Serenity

From $AMD(AMD.US) Advancing AI, channel checks:

Helios networking partners told me they expect AMD to pursue CPO for future generations of Helios scale up networking.

And expects other hyperscalers + industry to go down the CPO route.

One AI infra executive at a company involved in a hyperscaler CPO programs said they had seen no indication of CPO scale up supply chain delays.

Pushing back on recent 3rd party delay reports.

This was informal commentary, not official AMD or partner guidance.

$Intel(INTC.US) and $AMD(AMD.US) to sign CPU LTAs with Chinese customers for AI DCs (Reuters).

- Prices of some CPU products have risen more than 40% in China since the start of the year from sources.

- Month-on-month increases topping 10% for some products

CPUs were already a bottleneck, following CPU ratios due to AI inference...

But the broader trend of LTAs seems to be appearing from:

- Memory, with $Micron Tech(MU.US), Samsung, $Sandisk(SNDK.US), and SK Hynix signing DRAM/NAND LTAs.

- Photonics, with $Lumentum(LITE.US), $Coherent Corp.(COHR.US) signing EML LTAs. And recent Trendforce reports that $AMD(AMD.US) and hyperscalers are now pursuing CW LTAs.

And I'm sure there's many more from MLCCs to all the way to substrates.

+1 for the bottleneck investors... hard to be a "bubble that pops" if you have take or pay demand spanning multiple years.

Just some notes from $Tesla(TSLA.US) transcripts:

Elon Musk: "I think Optimus will be the biggest product ever" from Q2 earnings transcripts.

Even from Q1, Elon Stated: "I think Optimus will be our biggest product. Not just Tesla's biggest product ever, but probably the biggest product ever.

And I remain convinced of that conclusion".

It's very rare to see such high conviction reiterated from Elon regarding humanoids both directionally and TAM-wise.

In terms of upstream suppliers opportunities, Karn Budhiraj, VP of $Tesla(TSLA.US) supply chains stated:

"We’re seeing the same level of investment going to memory and also new specific items like metal injection molded parts, flexible printed circuits, and all sorts of nonlinear technologies..."

And that they had to build new supply chains from scratch. So would be interesting to look at for new supply chain supplier opportunities apart from the known ones.

Just pulled older slides and it looks like capacity targets for Fremont are 1M and Giga Texas was 10M, so at scale, $Tesla(TSLA.US) seems to be the clear US leader. If you're looking at that S-Curve opportunity ramp.

On a side note, Elon also gave a shoutout to $Micron Tech(MU.US) with likely reasonably priced LTAs, given his statement:

"We really appreciate Micron making room for Tesla in the years to come and giving us actually a very significant allocation on reasonable terms given the pretty insane pricing of memory these days." (Just for structural memory demand).

Elon also gave a shoutout to Panasonic which had invested many billions in increasing battery cell production (will need to do deeper DD into Panasonic after this), then Samsung and $Taiwan Semiconductor(TSM.US).

Just takeaways from Tesla earnings TLDR:

Elon tends to be directionally correct on where the future is heading, such as with EV or Space.

He seems extremely bullish on humanoids.

$SIVE CEO buys 70K shares today.

I don’t really comment much on insider buying/selling, since it’s not exactly a fundamental catalyst.

But given the company’s future is dependent on the CEO’s vision + execution, from NASDAQ listing, M&A, to scaling:

It’s extremely positive to have Vickram be completely aligned with the shareholders

If you’re curious why $Nebius(NBIS.US) is up after hours.

$NVIDIA(NVDA.US) disclosed it owns 9.3% beneficial ownership of Nebius via SEC filings.

Not exactly too new since it dates back to Nvidia’s existing share position of 1.19M shares + $2B prefunded warrant of ~21M shares.

But great for sentiment having Nvidia be a large shareholder of the Neocloud leader.

