• Tech Stock Insider
  • Posts
  • The Best AI Trades Left On The Board Are Not The Chipmakers You Already Own

The Best AI Trades Left On The Board Are Not The Chipmakers You Already Own

The picks-and-shovels layer is repricing while the crowd stays fixated on GPUs.

September 22, 2026

Everyone is still crowded into the same handful of accelerator names, but the real repricing this quarter is happening one layer down the stack.

Contract manufacturing, programmable logic, zero-trust networking, and sovereign AI cloud each have a catalyst inside the next 90 days. You get four ways to own the buildout without paying the multiple everyone else is paying.

Built Different (Sponsored)

Billions have poured into electric aviation.

AirCar took a different approach: build first.

Six full-scale aircraft. 100+ manned flights. And 270+ deposit-backed reservations worth roughly $27M at list price.

Now AirCar is raising at $1 per share.

See why investors are looking at AirCar

Data Centers

Alibaba Builds a Complete AI Stack From Chips to Data Centers

Alibaba Group (NYSE: BABA) has introduced the Zhenwu V900, a processor designed to train and run advanced AI models. The company calls it China’s most powerful AI chip and claims three times the performance of its previous-generation M890.

The processor can be connected in clusters of up to 500,000 chips for exceptionally large computing workloads. Alibaba expects commercial production and wider customer availability to begin early next year.

Much Larger Qwen Models Are Coming

Alibaba is also training Qwen 4, the next generation of its expanding family of foundation models. Future Qwen 4.5 and Qwen 5 systems are expected to contain between five trillion and ten trillion parameters.

The additional scale is intended to support assignments requiring longer planning, deeper reasoning, and more independent decision-making.

Alibaba also says its models are improving their ability to identify weaknesses, conduct experiments, and generate training data with limited human involvement.

Alibaba Builds the Entire AI Stack

The company plans to increase its worldwide data-center capacity to more than 20 gigawatts by 2032. That infrastructure will provide the computing power needed to train larger models and offer them commercially through Alibaba Cloud.

Alibaba is combining its processors, Qwen models, cloud platform, and 

data centers inside one connected technology business. The Zhenwu V900 moves the company closer to controlling every major layer required to compete in advanced AI computing.

Consumer Hardware

Google Creates a Premium Laptop Platform for Android Users

Alphabet (NASDAQ: GOOGL) has opened preorders for Googlebook, a new premium laptop platform starting at $899. Initial models from Acer, Asus, Dell, HP, and Lenovo will use Intel or Qualcomm processors and offer up to 14 hours of battery life.

The hardware includes dedicated neural processing units capable of more than 45 trillion operations per second. Buyers can also choose models with premium materials, haptic glass trackpads, and touchscreen displays reaching 2.8K OLED resolution.

Android and Chrome Come Together

Googlebook combines the Android technology stack with desktop foundations developed for ChromeOS. The result brings Android applications, a full Chrome browser, browser extensions, and Linux development tools onto the same machine.

Deeper phone integration allows messages, files, settings, and unfinished tasks to move between an Android device and the laptop.

Gemini is also built into the software to organize work, answer questions about screen content, create widgets, and support writing or coding assignments.

Google Targets Windows and Mac

Google has positioned Googlebook above the affordable Chromebook category instead of replacing it. The company wants to offer Android users the type of connected phone-and-computer experience Apple provides across the iPhone and Mac.

Each laptop includes one year of Google AI Pro, five terabytes of cloud storage, and software updates for up to ten years.

Googlebook gives Alphabet its strongest personal-computing platform yet, bringing hardware manufacturers, Android applications, Chrome, and Gemini into one premium ecosystem.

Elite Picks (Sponsored)

This report focuses on a narrow group of stocks identified through a detailed screening process.

Analysts apply a combination of metrics to narrow down potential opportunities.

Past selections have shown strong momentum, but no outcomes are guaranteed.

The newest edition is now open for access.

Get the report now.

*This free resource is being sent by Zacks. We identify investment resources you may choose to use in making your own decisions. Use of this resource is subject to the Zacks Terms of Service.
*Past performance is no guarantee of future results. Investing involves risk. This material does not constitute investment, legal, accounting, or tax advice. Zacks Investment Research is not a licensed dealer, broker, or investment adviser.

Enterprise Computing

Apple Starts Shipping New Macs Built for Local AI Work

Apple (NASDAQ: AAPL) has started shipping its upgraded Mac mini and Mac Studio computers, bringing its newest processors into the desktop lineup. The Mac mini offers M6 and M5 Pro chips, while Mac Studio configurations use the more powerful M5 Max and M5 Ultra.

Apple is positioning the machines for developers and businesses that want to run demanding AI models locally. Processing information on the computer can keep sensitive files out of the cloud while removing repeated fees tied to cloud-based AI usage.

Four Macs Run a Trillion-Parameter Model

Apple demonstrated four connected Mac Studios running an AI model containing one trillion parameters. The system used the model to identify and correct a graphics coding problem while drawing power from a single wall outlet.

