fJscaler Announces Sampling of Breakthrough 100 GHz BiCMOS TIA for Next-Generation AI and Cloud Datacenter Optical Interconnects

fJscaler began sampling its BiCMOS transimpedance amplifier (TIA) to strategic customers at ECOC 2026 for IM-DD optical interconnect applications. The product targets AI and cloud datacenters, with the company highlighting improved energy efficiency and sensitivity, though the announcement provides no commercial volumes, financial metrics, or customer commitments.

Analysis

This is an early design-in signal, not a revenue event. Sampling suggests the relevant read-through is to the AI optical-interconnect bottleneck: lower-power receiver components can improve rack-level power budgets and make higher-speed scale-up architectures more economical, supporting the migration from copper and lower-speed optical links toward 800G/1.6T deployments. The near-term public-market beneficiaries are likely established optical-component vendors with customer qualification infrastructure—COHR, LITE and AAOI—rather than a private component startup whose claimed performance remains unverified.

The second-order risk is margin pressure within the optical supply chain. If a BiCMOS-based TIA achieves comparable sensitivity at materially lower power or cost than incumbent III-V/SiGe solutions, module makers could gain sourcing leverage and commoditize discrete receiver electronics; that would be more negative for differentiated analog component suppliers than for vertically integrated transceiver vendors. Over 6-18 months, the key question is whether hyperscalers accept IM-DD at the targeted speeds and reaches, or favor silicon photonics/co-packaged optics, which would redirect value toward AVGO, MRVL and COHR.

No immediate trade is warranted absent disclosed data-rate, power-per-Gbps, qualification customer, pricing, and production timeline. The actionable catalyst window is 1-3 months around ECOC customer commentary and 2027 optical order guidance: evidence of accelerated 1.6T deployments would validate broad optical demand, while delayed qualification or a shift to co-packaged optics would weaken the discrete-TIA thesis. Falsify a long optical-equipment/component view if hyperscaler capex guidance decelerates or 1.6T shipment forecasts are pushed beyond 2027.

Key Decisions for Investors

• Maintain a watchlist long bias in COHR and LITE for the next 1-3 months, but do not initiate solely on this release; require independent confirmation of 1.6T module demand or raised datacom guidance. Upside comes from optical-content expansion, while the principal risk is pricing pressure from lower-cost receiver architectures.
• Use AVGO/MRVL versus a basket of discrete optical-component names as a 6-18 month structural pair-trade monitor: go long AVGO or MRVL / short a diversified optical basket only if customer roadmaps indicate co-packaged optics adoption ahead of discrete IM-DD upgrades. The trigger is hyperscaler architecture disclosure, not component sampling.
• Set alerts for AAOI, LITE and COHR earnings: track datacenter revenue growth, gross-margin commentary, 1.6T qualification timing, and any reference to TIA sourcing. A guidance increase tied to 800G/1.6T volumes is a more investable catalyst than a private-company performance claim.
• Avoid shorting incumbent optical suppliers on this news. Sampling-to-volume conversion commonly takes 12-24 months and qualification failure, thermal constraints, or inadequate yield could leave incumbent component economics unchanged.