Who Will Disrupt the Magnificent Seven

While the U.S. government alternates between subsidizing big tech and threatening to bust up the best companies in America, under the radar fly hundreds of small companies with potentially disruptive technologies that might become the big tech of the future.

As Trump advisor Stephen Moore quipped in an excellent recent piece at Hot Air, “raise your hand if you had heard of Nvidia or OpenAI five years ago.”

The threat the most interesting emerging tech stars pose to the big guys is that they will do more for less, riding the learning curve, cutting prices, and making the super-expensive devices of the leaders unnecessary.

Consider a company we’re sure you’ve never heard of (unless you subscribe to our sister publication, George Gilder’s Moonshots): Skywater Technologies (SKYT). SKYT is the only U.S.-based and solely U.S.-owned pure-play semiconductor foundry (contract manufacturer). It offers specialized wafer manufacturing for both defense and commercial applications.

SKYT is a Department of Defense DMEA Category 1A Trusted Supplier. Not surprisingly, its customers tend to be cutting edge as well: teams with breakthrough concepts that they can’t execute without a deeply engaged manufacturing partner intimately involved in device design.

Long gone are the days when progress in the chip industry was confined to one more scale-down of the circuitry with the resulting homogenous chips sucked up by design firms that were largely siloed from the foundry manufacturers. Even the biggest foundries today, like Taiwan Semiconductor (NYSE: TSM), work intimately with the chip-design firms, such as Nvidia (NASDAQ: NVDA), Qualcomm (NASDAQ: QCOM, and Advanced Micro Devices (NASDAQ: AMD).

Skywater pushes that envelope. The company’s Technology as a Service (TaaS) model and Advanced Technology Services (ATS) business engage small innovative customers in exceptionally early collaboration. Good for the customers, as few are likely to get the attention they need from a behemoth such as TSM. Good for Skywater because this approach builds a pipeline of future customers and revenues.

Catering to innovators, SkyWater’s foundry services span specialized processes including mixed-signal devices, radiation-hardened integrated circuits, power management, MEMS (microelectronic machines), superconducting integrated circuits, photonics, and carbon nanotubes.

This spread of specialized capabilities has allowed SKYT to support emerging markets and technologies, particularly in quantum computing, AI, and biomedical applications. As testimony to their success: we find ourselves consulting SKYT’s customer lists to find other innovators for our reader/investors.

Our readers know that we are skeptical of what is called “quantum computing.” SKYT’s clients, however, include at least one company taking an approach to quantum computing we find interesting.

Quantum computers use “Qubits” that multiply computing power because a Qubit can simultaneously be in an off and on state, rather than the simple off and on of digital devices. Alas, Qubits are unstable, and it has been very hard to string enough of them together to do effective computation.

Qubits take a variety of physical forms, often subatomic particles such as a single electron.

PsiQuantum uses photons as Qubits. This isn’t just another Qubit architecture—or another quibble. It’s a fundamentally different paradigm with distinct advantages. PsiQuantum’s system Photonic operates at room temperature rather than requiring extreme cryogenic cooling as most quantum computers do. And while most quantum efforts so far struggle to generate 100s of Qubits, SkyWater is working with PsiQuantum to develop silicon photonic chips targeting over 1,000,000 Qubits for fault-tolerant systems.

Another Skywater client to which we have alerted readers of the Gilder Technology Report is Weebit Nano, a pioneer in innovative “ReRAM memory (resistive random access memory),” which stores data by changing the resistance of a solid material and is non-volatile, keeping its data when the power is turned off, unlike DRAM in your PC.

Okay, all very cool but what was that you said about less expensive?

Skywater’s pathway to lower costs is using old technology to make new technology.

Most industry players still compete by making ever smaller circuits—7nm, 5nm, 3nm, and now 2nm—in ever more complex packages. SKYT is following a different path. It is using legacy manufacturing processes an order of magnitude less expensive in the development stages. Nobody can make a 2nm chip unless the development costs are amortized over hundreds of millions of devices. (See graph below, which stops at 5nm.)

Image courtesy of WCCF Tech.

Photonic computing, which PsiQuantum does with SKYT’s help, has no need for the tiniest of circuits because photonic components are fundamentally limited by the wavelength of the lasers used—typically 700 to 2000 nm—making further miniaturization unnecessary. This constraint aligns perfectly with mature semiconductor manufacturing nodes, particularly 28nm and 45nm processes, which provide all the precision needed for photonic components.

This resurgence of “legacy” processes for the most theoretically advanced computation is a strategic pivot from the traditional semiconductor roadmap. Instead of chasing Moore’s Law toward 2nm and beyond, legacy processes—backed by decades of industry investment and experience—offer a cost-effective path to quantum computing. It’s a powerful example of how mature technology can enable revolutionary advances through strategic repurposing of proven manufacturing excellence.

Similarly, WeebitNano is enabled by SKYT’s legacy processes. A ReRAM memory cell uses only a single transistor, a huge advantage over other technologies such as SRAM, which requires six transistors to hold one bit. ReRAM thus has the huge advantages of a smaller size and lower power consumption. ReRAM’s single transistor, however, is considerably larger than most of today’s circuitry, meaning it can be manufactured using cheaper legacy processes.

Skywater wins clients like PsiQuantum and Weebit Nano by starting with legacy processes and then intensely customizing those processes for the special needs of its customers. Figuring out how to be more innovative by starting with older technologies is SKYT’s competitive edge.

The moral: as ever, the most serious challenges to the current big boys will come from below… and no one in Washington will see it coming.