Do Broadcom's FY27/FY28 AI revenue targets rebut Monday's "cycle end" tape?

Opening

Operator multi-year AI revenue guidance has the better of overnight cycle-end takes. Hock Tan reaffirmed Broadcom's ~$115B FY27 / ~$230B FY28 AI semiconductor path, with Anthropic on track as largest custom silicon customer in 2027 — a harder datum than weekly slowdown essays. What remains open is delivery (secured supply, CSP concentration, GW ramps) and whether Monday's SOXX tape was pricing demand or rates/oil; deep tail unread (top-replies).

Sides

FOR — reaffirmed operator targets outweigh cycle-end narrative

• @jpinsights [named] — Hock Tan reaffirmed ~$115B FY27 / ~$230B FY28; Anthropic largest custom silicon customer in 2027; weight operators over cycle-timers
• @tonytonggg [named] — reaffirmed multi-year target from the person running the company is harder evidence than cycle-timing with nothing at stake
• @Little_K_Story [pseudo] — take Hock Tan's backlog numbers over another weekly end-of-cycle take

RELATED — bottom-up usage bills still rising

• @AbhishekYps_ [named] — solo founder AI spend up every quarter; more use cases as token prices fell, cost per task crashed

AGAINST — founders/operators can mislead; wait for firmer signals

• @Sean_Li0525 [anon] — almost all founders will lie (几乎所有创始人都会说慌)
• @TheWaitingGameX [pseudo] — start doubting the AI cycle when Jensen says something negative

Room: named semis voice citing a principal CEO outlook (Hock Tan [principal] [book] via Broadcom); anon founder-skepticism is stance, not evidence.

Quotes

Listen more to experts actually running these businesses and less to people trying to predict the end of the AI cycle every other week.

Hock Tan just reaffirmed Broadcom’s ~$115B AI semiconductor revenue target for FY27 and ~$230B for FY28.

Anthropic is also on track to become $AVGO’s largest custom silicon customer in 2027.

I know whose view I put more weight on. — JP Insights (@jpinsights)

@jpinsights A reaffirmed multi-year revenue target from the person running the company is a harder data point than a cycle-timing prediction from someone with nothing at stake in being right.

Weighting them equally treats operational commitment and punditry as the same kind of evidence. — Tony Tong (@tonytonggg)

@jpinsights You write about the boring stuff, so here's a boring data point: solo founder, my AI spend has gone up every quarter and never once down. no forecast in it, just a bill I keep paying. is that the kind of signal you weight, or too anecdotal? — Abhishek Sharma (@AbhishekYps_)

@jpinsights more use cases, not pricier work. token prices dropped, which made a dozen new daily tasks viable to automate. total bill grew, cost per task crashed. and yes 100% worth it, it pays for itself in raw shipping speed alone. — Abhishek Sharma (@AbhishekYps_)

@jpinsights I will start doubting the AI cycle when Jensen says something negative about it. It doesn't look like that is going to happen soon. — The Waiting Game (@TheWaitingGameX)

What would settle it

Broadcom fiscal reporting that AI semiconductor revenue is tracking toward ~$115B in FY27 (or an explicit cut to that path) Knowable — Broadcom FY27 quarterly prints / guidance updates — through 2026-12-15 earliest checkpoint

Anthropic (or Broadcom) disclosure confirming Anthropic as largest Broadcom XPU/custom silicon customer by 2027 revenue or GW deployed Knowable — earnings Q&A / customer commentary — through 2027

Hyperscaler capex and custom-ASIC commentary that either sustains or cuts the Broadcom GW/XPU pipeline named on the Q3 call Knowable — next major cloud earnings cycle — fall 2026