Futurist
Photonics, Space, AI Infrastructure & Robotics Commentary
Applied Optoelectronics and Bloom Energy show how the next phase of AI ties photonics to energy: more AI needs more bandwidth, more optical capacity and more power. Applied Optoelectronics is expanding optical capacity in Taiwan, and a Bloom Energy fuel-cell system is being deployed there to support the expansion and improve power reliability, with the project expected online in Q1 2027. x.com
Public markets follow private capital, which is pouring into AI, defense, robotics, photonics, space and AI energy. Matching names: NVIDIA, Vertiv, Arista and Nebius in AI infrastructure; Microsoft, Alphabet and Palantir in platforms; Kratos and AeroVironment in defense; Tesla, Ouster and Ambarella in robotics; Lumentum, Coherent and Applied Optoelectronics in photonics; Rocket Lab, AST SpaceMobile, Redwire and SpaceX in space; Vistra, Constellation, GE Vernova and Eaton in energy. x.com
Applied Optoelectronics' success depends on how fast it expands capacity, and the buildout looks to be going well after its large at-the-market raises. Per @carrioresearch, its Taiwan plant, which makes more than half of what it sells, will install Bloom Energy fuel cells to generate on-site power during the expansion, with the system running in Q1 2027. x.com
Optics and photonics should be among AI's biggest winners as faster GPU networks need optical connections that suppliers cannot expand quickly enough. Lumentum says some components are sold out through early 2029; Sivers is investing $30M to target more than 100M lasers a year, with operations expected in Q4 2027. Potential US restrictions on Chinese transceivers could reshape supply in 2027. Applied Optoelectronics and Sivers are the favorites. x.com x.com
Applied Optoelectronics looks attractive as Bloom Energy fuel cells for its Taiwan expansion could get transceiver capacity running sooner. Lumentum's CEO says the company is sold out through early 2029 despite a roughly 12x expansion of its main Japan factory; with new laser capacity taking three to five years to build, some unmet demand could flow to AAOI. x.com x.com
Power constraints can spread AI compute across sites, increasing the optical networking needed to connect them. Ciena's Gautam Billa puts power and infrastructure lead times at at least 18–24 months; a planned CtrlS deployment uses Ciena equipment to connect nine data centers. Cignal AI forecasts line systems rising from 29% of inter-site AI network spending in 2026 to 57% in 2030, putting more weight on that equipment in Ciena's and Nokia's opportunity. x.com x.com
Optics could be one of the biggest AI trades of 2027, the next layer of the buildout after memory led 2026. As AI clusters scale and networking moves from 800G to 1.6T and beyond, demand for optical infrastructure should accelerate. Applied Optoelectronics' earnings will set the tone. The watchlist spans optical modules, silicon photonics, AI networking, connectivity and AI fabric, from Lumentum, Coherent and Fabrinet to Marvell, Arista, Broadcom, Corning and Astera Labs. x.com x.com
Bloom Energy remains bullish as it is supplying fuel cells through local partner Leadray Energy for Applied Optoelectronics' Taiwan expansion. AOI's decision to bring power on-site to start transceiver capacity sooner illustrates demand for Bloom where grid access slows an optics buildout. x.com
Rising AI equipment demand could sustain the buildout behind Micron, Nebius, Marvell, Bloom Energy and Applied Optoelectronics. ASML says customers are raising short- and medium-term forecasts, some are interested in orders for 2029 or 2030, and its currently planned 2027-28 supply may not fully meet demand. The dip in these AI stocks might be a buying opportunity. x.com
Optics is the next big opportunity in AI hardware, with a 2-3 year runway of incremental demand as it spreads from scale-out beyond the data center to links between data centers from 2026 and inside servers from 2028. Lumentum is best positioned because current tightness is in lasers rather than transceivers. Coherent has a promising 6-inch expansion, though there is pushback on whether it adds to margins. Applied Optoelectronics expanded capacity most aggressively, so its bet is on execution. … Show more
Applied Optoelectronics is a domestic AI data-center bottleneck winner whose results show exponential operating leverage. EPS jumped 186% quarter over quarter and 138% year over year on an 86% revenue surge, as 800G transceiver economics improve with its Texas and Taiwan fabs reaching prime yields. With the US tightening restrictions on Chinese optical supply chains and the FCC banning new Chinese data-center hardware, hyperscalers are flooding it with high-margin orders. x.com
Electricity is a clear bottleneck for AI, favoring AI-focused power and data-center infrastructure names over residential utilities. The US could be short 107 GW to power data centers through 2029, while China adds capacity year after year. Medium-term preferences are Cipher, Hut 8, Riot, TeraWulf, Galaxy, Equinix, Bloom Energy, Solaris, Vistra, GE Vernova, INIO, Williams, Liberty Energy, EROC and Cummins. x.com x.com
