @thsottiaux now reports 40M active users across Codex + Work.
BEP’s hypothesis: dots, new models and tiers, plus coding and enterprise adoption, could drive $100B by year-end. That requires ~26% monthly growth on the same revenue basis. Ambitious, not confirmed.
As @altcap argued at All-In, downstream revenue must justify upstream capex.
FUNDA retweeted
OAI (and Ant) need to explain their ARR's.
Rightly or wrongly, the entire market assumed OAI's ARR was $70B as per Axios' earlier reporting and accounting magic.
Regardless of being private companies, the stakes are way too high -- as seen with the subsequent market sell-off after FT's report.
Letting people come to their own conclusions and assumptions is a dangerous game to play when you directly impact dozens of publicly listed companies.
Also note - @FundaAI have OAI's ARR at $48.4B...
FUNDA retweeted
Funda is the GOAT when it comes to optics, so I take what they say very seriously. Bullish $LITE. $20B+ OCS TAM by 2030 is crazy.
Deep|Optics: OCS Content per Accelerator Grows with Each Generation as Demand Widens Beyond Google
A year ago, optical circuit switches were mostly a Google program with one outside supplier shipping under $10m a quarter. Lumentum now expects its September-quarter OCS revenue to exceed $100m, and we believe its OCS revenue will pass $2.5b next year. TPU 8i put OCS ports into inference pods for the first time and the 6D torus would give the next training chip four times as many OCS ports as today.
Demand for OCS is widening beyond Google. We expect Nvidia to adopt OCS once their scale-up domains pass eight racks, from late 2028. We see more than $20b a year of OCS demand by 2030. Lumentum leads merchant OCS shipments today, and Coherent is the other US-listed supplier. Google remains the only hyperscaler confirmed to buy OCS and nothing at Nvidia is in production yet.
An OCS connects one fiber to another with a mirror, a liquid crystal cell or a chip and leaves the light untouched. With no transceiver and no DSP inside it, the same switch works at 800G, 1.6T or 3.2T and uses about 80% less power than a packet switch on Lumentum’s figures. Google has run OCS in their datacenter network for a decade and in their TPU pods since TPU v4. Until last year they built almost all of their switches themselves. Below is an illustration from Lumentum of how an OCS can fit within a datacenter.
A TPU in Google’s 3D torus has 1.5 OCS ports. Google’s inference pods had none until TPU 8i, which links 36 groups of chips through OCS at about 1.25 ports per chip. The 6D torus reported for the next training chip would take the count to six ports per chip from 2027, or three if two transceivers share a port. Nvidia has no OCS in production. Direct fiber links tie up to eight racks together. Beyond eight racks, our model gives each GPU four ports, $800–1,200 of content at $200–300 a port.
Some of this is already in reported numbers. Lumentum targets $400m of OCS for the second half of 2026 on a backlog from three customers. Coherent doubled its addressable market estimate to $4b by 2030 between February and March. Nvidia took part in iPronics’ $125m round on September 2.
We attended ECOC in Málaga in late September and met with teams from Oriole Networks, Salience Labs, iPronics, LightDance and others. OCS was one of the topics discussed during the week. This note updates the OCS market outlook and presents both the bull case and the pushback we heard from the industry.
The companies discussed in this report include:
1. Lumentum
2. Coherent
3. Huber+Suhner
4. Google
5. Nvidia
6. Broadcom
7. Marvell
8. Astera Labs
9. Arista Networks
10. Ligent
11. nEye (private)
12. iPronics (private)
13. Salience Labs (private)
14. LightDance (private)
15. Oriole Networks (private)
Detailed Report
fundaai.substack.com/p/deepo…
FUNDA retweeted
Last November @FundaAI called $LITE "the name that benefits the most from the TPU." It closed at $253.81 that day. 🧐 It closed at $1,111.07 on Wednesday, up 338%! 🚀
Now they say Lumentum's OCS revenue passes $2.5B next year!