$IREN(IREN.US) +19.69% after revising ARR target to $4B+ from new AI cloud contracts. Its customer base now includes Microsoft, Nvidia, Perplexity, and Figure.

$Hut 8 Mining(HUT.US) +10.45% off a 15Y $9.8B AI DC lease

$Cipher Digital(CIFR.US) +16.76%

$CleanSpark(CLSK.US) +13.7%

$Whitefiber(WYFI.US) +9.18%

$Nebius(NBIS.US) and $Coreweave(CRWV.US) are taking their time in the shower.

But does look like Neoclouds/Colo players are sharply recovering following new catalysts, such as Kimi compute shortage news and new contracts.

Some updates on bottleneck timelines:

- ABF substrates: 1 year

/ eg. Ajinomoto (2802) for ABF film, Ibiden, Unimicron, Nan Ya PCB, etc make the substrates

- HDI Boards: Over 6 months

/ Victory Giant (300476), Zhen Ding (4958), Unimicron, Compeq (2313), Meiko (6787)

- Multilayer boards: ~6 months

/ Victory Giant, WUS (002463), $TTM Tech(TTMI.US), Gold Circuit Electronics (2368),

- CCL: Over 6 months (severe shortages of materials)

/ Elite Material (2383), Shengyi (600183), TUC (6274), ITEQ (6213), Nan Ya Plastics (1303), Resonac (4004)

Mitsui Kinzoku (5706) / HVLP copper foil. Nittobo / glass fiber/cloth

- MLCCs, chip resistors, tantalum capacitors, aluminum electrolytic capacitors: 15-20 weeks to 1 year+

MLCC: Murata (6981), Samsung Electro-Mechanics, TDK (6762), Taiyo Yuden (6976), Yageo (2327), Walsin (2492)

Chip Resistors: Yageo, Walsin, KOA (6999), Rohm (6963) and $Vishay Intertech(VSH.US)

Tantalum capacitors: Kyocera/AVX, Yageo/KEMET, $Vishay Intertech(VSH.US) again

Aluminum electrolytic capacitors: Nippon Chemi-Con (6997) Nichicon (6996) | Timelines sourced from Digitimes, did the honors of adding in related companies.

But it's slightly more nuanced since MLCCs and others mix. in commodity types vs. AI DCs.

Regardless, just some helpful mapping.

$Meta Platforms(META.US) in talks to lease compute to Anthropic in a $10B dollar deal.

Seems like they saw how profitable

$SpaceX(SPCX.US) $45B compute deal was.

But just goes to validate Neocloud business models like $Nebius(NBIS.US), $IREN(IREN.US), and co if hyperscalers are copying their homework.

Source: NYT

$Apple(AAPL.US) looks to acquire AI chip companies for running AI (Source: The Information)

Right now, $Microsoft(MSFT.US), $Meta Platforms(META.US), Amazon, Google are carrying AI capex spend.

But what if Apple joined the others after M&A?

A possible scenario is that they revise capex largely upward for their own AI buildout.

Since they probably witnessed Google cutting off Meta from compute constraints... or what happens when you partner with OpenAI for LLMs.

Then learned how important it is to have your own infrastructure.

This scenario would be quite bullish thematically from optical networking to foundries and something markets would not expect? We'll see what happens.

$Nebius(NBIS.US) signs $1B+ compute agreement with Reflection AI, for GB300 access through 2029.

Reflection also signed a multi-billion dollar agreement with $SpaceX(SPCX.US) earlier.

Interesting to say the least, seeing Nebius drop -5% off the news today.

Also... counterparty to get this done kinda reminds me of OpenAI, where they might not have the funds to actually execute on these LTAs yet compared to $Meta Platforms(META.US) or $Microsoft(MSFT.US).

But generally positive long term developments, customer diversification was one of the core strengths of Nebius.