The machines communicate through Apple’s RDMA over Thunderbolt technology, allowing memory and computing resources to work across several connected Macs.

Mac Studio can also include as much as 512GB of unified memory for models that would overwhelm a standard desktop computer.

Apple Pushes Deeper Into Enterprise Computing

Apple’s unified memory architecture places processing and memory close together, reducing the time required to move large amounts of information.

The same design principles now stretch from iPhones and Mac minis to Apple’s highest-performance desktop systems.

The company is presenting these Macs as an alternative to expensive AI workstations and rented data-center capacity. Apple is giving the Mac a new enterprise role as a private AI machine that can run complex models repeatedly without paying for every processed token.

Apple's forgotten third co-founder sold his 10% stake back to Jobs and Wozniak 12 days in. For how much?

Login or Subscribe to participate in polls.

Recent Tech Movers

Celestica Inc. (NYSE: CLS)
Riding the Hyperscaler AI Buildout

Celestica is one of the cleanest ways to play hyperscaler AI capex, and the market backs it up. Shares trade near the low end of a 52-week range of $315 to $656, and institutional accumulation keeps building. Nykredit A/S just disclosed adding another 25,458 shares to your radar.

The story is the mix shift. Celestica's Connectivity & Cloud Solutions segment, which builds the servers, switches, and storage sitting inside AI data centers, is now the growth engine you want to watch.

Hyperscaler orders keep expanding, and management has been pulling forward capacity as backlog builds.

Q3 results land in late October and give you the next real read on how fast the AI buildout is translating into revenue. Want the picks-and-shovels layer without paying Nvidia multiples? Celestica is one of the better setups on the board.

The risk to watch: A hyperscaler capex pause would hit this name harder than most.

Lattice Semiconductor (NASDAQ: LSCC)
200-Day Reclaim Signals Cycle Turn

Lattice just reclaimed its 200-day moving average for the first time in months, often the first signal a semi cycle is turning. The programmable logic maker got left behind while the AI names ran. Now that setup is flipping in your favor.

Two things worth your attention. First, the industrial and automotive markets Lattice sells into are showing early inventory normalization, which means design wins from 2024 and 2025 finally start converting to revenue.

Second, edge AI at the chip level (small, low-power FPGAs sitting next to accelerators) is a category Lattice owns. That tailwind hasn't been priced in.

The stock still trades at a rich multiple on trailing sales, so don't confuse this with a value name. A close above $122 confirms the higher-low structure and opens a path toward $135. Set your stop below the 200-day.

The risk to watch: Another leg down in industrial demand invalidates the whole setup.

Netskope (NASDAQ: NTSK)
SASE Re-Rate Story Taking Shape

Netskope is a cleaner cybersecurity re-rate story than the market credits.

The post-IPO overhang is fading, subscription mix keeps shifting toward higher-margin NewEdge AI modules, and free cash flow is finally trending in the right direction. The ingredients for a multiple re-rate are lining up in front of you.

Bigger picture: SASE and SSE (the secure-access categories Netskope owns, essentially the modern way to deliver zero-trust to remote workers) are what every enterprise board wants right now.

Netskope is one of only two or three vendors that can deliver it at hyperscaler-grade performance, so keep it on your radar.

Want to build a position? Start scaling in and add on weakness. The next hard catalyst is fiscal Q3 results in early December, where a clean beat-and-raise likely pushes shares toward the mid-$20s.

The risk to watch: Enterprise IT budget freezes into year-end would delay the whole thesis by a quarter or two.

Trillion Shift (Sponsored)

Elon Musk, Jeff Bezos, Bill Gates, and Jensen Huang are all backing the same little-known corner of AI.

It is not chips or software.

And one forecast cited in the presentation puts the market opportunity at $24 trillion.

Now one analyst is revealing his #1 stock pick positioned for this emerging trend.

Reveal the #1 Stock Behind This $24 Trillion AI Trend

The AI Cloud Wildcard With Real Backlog

Nebius Group (NASDAQ: NBIS)

Why it's on your radar

Nebius is running one of the most aggressive AI cloud capex programs in the market, with $20 to $25 billion planned for 2026 alone.

Contracted revenue climbed more than 450% year-over-year last quarter to $582 million, and the backlog is building faster than analysts modeled.

Your bet here is that AI compute demand keeps outpacing supply through 2027, and that Nebius can convert its capacity into signed enterprise contracts fast enough to justify the spend.

The catalyst and the risk

Over the next 90 days, watch for two things: incremental hyperscaler or sovereign-cloud contract announcements, and Q3 commentary in November that confirms capacity-to-revenue conversion is on track.

If both land, the setup opens a path back toward the 52-week high near $300. The risk cuts both ways. This capex plan requires equity or debt raises that could dilute you meaningfully, and any pause in AI capex from the majors would blow a hole in the thesis. Size it small.

Everything Else

That's our coverage for today; thanks for reading! Reply to this email with feedback or any tech stocks you want me to check out.

Best Regards,
—Noah Zelvis
Tech Stock Insider