A Full Analysis of Celestica ($CLS)

In partnership with

Hi everyone,

Today we’re be breaking down a company that has become a “golden screw” in the AI buildout.

Stock Deep Dive: Celestica Inc. (CLS-US, $35.2B MCAP)

Celestica is one of the cleanest “picks and shovels” plays on the AI data center arms race.

Hyperscalers are spending tens of billions to scale GPU clusters, and that money has to flow into the unglamorous but mission-critical hardware that makes AI run: high-speed switches, servers, rack integration, and increasingly liquid cooling.

Celestica sits right in that spend path, with 800G upgrades today and 1.6T plus full-rack AI systems ramping into 2026.

The stock has already re-rated hard, so the bar is higher, but if AI capex stays durable and Celestica keeps executing, this is the kind of under-the-hood infrastructure winner that can keep compounding.

• Why Now? 👉 The AI-Fueled Momentum

• Overview 👉 What Does Celestica Do?

• Business Units 👉 Segment Breakdown

• Celestica’s Role in the AI Datacenter Ecosystem

• Key Drivers 👉 Unlocking Performance

• By The Numbers 👉 Key Metrics

• Competition and Outlook

• Risks 👉 Potential Pitfalls

• Wrapping Up…

Why Now? 👉 The AI-Fueled Momentum

Celestica has re-rated from a sleepy electronics manufacturer into a leveraged play on hyperscaler AI capex. The setup is straightforward: cloud giants are rebuilding data centers around AI training and inference, and that requires a step-change in networking, compute density, and rack-level integration. Celestica is benefiting because it is not just assembling boxes, it is increasingly co-designing and producing high-speed switching platforms (800G today, 1.6T next), AI server deployments, and full-rack solutions that bundle servers, switching, and liquid cooling.

Recent program wins matter because they extend beyond one-off builds and move Celestica deeper into customers’ multi-year roadmaps, with production ramps expected to build through 2026. Management’s tone has turned structurally bullish on data center demand, raising guidance and leaning into capacity expansion to support booked business. In a market looking for “picks and shovels” exposure to AI infrastructure, Celestica sits in the spend path whenever a hyperscaler upgrades network fabrics and scales GPU clusters.

Overview 👉 What Does Celestica Do?

Celestica is an electronics manufacturing and supply chain partner that has evolved toward higher-value design, engineering, and platform-led manufacturing. Historically, it looked like a classic contract manufacturer, building electronics to customer specifications. Over the last several years, it has pushed up the stack into more complex programs where it can contribute design expertise, system integration, test, and end-to-end supply chain execution.

The business is organized into two segments. Advanced Technology Solutions (ATS) serves aerospace and defense, industrial, health-tech, and capital equipment, typically lower-volume programs that prioritize reliability, regulatory compliance, and specialized engineering. Connectivity and Cloud Solutions (CCS) serves communications and enterprise customers, including networking hardware, servers, and storage tied directly to data center infrastructure. CCS has become the growth engine and the key driver of investor interest because it is exposed to AI clusters, high-speed switching upgrades, and hyperscaler buildouts. In plain terms, Celestica is a behind-the-scenes builder of mission-critical hardware systems that power modern data centers, while maintaining a diversified base of non-cloud programs through ATS.

Business Units 👉 Segment Breakdown

Celestica’s mix has tilted decisively toward cloud and communications hardware, which is why the equity now trades like an AI infrastructure beneficiary rather than a generic EMS name. CCS is roughly three-quarters of revenue, with ATS around one-quarter, and the gap has widened as hyperscaler demand surged. Recent quarters illustrate the divergence: CCS has delivered rapid growth while ATS has been comparatively stable to slightly down depending on end-market timing. Within CCS, Celestica effectively straddles two demand streams: communications, which includes high-speed networking and telecom-related platforms, and enterprise, which includes servers and storage. The communications side has been the standout as hyperscalers migrate network fabrics from 400G to 800G, a transition that pulls through large volumes of advanced switches and related integration. Enterprise has also contributed via next-gen AI and ML compute deployments, but networking has been the headline driver.

A key proof point is Celestica’s High-Performance Systems business, which has become a meaningful slice of total revenue and reflects the shift from “build to print” assembly toward more proprietary, higher-value platforms. ATS remains strategically important as a stabilizer, but the growth and margin narrative is overwhelmingly CCS-led.

Become An AI Expert In Just 5 Minutes

If you’re a decision maker at your company, you need to be on the bleeding edge of, well, everything. But before you go signing up for seminars, conferences, lunch ‘n learns, and all that jazz, just know there’s a far better (and simpler) way: Subscribing to The Deep View.

This daily newsletter condenses everything you need to know about the latest and greatest AI developments into a 5-minute read. Squeeze it into your morning coffee break and before you know it, you’ll be an expert too.

Subscribe right here. It’s totally free, wildly informative, and trusted by 600,000+ readers at Google, Meta, Microsoft, and beyond.

Celestica’s Role in the AI Datacenter Ecosystem

Celestica’s differentiated role is that it sits at the bottlenecks of AI scaling, especially networking and rack integration. In large AI training clusters, GPUs are only as productive as the network fabric that keeps them fed with data and synchronized during training. That is why the shift from 400G to 800G switching has been so powerful for Celestica, and why the move to 1.6T is the next catalyst. The company has effectively become an early, trusted supplier for hyperscaler networking upgrades, capturing share as customers standardize on higher-speed platforms.

More importantly, Celestica is expanding from building switches into building systems. Recent wins include full-rack AI programs where it co-develops and manufactures integrated racks combining servers, high-speed switching, and liquid cooling, which is increasingly required as power density rises. That pushes Celestica deeper into customer roadmaps and makes it harder to swap out with a pure low-cost assembler. The “picks and shovels” angle is real: regardless of which AI model wins, hyperscalers must buy racks, switches, interconnects, and integration services to scale capacity. Celestica is positioned to capture that spend through 2026 to 2027 as architectures get denser and networking requirements intensify.

Key Drivers 👉 Unlocking Performance

First, the strategic pivot toward design-led, higher-value programs has changed the earnings algorithm. By contributing engineering, proprietary platforms, and system-level integration, Celestica captures more margin than traditional assembly work and becomes embedded earlier in product cycles. Second, operating leverage has emerged as volumes scale. Once facilities, labor, and qualification processes are in place, incremental units in data center programs can carry attractive flow-through, supporting rising operating margins even as the mix shifts between networking and servers. Third, customer partnerships and visibility matter more than most investors appreciate.

Subscribe to GritAlpha Premium to read the rest.