Morgan Stanley note on CPO today. Key participants include:

- $SIVE, $Coherent Corp.(COHR.US), and $Lumentum(LITE.US) in laser supply

- Broadcom and Nvidia in switch platforms

- Lightmatter, Ayar Labs, Marvell/Celestial, and POET in optical engines and photonics,

- $Taiwan Semiconductor(TSM.US), $GlobalFoundries(GFS.US), and $Tower Semicon(TSEM.US) in silicon photonics foundry capacity

I’ve covered all of these before, like Tower Semi.

But I’m especially happy that Morgan Stanley validated my research that $SIVE is one of the critical global players in CPO.

A small $1.5B laser company next to your two leading $60B+ companies…

If you also synthesize Rosenblatt Securities recent note that China laser suppliers were quite far from having CPO lasers.

This kinda magnifies importance of the three Western leaders of that laser chokepoint.

Just a semi recap TLDR:

- $Corning(GLW.US) glass bridge per Morgan Stanley has potential, but hard to displace FAU (like FOCI) in short term

- $SpaceX(SPCX.US) Starlink Gen 3 is scaling to 100,000 units (10x prev gen) creating possible capacity constraints for suppliers of switches to CCL.

- PCB supply shortages are projected to persist until 2028, and component shortages/price hikes are already forcing ODMs like Inventec to issue conservative H2 shipment

- DeepSeek and Zhipu are developing custom ASICs to bypass $NVIDIA(NVDA.US) (kinda expected by now).

- Anthropic has achieved a $30B ARR and is projected to hit >$1B in Q3 profit. Turns out these frontier labs are more profitable than people think.

- US admin pressured $Apple(AAPL.US) to source from $Intel(INTC.US), in exchange for tariff exception. Kinda pressuring alignment away from TSM.

- $Taiwan Semiconductor(TSM.US) plans a 30x expansion of its Photonic Integrated Circuit (PIC) capacity by 2028, growing from 500 to 25,000 wafers per month

- Hanmi Semiconductor is entering the CoWoS packaging equipment market

- $NVIDIA(NVDA.US) and NTT hosting a conference July 24th to discuss CPO strategies.

- 5x general NAND flash price hike this year, market size $489B by next year triggered by RAG/inference. Samsung and SK Hynix make emergency fab investments, NAND equipment suppliers go brrrr.

- Gas turbine bottleneck, 40% supply gap accompanied by staggering 5-year delivery cycles (mega-fabs need them for mass production).

- AI probe card assembly has hit a severe bottleneck apparently. Which is attempted to by solved with things like Innovation Service's machines.

- Nanya, 79.5% gross margins reported, memory go brr. 4x capex for capacity/advanced packaging.

- CXMT's STAR Market IPO on July 16, so should bring a lot of attention to memory players in that supply chain.

- KYEC $1.4B investment into US for $Taiwan Semiconductor(TSM.US) Arizona output test facilities. Lot of these players like $Amkor Tech(AMKR.US) and others should go brr in 2028.

- $Intel(INTC.US) CEO warned last month that helium could hinder the manufacturing costs and delivery times of AI chips

- 3D NAND word lines are shifting from Tungsten to Molybdenum starting at the 375-layer node

- SambaNova secured JPM for AI inference and raised $1B at an $11B valuation.

- HBM prices projections to double in 2027 as $NVIDIA(NVDA.US) Rubin platform drives demand.

- $Micron Tech(MU.US) provides $500 million in funding to support GlobalWafers' US manufacturing capacity

- $Meta Platforms(META.US) 'Iris' will enter mass production in September via $Broadcom(AVGO.US) and TSMC and Meta aims to double computing power to 14GW by 2027.

- Largan Precision has secured its first CPO FAU order, mass production slated for middle of next year (kinda indication around Foci and others).

- Samsung and SK Hynix have delayed the implementation of hybrid bonding packaging technology for HBM4 apparently

- Memory costs (DRAM/NAND) have reached 60% of the BOM for sub-$400 smartphones, causing a severe volume contraction.

