KEYSKeysight Technologies, Inc.

“the opportunities that we size as our AI business, which is, I think, really the heart of your question, finished the first half in the $500 million to $600 million range, almost in line with what we did the whole of last year. So quite pleased with the progression of the opportunity in the wireline business. And so the AI portion of our business as we size it for you, is largely in the wireline segment.”

“In commercial communications, we continue to see accelerating momentum in our wireline business, driven by the ongoing AI data center expansions. Wireline delivered record orders again this quarter with robust demand for both R&D and manufacturing solutions. In the first half of fiscal 2026, our AI-related business has already surpassed the levels achieved in all of 2025.”

“this quarter, as an example, and even for the half, both R&D and manufacturing components of our business and portfolio doubled. If you look at our wireline business, still a very high percentage of R&D in the portfolio. The manufacturing is obviously up given the scaling that's occurring on the AI clusters.”

“On the -- I would just say the AI infrastructure, as I mentioned, we saw a stronger sense of customer urgency that manifested in stronger-than-expected bookings. Equally, we're pleased by the doubling of the number of customers in the AI space. So the ecosystem is broadening, and we're participating in it.”

“It's still for the wireline business, largely, things move quarter-by-quarter, but I think it's largely -- we still -- it's in the 70-30 range with both R&D, as I mentioned, in the AI space, both R&D and manufacturing doubling as we think about the first half business.”

“Second, the industry continues to navigate multiple overlapping speed transitions, with continued 800-gig deployments, accelerating adoption of 1.6-terabit architectures and increased R&D activity around 3.2-terabit technologies. The Optical Fiber Conference and NVIDIA's GTC this quarter reinforced the accelerating importance of networking as a critical enabler of AI data center scaling. At OFC, Keysight demonstrated our 1.6-terabit physical layer solutions with over 20 industry leaders. We also showcased 1.6-terabit traffic emulation, link reliability validation, and SerDes signal integrity solutions for switch and system vendors. And we collaborated with Broadcom on the industry's first public interoperability demonstration of Ultra Ethernet Consortium specifications, marking a major step towards production-ready AI-optimized Ethernet fabrics.”

“Third, activity in silicon photonics and co-packaged optics continues to expand. Our early engagements in co-packaged optics positions Keysight well to capture value as the industry transitions to these architectures. We're also seeing strong demand from next-generation optical component and transceiver development and deployment driven by expansion and scale-out networks. We recently expanded our optical portfolio with the industry's first 220 gigahertz Lightwave Component Analyzer to support advanced transceiver and photonics designs.”

“this heterogeneous environment really fits the kind of portfolio that we have, the breadth we have, we're able to participate in it broadening. And this is primarily all the stuff I talked about is in the physics of the AI infrastructure. We're equally excited by some of the announcements we made around emulating data center infrastructures, emulating inference infrastructure. That's still a very small part of the overall business, one we're getting considerable traction from customers.”

“This quarter, Keysight announced new scale-up validation solutions for performance characterization. As systems become more complex and expensive, additional investments in deeper manufacturing validation and production test coverage are needed to improve yields and reduce post-deployment failures. We saw strong adoption for newly introduced ultra-high-density interconnect solutions that enable rapid characterization of rack backplanes for next-generation scale-up networks.”

“Finally, customers need system-level emulation and benchmarking capabilities for data centers at scale. We saw strong adoption of our AI workload emulation solutions among hyperscalers as they work to improve utilization of GPU, power resources while addressing growing system and security complexity. This quarter, we expanded Keysight's AI portfolio with the release of Keysight AI Inference Builder designed to support emerging inference applications.”

“First, the scaling challenge is intensifying as AI clusters integrate GPUs, CPUs, DPUs, switches, NICs, memory fabrics, and storage across multiple vendors and the networking technologies, including Ethernet, UALink, PCIe, NVMe, and CXL. Customers are adopting Keysight solutions for end-to-end interoperability and system validation to ensure that these components function reliably together at scale.”

“AI ecosystem demand further accelerated this quarter across advanced node, memory, and silicon photonics. Our collaborations with leading foundries from R&D to production are enabling faster development and commercial ramp time lines for increasingly complex chip architectures and packaging.”

“for the AI business, there was a stronger sense of urgency from our customers to convert, which translates to a velocity in the pipeline where things or opportunities moved faster. But that's about it. There was no pull-forwards that if any -- that we can discern from the data.”

“we have a higher backlog in our AI business due to the strong demand in the first half.”

“we know one thing for sure is that there's going to be a few hundred gigawatts of capacity that is going to come online through 2030. And if you think of the big headline numbers, people talk about the multihundred billion dollars of spend commitments being made, it takes quite a bit of time for all of that to trickle through the ecosystem and the supply chain and manifest itself into demand that is actually implemented in a data center.”

“we're still in the very early innings in the overall AI landscape. And yet, we've all seen multiple turns as this market has moved. And I would just say, reflecting on what we have seen play out over the last couple of years even is the focus was only on training and models. And now you're moving from training to inference being the focus with the promise of agentic yet to come.”

“When we called out AI in 2023, it was just about the time of the ChatGPT moment and -- but we felt really good about the long-term opportunity. So we identified it, we invested in it, and we're excited by not only AI, but also the other opportunities that we laid out as part of this accelerating technology trends.”

“As we are focused on capitalizing on our early leadership in the AI data center infrastructure ecosystem, we're equally excited by the broader set of secular growth opportunities we're progressing, including defense technology, space, 6G, and quantum computing.”

“As I mentioned in our Q1 earnings call, this momentum continues to be driven by 4 key pillars of opportunity that we expect to continue. AI infrastructure scaling, speed transitions, optical and photonics technologies, and system-level emulations.”

“AI was obviously the strong theme. Equally, aerospace, defense and semiconductor were key contributors to that growth. And the growth came across all our businesses and across all of our sales regions, which we are very pleased by.”

“Orders saw robust growth in the quarter with activity in nonterrestrial networks, 6G research, and increased demand to support the supply chain associated with AI expansion.”

“In general electronics, double-digit order and revenue growth was driven by ongoing momentum in AI-related innovation and infrastructure investments.”

“at Mobile World Congress conducted a joint demonstration with Samsung on AI-RAN workflows.”