Direxion Daily AVGO Bull 2X Shares

Zhongji Innolight's Interim Report covering optics:

> Silicon photonics penetration in the global datacom optical interconnect market is expected to rise further, from about 38% in 2025 to about 73% in 2030.

> Optical interconnects accounted for about 5% of global AI capex in 2025, and are expected to reach about 10% over the long term. The global datacom optical interconnect market is projected to grow from $19.4 billion in 2025 to $98.6 billion in 2030, a compound annual growth rate of 33.8% from 2026 to 2030.

Regarding CPO:

> CPO is described as currently viewed as one of the most highly integrated optical packaging solutions, assembling the switch ASIC chip and optical engine together on the same substrate.

> The report states CPO is still in the development stage, facing challenges around technology and ecosystem maturity, manufacturing consistency, cost, and maintainability.

> Over the long term, CPO will coexist with pluggable optical modules and other connectivity solutions (rather than fully replacing them).

$Lumentum(LITE.US) $Coherent Corp.(COHR.US) $Applied Optoelectronics(AAOI.US) $Broadcom(AVGO.US) $NVIDIA(NVDA.US)

Goldman Sachs: Optical Networking Supply Chain

Too detailed to fit into just one pic...

$NVIDIA(NVDA.US) $Tower Semicon(TSEM.US) $Broadcom(AVGO.US) $Marvell Tech(MRVL.US) $Lumentum(LITE.US) $Coherent Corp.(COHR.US) $AXT(AXTI.US)

SCHWAB CLIENTS IN JULY WERE NET BUYERS BY A 2:1 RATIO IN ETFS AND EQUITIES.

Top stocks bought:

- $SpaceX(SPCX.US)

- $Micron Tech(MU.US)

- $Intel(INTC.US)

- $Oracle(ORCL.US)

- $Tesla(TSLA.US)

Top stocks sold:

- $Apple(AAPL.US)

- $Broadcom(AVGO.US)

- $Adobe(ADBE.US)

- $Paypal(PYPL.US)

- $AMD(AMD.US)

Source: amit

$NVIDIA(NVDA.US)

Okay, so let’s talk about what Nvidia actually announced today…

I’ll present the bear and the bull case. I lean more bullish for reasons that are below, but I want to highlight both sides and what this means for the broader AI industry.

So, what did Jensen do today that required 6 of the most important bankers to sit with him live on national television?

Well, he just secured $500B of revenue for Nvidia.

The entire press release is centered around financing supply for NVIDIA compute with a “rich ecosystem of offtakers built upon NVIDIA’S CUDA platform.” The announcement literally ends by saying “this will create dedicated pools of capital at significant scale at attractive rates for NVIDIA customers.”

This move is not to increase supply for the demand that Broadcom, AMD, Google TPUs, etc. are seeing. This move may help all those companies find ways for more buyers to potentially finance their products as more money enters the ecosystem, but this is explicitly a move that Jensen is making to secure $500B of demand that will flood into Nvidia. It’s a very, very smart move.

How do I envision it playing out? Let’s say a neocloud needs $5B of capital to buy Nvidia GPUs. Now, Nvidia does $100B+ of FCF, they obviously could invest in that neocloud or just give that neocloud a loan to buy their chips. But, that ends up putting some strain on their balance sheet. Nvidia has an excellent balance sheet, just increased their dividend by 25X, and has committed $80B to buybacks.

So, instead of Nvidia taking on the financial risk, they essentially vet the neocloud and tell Goldman Sachs that this company deserves a $5B loan. Goldman trusts Jensen, they give the neocloud the money, the neocloud then spends it on Nvidia chips, eventually the ROI from that compute goes back to paying Goldman with interest, which is the entire business of a bank, lending capital to customers with an attractive risk profile to make a return on that capital.

The combined banks Nvidia is working with have 25T in assets. They are committing $500B of those assets for this project.

Okay…the bull and bear cases.

Bull:

- Obviously, great for Nvidia has they remove all financial risk but get financial institutions to make a loan that directly will go back to Nvidia in the form of chip sales.