- SK Hynix successfully rasied $26.5 billion through a Nasdaq ADR

- $Tesla(TSLA.US) issued procurement guidelines requiring suppliers to reach weekly production of 1,000 units by September and double to 2,000-2,500 by year end for its 3rd gen Optimus robot. Apparently Alliance Technology and A-Link may be in this supply chain harmonic reducers and vision lenses?

Kinda go through all this stuff every day, but don't wanna be a news reporter so just consolidated stuff I found interesting.

Apparently there’s a new popular trend to bearpost trillion dollar companies.

Like $NVIDIA(NVDA.US) or $Taiwan Semiconductor(TSM.US).

By finding isolating technical details, then blowing it out of proportions by claiming their whole program is delayed.

-> if the company doesn’t respond, stock prices/sentiment are damaged.

-> if the company issues a response, they can’t go into detail since supply chains are confidential… but people still doubt them

-> Poster gain more impressions, then does it again with more views the next time.

Don’t quite think this is very healthy, with more folks copying that playbook. Especially around incentive structures.

Regarding $Amazon(AMZN.US) RNG + OpenAI’s MRC and VCSELs.

In what appears to be an indirect response on B. Riley's $Applied Optoelectronics(AAOI.US) sell report regarding optimizations that "flattens networks and cuts transceiver TAM 40-50%".

Rosenblatt TLDR: They've already modeled for this, even though each XPU may use fewer transceivers.

The number of XPUs is growing so quickly that total optical demand should still rise.

On lasers architectures for CPO: VCSELs are useful as a credible bridge for NPO/short reach.

Ideal architecture though for CPO is UHP CW over other lasers, and cites Broadcom for that statements. Which happens to supports my core $SIVE thesis.

I feel like Rosenblatt and I share the same views on a lot of things, I like the way they roll.

Mitsubishi Motors joins robotics race to deploy humanoid workers by 2027.

Pairing with Highlanders, a Japanese robotics company.

- Agility $Churchill Capital XI(CCXI.US) x Foxconn, Toyota

- Apptronik x $Jabil(JBL.US), Mercedes

- Highlanders x Mitsubishi

- Figure × BMW

- Boston Dynamics x Hyundai (subsidiaries)

- Optimus x Tesla (Internal)

- Rainbow Robotics x Samsung (subsidiaries)

Idk if it's just me of there's a lot of patterns with humanoid companies playing buddy-buddy up with a major auto player/manufacturer.

If they're not already vertically integrated by one already.

Then the Chinese players like UBTech/AgiBot/Unitree don't seem to care and work with a wide variety of auto players in China.

Curious how this trend plays out...

Just Bloomberg and $Meta Platforms(META.US) doing damage control after crashing the market with Meta Compute framing:

Spokesperson: "Meta is still hungry for even more computing power.

It is still moving forward with plans for expensive new data centers and recently inked major computing deals with $CRVW, Google, $Oracle(ORCL.US), and others."

Just dropped that in with the Meta Muse announcement, and evenn threw in the "expensive" framing with DCs to signal capex.

But little late given we're likely seeing a lot of margin liquidation cascades and heavy losses from media framing earlier.

Just putting it out there:

If everything crashes together from $Nebius(NBIS.US), $Marvell Tech(MRVL.US), $Intel(INTC.US), $Sandisk(SNDK.US), $AMD(AMD.US), $SIVE, $Micron Tech(MU.US), $Lumentum(LITE.US), and others...

Which are all down -4% to -10%+ today so far.

Probably doesn't have anything to do with individual fundamentals.

Indiscriminate selloffs from things like cascading margin liquidations, usually provide compelling opportunities if the underlying improves.

Was a sad few months for $Reddit(RDDT.US). But glad it's finally back above $200.

Reddit was doing:

- $663M revenue w/ 91.5% gross margins

- $204M GAAP net income

- 45%+ fwd Y/Y growth after 69%+ growth.