Deep|Optics: OCS Content per Accelerator Grows with Each Generation as Demand Widens Beyond Google
A year ago, optical circuit switches were mostly a Google program with one outside supplier shipping under $10m a quarter. Lumentum now expects its September-quarter OCS revenue to exceed $100m, and we believe its OCS revenue will pass $2.5b next year. TPU 8i put OCS ports into inference pods for the first time and the 6D torus would give the next training chip four times as many OCS ports as today.
Demand for OCS is widening beyond Google. We expect Nvidia to adopt OCS once their scale-up domains pass eight racks, from late 2028. We see more than $20b a year of OCS demand by 2030. Lumentum leads merchant OCS shipments today, and Coherent is the other US-listed supplier. Google remains the only hyperscaler confirmed to buy OCS and nothing at Nvidia is in production yet.
An OCS connects one fiber to another with a mirror, a liquid crystal cell or a chip and leaves the light untouched. With no transceiver and no DSP inside it, the same switch works at 800G, 1.6T or 3.2T and uses about 80% less power than a packet switch on Lumentum’s figures. Google has run OCS in their datacenter network for a decade and in their TPU pods since TPU v4. Until last year they built almost all of their switches themselves. Below is an illustration from Lumentum of how an OCS can fit within a datacenter.
A TPU in Google’s 3D torus has 1.5 OCS ports. Google’s inference pods had none until TPU 8i, which links 36 groups of chips through OCS at about 1.25 ports per chip. The 6D torus reported for the next training chip would take the count to six ports per chip from 2027, or three if two transceivers share a port. Nvidia has no OCS in production. Direct fiber links tie up to eight racks together. Beyond eight racks, our model gives each GPU four ports, $800–1,200 of content at $200–300 a port.
Some of this is already in reported numbers. Lumentum targets $400m of OCS for the second half of 2026 on a backlog from three customers. Coherent doubled its addressable market estimate to $4b by 2030 between February and March. Nvidia took part in iPronics’ $125m round on September 2.
We attended ECOC in Málaga in late September and met with teams from Oriole Networks, Salience Labs, iPronics, LightDance and others. OCS was one of the topics discussed during the week. This note updates the OCS market outlook and presents both the bull case and the pushback we heard from the industry.
The companies discussed in this report include:
1. Lumentum
2. Coherent
3. Huber+Suhner
4. Google
5. Nvidia
6. Broadcom
7. Marvell
8. Astera Labs
9. Arista Networks
10. Ligent
11. nEye (private)
12. iPronics (private)
13. Salience Labs (private)
14. LightDance (private)
15. Oriole Networks (private)
Detailed Report
fundaai.substack.com/p/deepo…
FUNDA retweeted
去年 11 月 ，FundaAI 说光通信里最好的票是 LITE。
当时股价 253.8 美元，现在 1000+ ，涨了 330%。
这次他说 OCS 明年过 25 亿美元🤔
Deep|Optics: OCS Content per Accelerator Grows with Each Generation as Demand Widens Beyond Google
A year ago, optical circuit switches were mostly a Google program with one outside supplier shipping under $10m a quarter. Lumentum now expects its September-quarter OCS revenue to exceed $100m, and we believe its OCS revenue will pass $2.5b next year. TPU 8i put OCS ports into inference pods for the first time and the 6D torus would give the next training chip four times as many OCS ports as today.
Demand for OCS is widening beyond Google. We expect Nvidia to adopt OCS once their scale-up domains pass eight racks, from late 2028. We see more than $20b a year of OCS demand by 2030. Lumentum leads merchant OCS shipments today, and Coherent is the other US-listed supplier. Google remains the only hyperscaler confirmed to buy OCS and nothing at Nvidia is in production yet.
An OCS connects one fiber to another with a mirror, a liquid crystal cell or a chip and leaves the light untouched. With no transceiver and no DSP inside it, the same switch works at 800G, 1.6T or 3.2T and uses about 80% less power than a packet switch on Lumentum’s figures. Google has run OCS in their datacenter network for a decade and in their TPU pods since TPU v4. Until last year they built almost all of their switches themselves. Below is an illustration from Lumentum of how an OCS can fit within a datacenter.