- Should be good for memory, CPUs, optics, packaging, photonics, all parts of the semi stack. You can’t just buy Nvidia chips, you need to buy everything that goes into it and so all the semi names likely see their earnings continued to be sustained with hyperscaler capex and now a fresh 500B entering into the ecosystem. The market did not take those names up afterhours on this deal which likely means the market is still trying to figure out how bullish this is and if it’s sustainable.

- Compute as an asset class is the main thing that was discussed today which is what the financial institutions want: they would like to build futures/derivatives products on compute like they do for oil, crypto, real estate mortgages, etc. and create a new financial asset class that allows liquidity and fees to emerge. This is, in my opinion, the number one reason the banks are interested: the more compute, the more financial applications to speculate on that compute while trusting that Jensen knows how to allocate this 500B to continue progressing AI globally.

Bear:

- It’s circular financing. If no ROI on all the companies these banks put money into, even with Jensen’s approval, the loans don’t get paid back and if Nvidia has to promise to backstop those loans in someway, it could create a major risk to their balance sheet.

- Nvidia picks the wrong labs or clouds to invest in. Nvidia has been pretty good at allocating capital, but if they end up giving money to companies that will buy their chips but ultimately can’t execute (like having a ton of GPUs but not strong enough to set up datacenters quickly) then there will be questions around why those companies got the money which could lead the banks to not want to give out more loans, these are just MOUs, not confirmed agreements to begin allocating capital.

- Too much leverage in the system. If we do start to see financial derivatives of compute as an asset class, then there may be too many people speculating and with that speculation comes for the potential to have bad underwriting and essentially create too many products around compute that don’t actually provide any value which eventually collapses.

Ultimately, I lean on the bullish side because I believe there is an ROI coming from this compute. If there wasn’t, it’s just money switching hands back and forth with no real value. Ultimately, the market will have to see additional innovations in AI beyond just agentic inference which effectively has defined 2026 to see demand continuing to grow but also believe that the real value to GDP and productivity shows up in earnings growth (which we are seeing this year with S&P earnings) to make sense of this $500B entering into the ecosystem. If the ROI continues along with earnings growth, I would imagine the market feels like the fresh capital is going to only create more innovation and more growth which would be rewarded.

What do you think, bullish or bearish on this announcement?

Source: amit

📊 𝐈𝐍𝐕𝐄𝐒𝐓𝐎𝐑 𝐍𝐎𝐓𝐄: Cantor Says Semiconductor Selloff Is "Overdone," Reiterates AI Bullish View

$Micron Tech(MU.US) $Sandisk(SNDK.US) $Western Digital(WDC.US) $Seagate Tech(STX.US) $NVIDIA(NVDA.US) $Broadcom(AVGO.US) $AMD(AMD.US) $ASML(ASML.US) $Applied Materials(AMAT.US) $Lam Research(LRCX.US) $KLA(KLAC.US) $Teradyne(TER.US) $Lumentum(LITE.US) $Coreweave(CRWV.US) $Super Micro Computer(SMCI.US), $Coherent Corp.(COHR.US) $Cerebras(CBRS.US)

👉 𝗞𝗲𝘆 𝗛𝗶𝗴𝗵𝗹𝗶𝗴𝗵𝘁𝘀:

➤ 𝗖𝗮𝗻𝘁𝗼𝗿 says the recent semiconductor selloff is 𝗼𝘃𝗲𝗿𝗱𝗼𝗻𝗲.

➤ 𝗦𝗢𝗫 fell about 𝟯𝟬% peak-to-trough versus 𝟯% for the S&P 500.

➤ Semiconductor stocks rebounded with 𝗦𝗢𝗫 up 𝟭𝟮% this week.

➤ Firm recommends 𝗮𝗴𝗴𝗿𝗲𝘀𝘀𝗶𝘃𝗲𝗹𝘆 adding positions in semiconductor stocks.

➤ Cantor views the 𝗔𝗜 𝗶𝗻𝗳𝗿𝗮𝘀𝘁𝗿𝘂𝗰𝘁𝘂𝗿𝗲 buildout as durable and long-term.

➤ Top sector preference is 𝗠𝗲𝗺𝗼𝗿𝘆/𝗦𝘁𝗼𝗿𝗮𝗴𝗲, followed by Compute and Semi Cap Equipment.