- Net profit is ~30.7% of revenue

Felt very weird to see a profitable company get dragged down after earnings.

But in hindsight, given the increase hyperscaler capex, the drop felt more like it's more relative opportunity cost more than fundamentals?

Since a lot of inflow poured into $Micron Tech(MU.US) / Sk Hynix that are bottlenecked into 2029. Or with laser bottlenecks like $Lumentum(LITE.US) that last into 2029 as well.

Same thing happens in reverse after selloffs ig, even if fundamentals didn't change.

Been seeing a lot of noise with memory optimization or Chinese players... but don't quite think you sign 16T+ LTAs if memory was getting flooded or not used anytime soon.

Same with optical players... if in earnings, they state anything they make gets bought, I don't quite think a lot of the noise has material effect on their fundamentals.

Regardless, nice to see some recovery with familiar faces like $Robinhood(HOOD.US), $Hims & Hers Health(HIMS.US), and $Reddit(RDDT.US) though.

Kelper upgrades $XFAB to "Buy" citing:

- Melexis demand strengthening “week-after-week, day-after-day” (their lead customer).

- AI giving XFAB a structural growth leg with SiC, GaN, and photonics offsetting auto cyclicality.

- Scarce Western speciality foundries like $SILEX commanding higher multiples.

"X-Fab remains the cheapest specialty foundry peer on every multiple, trading at 4.8x 2027 EV/EBITDA versus a peer median of 14.5x"

I treat this as good signal for the core fundamentals improving like auto players.

But in terms of unpriced long term growth vectors, I'm personally looking at XFAB from a different angle in terms of photonics growth with photonixfab. Given $NVIDIA(NVDA.US) as a potential key driver.

But it's hard for traditional firms to price these types of active developments into valuation models until there's volume contracts.

Nomura’s report cites $AXT(AXTI.US) and $IQE as the leading players.

Then regarding ASP updates:

- 2in InP substrate price hike 42-76%

- 3in InP substrates price hike 78%

- 2in EML Epiwafer price hike 50-75%

- 3in CW Epiwafer price hike ~40%+

- Prices lower domestically in China as expected.

Hyperscalers bottlenecking upstream photonics confirmation validation… on all fronts from pricing, demand, to supply shortages.

My personal expectation was that price hikes would keep going up like $Sandisk(SNDK.US) given ramping optical demand.

We’ll see if this turns out right given another 2 years of the demand curve.

Came across an interesting report from SVRC Research called "State of Robotics 2026", published in April.

Which listed:

1. Figure AI

2. Agility Robotics $Churchill Capital XI(CCXI.US)

3. Apptronik

4. $Tesla(TSLA.US)

5. Boston Dynamics

6. Physical Intelligence

7. 1X Technologies

8. $Amazon(AMZN.US) Robotics

9. Covariant

10. Skild AI

As the National Champions of the United States robotics program.

"The United States leads the world in where robotics is heading: Fundation models, OpenAI-style scaling laws applied to action, autonomous vehicles.

While losing the race on where robotics is shipping today."

Then it frames:

1. Rare Earths Exposure: from Neodymium for motors to samarium-cobalt for high-temp applications as a critical vulnerability.

2. Actuator dependency. Series elastic actuators, quasi-direct-drive motors, and precision reducers overwhelmingly sourced from Japan, Germany, and China

As one of the main vulnerabilities alongside Manufacturing velocity/data collection cost/regulations. Then their take was:

"With at least six well-funded US humanoid companies competing for a market still in early formation, we expect at least two significant consolidation events (acquisition or merger) in 2027".

With Logistics / E-commerce (like $Amazon(AMZN.US) / $FedEx(FDX.US)) and Automotive from $General Motors(GM.US) to $Forward Industries(FORD.US) as being the immediate top use cases for deployment.

