aiqbench takeChips & Infra

This partnership addresses a critical bottleneck in AI infrastructure by scaling SiGe capacity for high-speed analog front-end components essential to optical and copper interconnects. As data center architectures evolve toward higher bandwidth requirements, power-efficient signal integrity solutions become the primary constraint on accelerator performance rather than raw compute speed. The expansion ensures dedicated manufacturing resources for next-generation transimpedance amplifiers and drivers required to sustain the growth of AI clusters.

Source coveringTower Semiconductor

What this isA model's reading of the article and related coverage, written for aiqbench. Not the publisher's words.

Tower Semiconductor and Semtech Expand SiGe Partnership to Scale Power-Efficient Analog Connectivity for Data Centers

Tower SemiconductorOctober 13, 20267 min read

Home » Press Releases Tower Semiconductor and Semtech Expand SiGe Partnership to Scale Power-Efficient Analog Connectivity for Data Centers Multi-year agreement expands dedicated capacity across Tower’s global fab network to support the companies’ growth in optical and copper interconnects for next-generation AI infrastructure MIGDAL HAEMEK, Israel, and Camarillo, California, Oct. 13, 2026 – Tower Semiconductor (Nasdaq/TASE: […]

MIGDAL HAEMEK, Israel, and Camarillo, California, Oct. 13, 2026 – Tower Semiconductor (Nasdaq/TASE: TSEM), the leading foundry for high-value analog semiconductor solutions, and Semtech Corporation (Nasdaq: SMTC), a leading provider of high-performance semiconductors powering AI data center networking and intelligent, connected Internet of Things (“IoT”) devices worldwide, today announced an expanded multi-year agreement that builds on the companies’ 20-year partnership. Under the agreement, Semtech will invest in securing dedicated silicon germanium (SiGe) capacity at Tower in support of its strong design wins and demand for AI Data Center optical interconnect components. AI’s next constraint is power, and a growing share of that power is spent moving data rather than computing it. As AI clusters scale across racks, rows and campuses, the energy cost of moving each bit is becoming central to how much useful work each accelerator can…

Excerpt shown under the publisher's RSS terms · 138 words · full text and figures are on the publisher's site

Read at Tower Semiconductor ↗Ask about this article