➤ Top memory picks are 𝗠𝗶𝗰𝗿𝗼𝗻 (MU) and 𝗦𝗮𝗻𝗗𝗶𝘀𝗸 (SNDK).

➤ Preferred AI compute names are 𝗡𝗩𝗜𝗗𝗜𝗔, 𝗕𝗿𝗼𝗮𝗱𝗰𝗼𝗺, and 𝗔𝗠𝗗.

➤ Key upcoming catalysts include 𝗔𝗠𝗔𝗧 earnings and 𝗦𝗮𝗻𝗗𝗶𝘀𝗸 Investor Day.

👉 𝗪𝗵𝘆 𝗧𝗵𝗶𝘀 𝗠𝗮𝘁𝘁𝗲𝗿𝘀:

➤ Signals improving sentiment toward AI and semiconductor stocks after a sharp correction.

➤ Highlights continued confidence in long-term AI infrastructure spending.

➤ Identifies sectors and companies analysts believe offer the strongest recovery potential.

👉 𝗘𝘅𝗽𝗲𝗿𝘁 𝗦𝘁𝗮𝘁𝗲𝗺𝗲𝗻𝘁:

𝗖𝗮𝗻𝘁𝗼𝗿:

"After a massive drawdown over the last 5+ weeks (with SOX down ~30% peak to trough vs. S&P 500 -3%), Semiconductor stocks have finally caught a bid (SOX +12% this week vs. S&P500 +3%). We would be adding to positions aggressively as we view today's AI Infrastructure buildout as durable and elongated though recognize the rate of change from here is important too and that data points are no longer going to be uniformly positive. Thus, selectivity will matter. But with peak to trough declines of 32% for SMID Cap SPE, 27% for memory, 25% for Analog, and 24% for Large Cap SPE, we view the selloff as overdone. Our rank order of preference remains Memory/ Storage, then Compute, and then SPE. Our Top Picks are MU & SNDK in Memory, WDC & STX in Storage, NVDA, AVGO, & AMD in AI Compute, and ASML, AMAT, LRCX, KLAC, TER, MKSI, & ONTO in Semi Cap. Potential Catalysts: Earnings season continues with reports from covered companies CAMT and AMAT. SNDK is also hosting its Investor Day Thursday. Other key reports include LITE, CRWV, SMCI, COHR, and CBRS."

TSMC’s revenue +44.7% year-on-year in July to a fresh record high NT$467.58 billion (US$14.61B), amid robust global AI-related demand.

The previous high was NT$442.68 billion in June.

TSMC did not provide a US$ figure. The figure is based on its forex guidance for the 3rd quarter. 1/3 $Taiwan Semiconductor(TSM.US) $Apple(AAPL.US) $NVIDIA(NVDA.US) $AMD(AMD.US) $Broadcom(AVGO.US) $Qualcomm(QCOM.US) #semiconductors #semiconductor

Source: Dan Nystedt

Just some TLDR notes on $Applied Optoelectronics(AAOI.US) earnings:

- Expects full qualification of 1.6T products by their hyperscaler customer within next couple of weeks

(helpful revenue ramp #2, timelines)

- Continues to believe that AOI will have the largest AI DC transceiver production capacity in the US

(reiterating ambitions during a time when their competitors might get banned)

- Expects facilities toward InP capacity to come online in early 2027

(timeline FYI in terms of further ramp)

- Total capacity is approaching 200,000/u per month, EOY 650,000/u per month of 800g/1.6t. EOY 2027, 930,000/u

(this is the ramp i want to see)

- "increase our manufacturing capacity for our external light source or ELSFP. That's for co-packaged optics or CPO". we anticipate ramping production later this year and into 2027, ultimately culminating in about 400,000 pieces per month in 2028

(need some time to model this into revenue)

- "As we have mentioned before, we've been manufacturing lasers internally for many years. This has allowed us to avoid some of the shortages that have affected others in the industry"

(vertical integration bull case during CW/EML laser shortages)

- We believe that in the future, CPO will continue to drive increased demand for high-power lasers

(thesis validation on CPO sector)

- "to our long-term objective of returning non-GAAP gross margins to around 40%"

- "We ended the second quarter with $508.8 million in total cash equivalents"

I need to double check if the ATM finished or not

- "our ability to deliver revenue in general, and specifically when it comes to 800G products, is limited by our production capacity right now" "If we could produce more, we could ship more right now"

Demand > Supply validation.