I think it's just interesting to see a lot of my points I've been talking about reiterated by research firms. Regardless, I do think it's going to be a major frontier race between the US and China.

Agility Robotics (which I own), Tesla, Figure, and Apptronik as leaders representing the USA. Competing against Unitree, AGIbot, Ubtech, and others in China.

Just some consolidated updates on memory:

- $Micron Tech(MU.US) leads new 1.5T Yen investment in Hiroshima ~$9.3B. (bullish read through for Disco, Advantest, Resonac, Towa) since capex is localized.

- Morgan Stanley pointed out NAND will continue to be in short supply into 2027 so $Sandisk(SNDK.US) / Kioxia type players are happy alongisde $Silicon Motion Tech(SIMO.US) and upstream.

- MS remains especially positive on Macronix/Winbond

- UBS expects the average price of DDR contracts in the Q3 2026 to increase 32% | 18% Q4, vs. 17% and 12% est.

- UBS expects NAND flash to be raised 30% from prev quarter.

- Samsung reportedly plans 20% DRAM hike Q3. TrendForce recently forecast DRAM contract prices to rise 13 to 18 percent in the third quarter from the previous quarter, so this hike beats expectations.

Something to note:

Lot people see 20%... and don't think it's a lot compared to the 70-80% from previous quarters.

But if you hike something by 100%, then hike something by 100%, then hike something by 30%, it's a lot more than people estimate since it's compounded. Similar to tracking inflation.

I've already made projections going into 2028 from the start of the year on my memory names...

I'm just sitting back and watching things play out through all the "memory optimization" and "they can't keep price hiking like this!" noise.

Leaked Anthropic docs show plans to secure 1.4GW capacity from Australia, amounting to ~$21.6B.

Recently $IREN(IREN.US), $SharonAI(SHAZ.US), and other Neoclouds have been building sovereign DCs in Australia.

Guess like a lot of things are stating to make more sense connection wise?

(disclosure: no open positions in any of the above).

Nikkei - $JD.com(JD.US) E-Commerce Giant CEO warns "Robots Will Replace 700,000 Delivery Workers".

Per FT's earlier piece last month: The 700,000 delivery personnel working for the firm will be replaced by robots, highlighting how rapid automation has become a new threat to China's already severe job market.

Liu Qiangdong, revealed that the company has signed contracts with approximately 120 schools to retrain its delivery workers, enabling them to transition into new roles such as robot repair and maintenance...

"Beijing has started to track AI’s hit to jobs as a national priority."

This follows other corporate strategy reports like $Amazon(AMZN.US), of avoiding 600,000 future hires with robots.

But it's very interesting to see the shift in worker roles to robot support. Maybe $Doordash(DASH.US) to $Uber Tech(UBER.US) to $Mercadolibre(MELI.US), we'll all start to see this rollout following China.

There's a lot of retail disbelief over humanoid/robotics commercialization. But feels like the direction industry is heading down is pretty clear.

Always a fun time seeing semi markets crash from Bloomberg report framing once again…

UBS analyst on $Meta Platforms(META.US) report: “this is not new news”.

Regardless:

$Aehr Test(AEHR.US) down -18.3%

$Applied Optoelectronics(AAOI.US) down -17.13%

$SIVERS SEMICONDUCTORS AB(SIVEF.US) down -15.4%

$Sandisk(SNDK.US) down -14.8%

$Teradyne(TER.US) down -13.8%

$Corning(GLW.US) down -11.4%

$Marvell Tech(MRVL.US) down -11.2%

$Lumentum(LITE.US) down -10.2%

$Nebius(NBIS.US) down -7.8%

Getting PTSD from Bloombert’s previous misleading report on AMD/Nvidia global export controls earlier this year that crashed semi stocks.

FYI: Meta doesn’t randomly buy $48B+ worth of neocoud contracts if they overbuilt capacity and can cut capex…

Markets are stupid sometimes.