A TPU in Google’s 3D torus has 1.5 OCS ports. Google’s inference pods had none until TPU 8i, which links 36 groups of chips through OCS at about 1.25 ports per chip. The 6D torus reported for the next training chip would take the count to six ports per chip from 2027, or three if two transceivers share a port. Nvidia has no OCS in production. Direct fiber links tie up to eight racks together. Beyond eight racks, our model gives each GPU four ports, $800–1,200 of content at $200–300 a port.
Some of this is already in reported numbers. Lumentum targets $400m of OCS for the second half of 2026 on a backlog from three customers. Coherent doubled its addressable market estimate to $4b by 2030 between February and March. Nvidia took part in iPronics’ $125m round on September 2.
We attended ECOC in Málaga in late September and met with teams from Oriole Networks, Salience Labs, iPronics, LightDance and others. OCS was one of the topics discussed during the week. This note updates the OCS market outlook and presents both the bull case and the pushback we heard from the industry.
The companies discussed in this report include:
1. Lumentum
2. Coherent
3. Huber+Suhner
4. Google
5. Nvidia
6. Broadcom
7. Marvell
8. Astera Labs
9. Arista Networks
10. Ligent
11. nEye (private)
12. iPronics (private)
13. Salience Labs (private)
14. LightDance (private)
15. Oriole Networks (private)
Detailed Report
fundaai.substack.com/p/deepo…
FUNDA retweeted
@FundaAI is one of three that passed a recent audit of newsletters. The author says Funda's strongest signal comes with the fourth bullish mention of a stock - conviction building across successive research notes is a reason to follow the research over time.
quantdata.uk/research/financ…
FUNDA retweeted
On our weekly @FundaAI client call, we discussed what the microLED and microVCSEL demonstrations at ECOC mean for optical links moving into applications currently served by copper.
We also covered their potential role in scale-in architectures and the industry’s shift toward “wide and slow,” using more lanes at lower speeds. This was a recurring topic in major plenary sessions at ECOC. Avicena’s successful microLED demo and ams OSRAM’s microVCSEL demo showed progress on both technologies. For Credo, we are watching its expected ALC demo at OCP, which it has renamed “active light cable.” When we spoke with the team at ECOC, they did not confirm whether the demonstration would use microLEDs or microVCSELs. Seeing it in action should give us a better sense of Credo’s progress and where the product could fit into customers’ systems.
Client questions focused heavily on Lumentum versus Coherent, specifically the linewidth and phase-noise debate over Coherent’s 400 mW UHP CW laser. We still see Lumentum ahead, with customer feedback on Coherent’s UHP laser remaining limited.
We have not heard confirmation of successful qualification nor reports of specific problems with its laser. We also do not yet have enough visibility to assess whether Coherent can meet its target for December-quarter volume shipments.
fundaai.substack.com/p/resea…
FUNDA retweeted
Recommended reading for the weekend:
1. @a16z — State of Markets II
2. @GoldmanSachs — AI Is Becoming a Stock Picker’s Market
3. @FundaAI — Optics: FCC Transceiver Rules Might Be Better Than Feared, and Better for Top Chinese Players Than You'd Think
4. @abcampbell — There Are Bonds in the Singularity
5. @federalreserve — An Update on AI and the Economy
Have a great weekend everyone!
FUNDA retweeted
A joint deep dive with Ozeco of Crack The Market is out, and the whole piece is free. We set out to map who can actually fabricate the PIC, the silicon photonics chip that sits inside every AI optical link.
When we finished counting, the list of fabs shipping datacom PICs in volume had six names on it. Underneath them, one company supplies nearly all of the 300mm photonics substrate they build on.
What the piece covers:
- Each of the six, by platform, wafer size and the customers designing into it
- Why customers are already handing the merchant fabs cash, or being asked to, before the new capacity is built
- The 300mm expansions that all land around 2028, and what that could do to wafer pricing
- Why we see Nvidia's TSMC-based CPO and Tower/NewPhotonics NPO as complements today and rivals once lanes reach 400G
My side was the foundries and the packaging. Seen from packaging, CPO moves at the speed of what comes after the fab.
In TSMC's flow that means a hybrid bond nobody can redo, with the driver chip fused onto the photonic chip. Then every engine needs an optical check before it goes anywhere.
