Skip to the stories
Microsoft 365 & SharePoint

Toshiba QKD Goes Live in Albuquerque as SQS Builds a Managed Quantum Security Service

SQS has live quantum key exchange at two Albuquerque nodes, with quantum timing there too. Combining those keys with PQC is still in integration.

Toshiba QKD Goes Live in Albuquerque as SQS Builds a Managed Quantum Security Service

Secured Quantum Services (SQS) has moved past lab demonstrations. Its Albuquerque, New Mexico, site runs Toshiba quantum key distribution (QKD) systems alongside Celestica network switches, and the keys it produces are already flowing between nodes. What's still unfinished is the part most enterprise buyers will care about: wrapping those quantum-delivered keys in NIST post-quantum cryptography (PQC) algorithms. For IT teams, that gap defines what can be evaluated today and what has to wait.

SDxCentral Magazine - Issue 2
Albuquerque

What's actually running in Albuquerque

SQS's deployment ledger lists two Albuquerque nodes operating since September, along with a second pair in Mesa, Arizona. Key exchange is running at both sites. Quantum-based timing is listed only in Albuquerque, which means the Mesa deployment is currently a key-distribution footprint and nothing more.

That distinction matters when you read vendor announcements. SQS defines a site as live once its nodes are exchanging keys, not once every planned capability is switched on. That's a sensible engineering milestone, but it's a narrower claim than "fully operational quantum security service." If you're scoping a project, treat a live site as proof that the key-exchange layer works in the field, and ask separately about every other component you need.

Close-Up of an M.2 NVMe SSD

The company also says it has hosted hands-on technical sessions for U.S. evaluators and commercial prospects, giving them access to the deployment. SQS didn't name paying customers in its announcement, so there's no public reference list to lean on yet.

A managed service, not a parts list

The more interesting shift here is commercial and operational. SQS is positioning itself as the party responsible for integrating and running equipment from several vendors. Customers would connect their existing infrastructure to SQS nodes instead of sourcing, configuring and maintaining QKD hardware, timing gear and encryption software themselves.

Quantum

SQS founder and CEO Ray Schouten put the operational burden plainly: "The hard part is the work nobody sees: integration, timing, uptime, keeping systems from different vendors running together in the field day after day." That's a fair description of where quantum projects tend to stall. Hardware from one vendor usually works in isolation. The difficulty comes when key delivery, clock synchronisation and classical network gear have to behave as one system under real maintenance windows and failures.

Teams in this position typically find that the vendor relationship changes shape. Instead of managing three or four supplier contracts and their support tickets, you manage one service boundary with defined uptime expectations. Whether that trade is worth it depends on how much in-house quantum expertise you already have, and how much you'd prefer to hand off.

1 The AI Hardware Supplement_page-0001

How the deployment is built

SQS's deployment sequence starts with installing a pair of nodes and establishing QKD between them. Only after that does it integrate additional capabilities into the site. The sequence is logical: a working key path is the foundation everything else depends on, and it gives you a measurable checkpoint before more complexity arrives.

SQS's platform documentation describes a transport model in which customer traffic passes through its nodes without requiring application changes. For network and security teams, that's the detail to test hardest. Transparent transport sounds simple, but it still means traffic paths, latency, MTU behaviour and failover need to be validated against your own topology. The documentation is the place to start those conversations, not a substitute for them.

NEC

Where the encryption layer stands

The intended design combines post-quantum cryptography with a separate protection layer, and SQS says integration of that dual-layer approach is underway. The company describes the quantum-delivered keys and NIST PQC algorithms as working together to protect messaging between nodes. At both live sites, though, that dual-layer encryption remains in integration.

That's a meaningful caveat. Key distribution and key-based encryption are separate problems, and the combined stack is where interoperability issues tend to surface. If you're assessing this service for sensitive data, the questions to ask are concrete: which algorithms are in the production path, how key material moves between the QKD layer and the PQC layer, and what happens to traffic if one layer degrades. SQS hasn't published answers in the material available, so expect to request them directly.

Facepalm statue

Post-quantum migration is also a planning exercise in its own right, and QKD doesn't replace it. Many organisations will need PQC on their own systems regardless of whether they ever adopt quantum key delivery. Treat this service as one option within a broader cryptographic inventory rather than a shortcut around it.

Timing as a second application

Quantum timing is the second infrastructure application in the Albuquerque build. SQS says its optical clock distributes synchronisation without relying on GPS. The company names financial markets, settlement systems and other distributed operations as intended uses.

For operations teams, GPS-independent timing addresses a real concern: satellite signals can be disrupted, and many time-sensitive systems depend on them. A terrestrial timing source that doesn't rely on that signal path is worth evaluating on its own terms, even if you never adopt the encryption side. The practical questions are accuracy targets, how the clock feeds your existing time servers, and how the timing service is monitored. Nothing in SQS's announcement gives specifications for those, so they belong on your checklist.

Pilots, acceptance criteria and next sites

SQS proposes starting with pilots that cover an individual customer system or data flow. Scope and acceptance conditions would be agreed before installation, and results would be measured against those conditions. That's the right structure for a technology this new. A narrow pilot with written pass/fail criteria gives you evidence without committing your whole estate.

If you go this route, define success before anything is connected. Decide which data flow is in scope, what latency and availability you'll accept, how key exchange failures will be detected and escalated, and who signs off. Ask SQS how it reports on those metrics, and whether you'll have direct visibility into node status or only what the provider chooses to share.

The deployment ledger lists Colorado Springs, Colorado, and Brooklyn, New York, as the next locations. Nodes haven't been deployed at either site yet, so they're plans rather than capacity. Watch for the same two markers that matter in Albuquerque: whether key exchange goes live, and whether quantum timing and the PQC combination follow. Those milestones will tell you more about the service's maturity than the number of sites on a map.

My read is that SQS is betting that enterprises would rather buy a quantum security outcome than assemble one. That's a credible model, and the Albuquerque deployment gives it a working foundation. The unanswered part is whether the combined encryption layer and the managed-operations promise hold up under production conditions. That's the question to revisit once the dual-layer integration is finished.

Source: James Rodriguez · www.sdxcentral.com

Same flag