Third-quarter 2026 earnings call, November 10, 2026: management cites slower silicon photonics foundries and process qualification cycles as the main hurdle
Volume production slips from the second half of 2027 to 2028
Lightwave Logic said it now expects volume production of its electro-optic polymer technology to begin in 2028, a delay from the second half of 2027 targeted on the second-quarter call. CEO Yves LeMaitre attributed the change to "longer foundry wafer production schedules and process qualification cycles alongside the remaining customer development and product qualification work." He said 2027 will be spent completing these activities and helping customers advance toward Stage 4 of the company's design win process.
LeMaitre was direct about the cause. "The main challenge we have been facing in 2026 is a slower pace of silicon photonics foundries, both in terms of production cycles and PDK process qualification," he said, calling this the company's main hurdle to faster introduction of silicon organic hybrid, or SOH, modulators incorporating its materials. He said the engineering team is "100% focused" on enabling foundry partners at both the front end and back end, and on supporting the transition of its 5 lead customers to commercial revenue.
Management tied the foundry slowdown to the broader optical supply chain. Shortages of indium phosphide lasers, particularly electro-absorption modulated lasers, are accelerating interest in silicon photonics architectures that use simpler continuous wave lasers, which LeMaitre said typically require fewer lasers. He said this shift validates the opportunity for electro-optic polymers. It also creates a near-term problem: foundry partners are prioritizing the immediate need for additional silicon photonics production capacity and managing their own capacity investment, which affects "the resources and schedules available to qualify new materials and processes." LeMaitre said these pressures have extended development timelines.
Foundry status: one usable run, one process failure, two still ahead
The company received wafers from 2 of its foundry partners in the third quarter. Chips from one, delivered on an expedited schedule, showed promising results and were used in the Foxconn demonstration. Additional wafers from that foundry are expected by the end of 2026, and the customer plans to launch another run in early 2027 targeting the final product configuration.
The second foundry partner fared worse. It "ran into process issues at one of their facilities, which translated into a limited number of usable chips." That partner is now expected to deliver 2 new sets of high-performance chips from 2 different manufacturing locations in the first half of 2027. Wafers from the third foundry partner are expected in December 2026, and the run with the fourth partner is tentatively scheduled for the first half of 2027. The company's foundry ecosystem includes GlobalFoundries, Tower Semiconductor and SilTerra, plus a fourth partner that remains unnamed. Management said each run feeds findings on device performance and yield back into the process design kit, or PDK, used for future customer designs.
Foxconn demonstration reaches system level, but qualification work remains
On October 29, Foxconn demonstrated an optical transceiver at Hon Hai Tech Day in Taipei using a silicon organic hybrid optical engine enabled by Lightwave Logic's polymers. The company said work with Foxconn began in the first half of 2025 and has produced a first-iteration optical engine capable of 200 gigabits per second per lane. LeMaitre described the significance as going "beyond the performance of individual materials or devices," showing the technology operating inside a functional optical engine and transceiver.
He was careful to define what the milestone does not mean. "A functioning transceiver demonstration is an important step towards product qualification," he said, adding that further chip design iterations, integration, reliability testing, manufacturing preparation and customer-specific qualification are still required before a program can reach Stage 4. Lightwave Logic considers Stage 4 reached when the customer's product is fully qualified, accepted by the end user and ready to ship in volume with consistently high yields.
LeMaitre also laid out the qualification path, which he said is a recurring question from shareholders. It typically begins with alpha devices or early optical engine prototypes on test platforms, followed by beta devices meeting the target specification for the customer's transceiver. Testing then occurs at both chip and assembled transceiver level, covering optical and electrical performance, mechanical and thermal requirements and, where applicable, software functionality. Reliability testing follows on both the photonic chip and finished product. This all assumes the silicon foundries have qualified their own equipment and processes and achieved predictable yields, usually signaled by a new PDK release that includes the steps needed to produce SOH modulators. In parallel, customer new product introduction teams must install and qualify equipment, train operators and validate assembly and test systems. The final step is end-user qualification, where a hyperscaler or AI infrastructure supplier tests samples in its own systems.
Beyond Foxconn and Marvell, the company has 3 other Fortune Global 500 customers in Stage 3, each on its own development and qualification path. LeMaitre noted that Marvell's recently announced agreement to acquire Lumentum reflects the push to combine semiconductor and optical capabilities, which he said supports the convergence of photonics and semiconductors.
Back-end processing to be outsourced
On manufacturing, the company continues to expand its Perkinamine production infrastructure in Denver, Colorado. LeMaitre said he is pleased with progress over the last few quarters. The back-end-of-line polymer processing, which includes depositing and poling the polymer and encapsulating devices, currently resides in Denver, but the company plans to outsource it. It is in advanced negotiation with a leading specialist in wafer-level optical manufacturing and kicked off a technical project with that firm a few months ago. Customers are focused on wafer-scale back-end processing with a path to high volume and 300-millimeter wafers.
On the second material supply and licensing agreement previously discussed, CFO Fred Graffam said only that it is "an ongoing and active negotiation" protected by nondisclosure agreements, and he declined to provide further detail.
Roadmap: 200G as a stepping stone to 400G and plasmonics beyond
Asked whether the delay reflects reduced R&D for next-generation 400G solutions amid pressure on AI infrastructure stocks, LeMaitre said "No, not at all." He said all programs continue with the same intensity and customers continue to invest toward 400G per lane across applications from scale-across to scale-up. Both 200G and 400G per lane are active priorities, with 200G viewed as a stepping stone and most technical activities overlapping, though 400G has more demanding bandwidth requirements. LeMaitre said activity is also increasing beyond 400G per lane, particularly in plasmonic organic hybrid solutions, where "new solutions might be required beyond SOH and where plasmonics might play an important role." The company's CTO and advanced technology group is leading that work.
Financials: widening loss, cash runway into 2028
Net sales were approximately $38,000 versus approximately $29,000 a year earlier. Graffam said revenue will be driven principally by technical and commercial milestones under existing development contracts, and the company held approximately $100,000 of deferred revenue to be recognized upon achieving them. Net loss was approximately $7.2 million, or $0.05 per share, compared with approximately $5.1 million, or $0.04 per share, a year earlier.
Research and development expense rose to approximately $4.3 million from approximately $2.9 million, and general and administrative expense rose to approximately $3.7 million from approximately $2.3 million. Graffam said the increases reflect investment in polymer development and production, device performance and reliability, foundry integration, prototype testing, intellectual property, targeted hires in engineering, manufacturing and commercial functions, and upgrades to IT, financial and operational systems.
The company held approximately $87.8 million in cash, cash equivalents and marketable securities at September 30, 2026, down from approximately $95.9 million at June 30. Cash used in operations was approximately $15 million for the first 9 months of 2026, versus approximately $10.4 million a year earlier, including approximately $5.2 million in the third quarter. Purchases of property and equipment were approximately $2.9 million for the 9 months, up from approximately $1 million, covering laboratory and manufacturing equipment, clean room and quality control capabilities. Graffam said the company has sufficient liquidity to fund its operating plan into 2028, which is the year volume production is now expected to begin. He said the revised timing "does not fundamentally change how we think about capital allocation" and that investments will be aligned with verified customer requirements and expected business benefits. He referred to the company's "current organic commercialization plan."