- "Most of the increased capacity will be in U.S. Even so, let me say that, like I keep emphasizing, that is not good enough for the customer demand. The customer demand is 20%-40% higher."

Unholy photonics demand validation across the whole sector, read through for $Lumentum(LITE.US), $SIVE / $Jabil(JBL.US), and others is amazing.

- "Not in the next two, three years, especially the demand is so big. Okay? Even combined AI, $Lumentum(LITE.US), $Coherent Corp.(COHR.US), $Broadcom(AVGO.US) all together, it's still very tough to meet the customer demand in the next few years"

More optical sector demand validation.

- CPO Timelines: "If you're talking about really high volume manufacturer [for CPO market], I would say more like the late Q3 next year" and "We have been working very close with at least five customer"

If you care about current earnings (which I'm not really looking at closely)

Revenue: $191M vs. $190m

EPS: $.06 vs. $.02

TLDR: Extraordinary demand across the laser + optical sector read through. Kinda supporting Lumentum CEO statement that laser shortage is worse than memory shortages.

2027 capacity ramp on track. To map inflection period with timelines, would be around early next year, as stated in their previous earnings call.

AAOI has the customers now. Limitation is making enough lasers and transceivers.

Citi: Optical

Core Market Growth

> Market Opportunity: The global optical interconnects market is expected to hit USD 92 billion by 2028 (with a 65% 3-year CAGR), accompanied by an Average Selling Price (ASP) growth of an 18% CAGR.

Optical Chip

> Optical Chip Demand: Total demand is estimated at 1.714 billion units by 2028 (~62% to 63% CAGR from 2025).

> CW Laser Chips: Projected to see explosive growth to 987 million units by 2028 (~114% CAGR).

> EML Chips: Projected to reach 718 million units by 2028 (~34% CAGR).

$Applied Optoelectronics(AAOI.US) $Macom Tech(MTSI.US) $Coherent Corp.(COHR.US) $Lumentum(LITE.US) $Marvell Tech(MRVL.US) $Broadcom(AVGO.US)

📢 𝗝𝗨𝗦𝗧 𝗜𝗡: $Broadcom(AVGO.US) Broadcom Unveils AI-Powered VMware Security and Load Balancer Enhancements

👉 𝗞𝗲𝘆 𝗛𝗶𝗴𝗵𝗹𝗶𝗴𝗵𝘁𝘀:

➤ 𝗕𝗿𝗼𝗮𝗱𝗰𝗼𝗺 introduces major updates to 𝗩𝗠𝘄𝗮𝗿𝗲 𝘃𝗗𝗲𝗳𝗲𝗻𝗱 and 𝗔𝘃𝗶 𝗟𝗼𝗮𝗱 𝗕𝗮𝗹𝗮𝗻𝗰𝗲𝗿.

➤ New features strengthen multi-layer cybersecurity for 𝗩𝗠𝘄𝗮𝗿𝗲 𝗖𝗹𝗼𝘂𝗱 𝗙𝗼𝘂𝗻𝗱𝗮𝘁𝗶𝗼𝗻.

➤ 𝗔𝗜 𝗔𝘀𝘀𝗶𝘀𝘁𝗮𝗻𝘁 simplifies troubleshooting, remediation, and deployment.

➤ 𝗔𝗧𝗣 𝟭-𝟮-𝟯 enables advanced threat prevention deployment in weeks.

➤ Native 𝗔𝗣𝗜 protection secures VMs, Kubernetes, and AI workloads.

➤ New 𝟮-𝗻𝗼𝗱𝗲 𝗦𝗦𝗣 deployment reduces hardware requirements by up to 𝟯𝟯%.

➤ Distributed Firewall throughput reaches up to 𝟳𝟱𝗚𝗯𝗽𝘀 per server.

