Texas Instruments

TXN automotive segment revenue fell again and management still won't give a clear timeline for when EV-related design wins actually ramp into meaningful revenue. The content-per-vehicle story sounds great on investor days but I need to see it show up in the numbers. Getting tired of the wait.

People are so focused on the near-term revenue weakness that they're missing the fact that TXN is deliberately building out capacity during a downturn so they don't have to pay peak prices later. This is exactly what Haviv said the 300mm roadmap is for — counter-cyclical capex so they can serve customers when competitors are scrambling for wafers.

How do people think about the risk that TXN's heavy internal fab investment becomes a liability if the industry moves further toward chiplet architectures or if demand for traditional analog just grows slower than expected? At some point all that 300mm capacity has to be filled.

The dividend growth track record at TXN is one of the best in semiconductors — 20 consecutive years of increases and they've been explicit about prioritizing it even through the capex buildout. For anyone running a dividend growth portfolio, the entry point here after the revenue-driven selloff looks genuinely attractive with a yield approaching 3%.

Does anyone have a good handle on how TXN's embedded processing segment competes with STMicroelectronics and NXP in automotive microcontrollers specifically? It feels like NXP has been winning more of the ADAS socket designs and I'm wondering if TI is ceding that ground intentionally to focus on analog.

Texas Instruments received CHIPS Act preliminary terms for potential funding related to its fab expansions in Sherman Texas and other US sites. The Sherman facility, LFAB2, is projected to be one of the largest 300mm analog fabs in the world when fully operational. This is part of TI's broader plan to have internal fabs supply the vast majority of its chips by the end of the decade.

One thing that doesn't get discussed enough is TXN's TI.com direct sales model — they've been pushing customers to buy direct rather than through distribution, which compresses short-term revenue visibility but improves margins and customer data long-term. It's a structural change to how analog semiconductors get sold and it's actually pretty unique in the industry.

Texas Instruments reported Q3 2024 revenue of approximately $4.15 billion, down roughly 8% year over year, with the industrial segment continuing to be the primary drag. Management guided Q4 revenue between $3.7 billion and $4.0 billion, implying the downturn hasn't fully bottomed. Free cash flow per share on a trailing basis came in well below the levels they used to justify the current dividend payout.

TXN trades at over 30x forward earnings while revenue is declining and capex is elevated — that's a lot of faith in a cycle recovery that keeps getting pushed out. Renesas and Infineon are cheaper and arguably better positioned in automotive. The valuation premium feels like legacy brand worship at this point.

TXN's 300mm analog fab strategy is genuinely underappreciated. While everyone else is outsourcing to TSMC, they're sitting on internal capacity that gives them structural cost advantages that widen every year. When the industrial and auto cycle turns, the operating leverage here is going to be violent to the upside.