Spotlighting the Future

Quantum computing has long occupied that space between scientific curiosity and futuristic promise. But in the past few years, that promise has begun to look more like a plan. I work on quantum interconnects — technologies that help different quantum chips talk to each other — and in my view, the field is entering a critical new phase.

Let’s start with the basics. Traditional computers process information using bits — zeros and ones. But quantum computers use qubits, which are governed by the laws of quantum mechanics. That means they can exist in a state called superposition, where a qubit is both zero and one at the same time, at least until it’s measured. Qubits can also become entangled, which allows the state of one to be dependent on the state of another upon measurement, even across distance.

These properties open up new possibilities for solving certain types of problems. For example, factoring large numbers is notoriously hard for classical computers, and that difficulty underpins much of modern encryption. A quantum algorithm known as Shor’s algorithm could, in theory, factor those same numbers much faster than any known classical algorithm, making it a potential game-changer for cybersecurity. While we’re not there yet, the threat has spurred the development of post-quantum cryptography: algorithms designed to resist attacks even from quantum computers.

Another exciting application is quantum simulation. In this area, we use qubits to model other quantum systems, such as molecules or materials. This could drastically improve how we design new pharmaceuticals or study protein folding — tasks that are challenging on classical computers because of the complexity of quantum interactions.

So what’s holding us back? Simply put: errors. Quantum systems are fragile. A state-of-the-art classical processor operates with a low error rate of around 10⁻²⁰. For quantum computers, we're currently at about 10⁻³. It’s a massive gap. To do anything useful, we need millions of qubits working with extremely high fidelity. And to get there, we need quantum error correction.

Instead of relying on a single qubit to represent one quantum bit of information, error correction schemes encode one “logical qubit” into many physical qubits. The idea is that while individual qubits might fail, the system as a whole can still function reliably. In 2024, Google Quantum AI demonstrated a proof-of-principle version of this using superconducting qubits, showing that adding more noisy qubits could still reduce the overall error rate. That’s a major milestone.

And it’s not just superconducting qubits. Other groups, like one at Harvard, are working on neutral atom qubits—where actual atoms are used as information carriers and moved around with lasers to implement quantum logic gates. They too are beginning to demonstrate early forms of error correction. We don’t yet know which type of qubit will “win.” It’s entirely possible that the most practical quantum computers will be hybrids, using one type for memory and another for computation.

This brings us to scalability. Today’s most advanced processors, from industry leaders like IBM Quantum and Google Quantum AI, contain up to a few hundred qubits. But we’ll need far more than that. To make quantum computers practical, we’re starting to think in terms of distributed systems—multiple quantum chips working together. That’s where my research comes in: developing quantum interconnects to move quantum information between processors efficiently.

Looking five to ten years ahead, we may see quantum processors with a few thousand qubits operating in tandem, enabled by this modular, networked architecture. We’re still early, but the roadmap is becoming clearer. While quantum computing faces enormous technical challenges, the research community remains optimistic that with time, effort, and investment, these hurdles can be overcome.

6G is in flux. Network operators—companies like T-Mobile, AT&T, and Verizon—are asking what can create the market pull for the extra wireless capacity 6G will provide. That's unusual. 

Unlike previous generations, we don't have an immediate problem that more bandwidth can solve. The big question isn't whether engineers can make 6G work. They will. It's what transformative use case will generate the demand.

Networks evolve in roughly ten-year cycles. In 2007, the transition to smartphones fundamentally changed global communication – requiring networks that could handle vastly more data – fueling 4G's rollout around 2010.

The situation is markedly different today. Current wireless networks already provide reliable connectivity. 5G succeeded technologically when it arrived a decade after 4G. Yet some skeptics still call it a “failure,” because it's missing a killer application that fully exploits its advances, let alone ignites the momentum for another generational leap.

So where might that groundswell of demand come from? 

Many expect virtual and augmented reality, or connected vehicles, to create it. Meta and Apple have invested hugely in VR. 6G will likely offer very high data rates and low-latency that both require. If either scale widely enough, they may be 6G catalysts.

Another potential avenue is satellite integration. Starlink has revolutionized satellite communications by deploying over 8,000 active Low Earth Orbit (LEO) satellites, an unprecedented mega-constellation that is rapidly growing. Enabled by reusable launch vehicles, SpaceX has dramatically lowered the cost and frequency of satellite deployment, making global broadband coverage a reality and connecting areas where cellular infrastructure does not exist.  

One 6G vision is to connect phones directly to satellites the way they do to towers today. But because satellites orbit hundreds of kilometers above Earth, far beyond the short hops to nearby towers, delivering reliable, high-bandwidth service is a daunting challenge. Solving it could be a defining feature of 6G.

Security and cryptography could also be key demand drivers. Today's networks weren't designed for resiliency against hostile attacks, a gap that 6G could address. A major security breach or cyberattack could make people suddenly prioritize protected communications, creating the urgency that drives 6G adoption.

Further afield, so-called “privacy-preserving computation” techniques — such as homomorphic encryption, and secure multi-party computation — offer a transformative shift in how sensitive data is handled. Instead of moving raw data to centralized servers, these methods enable processing while keeping personal information encrypted.  

If deployed, they could redefine user privacy by allowing organizations to train models or perform analytics without ever accessing the underlying data, dramatically reducing exposure risks and enabling compliance with stringent data protection laws. These methods, however, explode the size of data, potentially creating a significant new demand for network capacity. 

Meanwhile, engineers are preparing the technical foundation for 6G. One key area is finding new radio frequencies to use. Just as 5G opened up very high frequencies called millimeter waves, 6G will likely tap into what's called upper mid-band spectrum, mid-range frequencies that fall between today's cellular bands and 5G's highest frequencies.

The complication is that satellite companies already use these frequencies. Making both systems work without interfering with each other will require careful coordination.  

A second bottleneck that needs to be addressed is the computational fabric – the microchips in handsets and network infrastructure on which wireless systems run. Wireless systems are supporting ever larger numbers of antennas, wider bandwidths, and greater data rates, all of which need more processing at lower power.   

The situation is similar to AI with continually growing models, and wireless can benefit from the strides in digital processors to support AI.  AI itself is also being widely used for improving wireless communication design and operation and could be a piece of the 6G puzzle. 

These technical preparations won't create demand on their own, but they'll be essential when new applications finally emerge. 

The NYU WIRELESS center is well-positioned to continue its role in developing future wireless systems. Most advances for 6G will come from collaboration across multiple areas of expertise including wireless communications, circuit design, AI, networking, and security, as well as potential developers using wireless including virtual reality, vehicular systems, robotics, and healthcare.  

NYU WIRELESS uniquely brings these diverse talents together along with our collaboration with leading industry partners.   

6G will likely roll out around 2030, consistent with the ten-year cycle. Whether 6G’s transformative use case is AR/VR, connected cars, advanced cryptography, or something no one has yet imagined remains to be seen. There's no technical reason 6G can't happen. It's just a matter of finding what people will actually need it for.