If the base-case arithmetic holds and the 2028 risk is surplus rather than shortage, who keeps the pricing power: the foundry, the substrate, or the bonding line?
FUNDA retweeted
I asked Claude yesterday on @FundaAI post about models not exposing full chain of thoughts as plaintext.
CPU + GPU up today too.
FUNDA retweeted
I invite you all to read the whole paragraph and the full article for a complete overview.
Credo will adopt both uLEDs and uVCSELs and have re-named their cables to Active Light Cables (ALC) which is more of an AOC now.
Avicena showed at ECOC a successful uLED demo.
From @FundaAI latest. Credit to Credo for admitting to reality and pivoting quickly.
uVCSEL >> uLED in every way. uLED has fundamental, unfixable issues. Blowing up fiber attach count for ultra slow lanes only to need to burn power on FEC anyway.
FUNDA retweeted
From @FundaAI latest. Credit to Credo for admitting to reality and pivoting quickly.
uVCSEL >> uLED in every way. uLED has fundamental, unfixable issues. Blowing up fiber attach count for ultra slow lanes only to need to burn power on FEC anyway.
FUNDA retweeted
We had a happy Happy Hour in Malaga during ECOC. It was a small but nice crowd. Lots of ideas bounced between the attendees. Looking forward to our next Happy Hour at OCP in San Jose!
FUNDA retweeted
Maybe a prick and not a puncture in equilibrium. type of curve that creates a generation of new startups
"GPT-6 Luna is the first U.S. model since DeepSeek launched its V4 series to undercut DeepSeek’s cheapest model on both input and output list prices."
Deep|LLM: Tiered Model Pricing Is Broadening AI Adoption; Limited Impact by Jev
Pricing cuts accelerate adoption: Anthropic launched Claude Opus 5.5 at $4/$20 per million tokens, 20% below Opus 5, and estimates typical workload costs are down 40%. OpenAI’s GPT-6 Sol at $2/$10 and GPT-6 Luna at $0.10/$0.50 are 50–58% below comparable GPT-5.6 prices, making more everyday AI tasks economical to automate.
Tiered models expand use cases: Frontier models such as GPT-6 Astra and Claude Opus 5.5 still command premium pricing for complex tasks, while lower-cost models like Luna support extraction, routing, and consumer-agent workflows. The same application budget can now support more users, more model calls, or longer agent workflows.
Jev is useful but narrow: TypeSafe’s Jev offers structured decisions at $0.042 per million input tokens with free output, averaging $0.0004 per call and 0.4-second latency in TypeSafe’s evaluation. It can replace some classification and guardrail calls, but it does not perform long-chain reasoning, so its impact on frontier-model demand should be limited.
Compute demand remains supported: Over the next 6–12 months, broader AI deployment is expected to increase both usage and compute demand, even as inference efficiency improves. Application companies are the clearest beneficiaries, while model developers that compete mainly on low API prices face greater margin pressure, especially standalone Chinese players.
Detailed Report
fundaai.substack.com/publish…
FUNDA retweeted
$SMTC flirting with ATHs once again - up 50% in a month lol. Not sure why, but @FundaAI just reported that 1.6T transceiver shipments are growing faster than expected. For context, more than 50% of 1.6T modules use Semtech's TIAs. Laser shortage should also benefit them if they can ramp up Hiefo fast enough.
@FundaAI also reports that DSP shortage is a huge constraint. Good news $CRDO, I'm sure.
FUNDA retweeted
More on AI pacing:
Sarah Friar CFO of OpenAI yesterday on CNBC: “from where I sit today there is so much opportunity to drive growth that I am still highly focused on getting more compute to keep that flywheel going.”
Sachin Katti VP of Compute Strategy at OpenAI @sk7037 yesterday: “The way we will make sure frontier models are safe and aligned is by spending compute. So the counterintuitive point is that we’ll need even more compute to make sure future models are more safe and aligned.”
If you thought “pacing” was negative for AI infrastructure demand, think again. Almost as bullish as open-weight AI taking share but not quite.
FUNDA retweeted
"Dario’s essay is, in part, an admission that safety engineering at AI labs has fallen behind model capabilities." - @FundaAI