➤ Avi Load Balancer throughput increases to 𝟭𝟮.𝟮𝟱𝗧𝗯𝗽𝘀 per controller.

➤ Automated migration tool accelerates transition from legacy firewalls.

👉 𝗪𝗵𝘆 𝗧𝗵𝗶𝘀 𝗠𝗮𝘁𝘁𝗲𝗿𝘀:

➤ Strengthens Broadcom's enterprise AI and private cloud security portfolio.

➤ Addresses rising AI-driven cyberattacks with integrated multi-layer protection.

➤ Helps enterprises reduce infrastructure costs while improving security performance.

➤ Expands VMware Cloud Foundation's appeal for AI and mission-critical workloads.

👉 𝗘𝘅𝗽𝗲𝗿𝘁 𝗦𝘁𝗮𝘁𝗲𝗺𝗲𝗻𝘁𝘀:

𝗨𝗺𝗲𝘀𝗵 𝗠𝗮𝗵𝗮𝗷𝗮𝗻, Vice President and General Manager, Application Networking and Security Division, Broadcom:

“As AI-fueled cyberattacks and vulnerability exploits redefine the threat landscape, a fragmented security approach is no longer an option. With today’s updates to vDefend and Avi Load Balancer, we are giving enterprise customers the protection, performance, and automation they need to defend their application infrastructure at scale.”

𝗦𝗰𝗼𝘁𝘁 𝗥𝗼𝘀𝗲𝗻𝗯𝗲𝗿𝗴, Lead Architect in the CTO Office at TeraSky:

"By integrating powerful AI directly into the product, the AI Assistant for vDefend and Avi streamlines operations by quickly showing real-time insights, assisting with troubleshooting, and accelerating deployments. This powerful combination can make day-to-day operations faster, simpler and more contextual."

𝗔𝗵𝘀𝗮𝗻 𝗞𝗵𝗮𝗻, Chief Technology Officer of ThinkOn:

“API protection has become paramount for securing our private cloud application and AI traffic. Avi WAAP extends our existing web security posture to defend against API layer threat vectors. Additionally, Avi’s software-defined scale-out architecture makes it very easy to roll out WAAP comprehensively across our VCF private cloud workloads.”

𝗖𝗵𝗿𝗶𝘀 𝗦𝘁𝗲𝗳𝗳𝗲𝗻, VP of Research at EMA:

"Frontier AI assisted attacks have made multi-layer defense a necessity for private cloud workloads. With prescriptive ATP 1-2-3 workflows, enterprise customers can enable virtual patching and advanced threat detection comprehensively. What could have taken many months can now be rolled-out in a matter of a few weeks."

𝗡𝗮𝗴𝗮𝗿𝗮𝗷 𝗥𝗲𝗱𝗱𝗶, VP of Technology at United States Senate Federal Credit Union:

"As a financial institution, securing virtualized, Kubernetes and AI applications is a top priority. With prescriptive ATP 1-2-3 workflows in vDefend, we can efficiently implement virtual patching and advanced threat prevention in a matter of weeks, rather than many months. Furthermore, Avi WAF and WAAP capabilities further protect web applications and API traffic as part of our defense-in-depth security architecture.”

Intel’s EMIB-T advanced packaging costs 50% less than TSMC CoWoS, media report, adding EMIB-T will begin mass production next year as yields now exceed 90% - though another report says a 50% substrate yield remains the bottleneck.

-EMIB-T has already won Intel at least one AI ASIC chip project from MediaTek (likely a Google TPU)

-Broadcom and Meta are now evaluating EMIB-T.

-Two Intel partners in Taiwan are rapidly expanding production: Powerchip as its exclusive silicon capacitor supplier, and UMC for silicon bridges.

-Taiwan’s Unimicron is the substate partner for EMIB-T.

-For Intel EMIB-T cost and yield data, Taiwan media cite Wccftech for Intel EMIB-T cost, yield data, and Wccftech cites X handle @imnotharsh $Intel(INTC.US) $United Microelectronics(UMC.US) $Alphabet(GOOGL.US) $Broadcom(AVGO.US) $Meta Platforms(META.US)

Source: Dan Nystedt