Nvidia ATH, Broadcom's $21.7B Guide, TSMC Preview: Oct. 4

After more than four months without a record close, Nvidia reclaimed its all-time high, and the broader AI hardware complex had a strong week: Broadcom posted a blowout AI chip guidance figure, TSMC heads into its Q3 print with expectations running high, and Micron’s financing picture is drawing fresh attention from analysts.

Key points

• Nvidia (NVDA) traded at an all-time high, capping a recovery of more than four months, according to Seeking Alpha.
• CEO Jensen Huang reaffirmed his projection that global data center capital expenditures will reach $3 trillion to $4 trillion annually by 2030, a significant step up from current spending levels.
• Broadcom (AVGO) guided Q4 AI semiconductor revenue to $21.7 billion, sending shares up roughly 3.5%.
• TSMC (TSM) is approaching its Q3 earnings with analysts anticipating strong results, bullish AI demand commentary, and guidance on capital expenditure plans, per a preview on Seeking Alpha.
• AMD (AMD) was upgraded to “Buy” by Wall Street Zen, with Wedbush maintaining its “Outperform” rating and a $600 price target.

What does Nvidia’s all-time high signal for the cycle?

Nvidia’s return to a record close is more than a technical milestone. The stock had spent over four months below its prior peak, a stretch that tested conviction among investors who had built positions on the AI infrastructure buildout thesis. The recovery, arriving alongside Huang’s reiteration of the $3-4 trillion data center capex outlook for 2030, suggests the market has renewed confidence in the durability of AI hardware demand rather than treating the earlier pullback as a structural break.

That projection is worth putting in context. Current annual data center capex across the major cloud providers runs in the hundreds of billions of dollars. A move to $3-4 trillion annually by 2030 would require a sustained, multi-year acceleration that touches nearly every part of the AI supply chain: chips, memory, networking, power, and cooling. Huang has floated this figure before, but reaffirming it now, as Nvidia sits at a new high, carries a different weight than when it was first offered as a forward-looking aspiration.

Broadcom’s $21.7B guide: how big is this number?

Broadcom’s Q4 AI semiconductor revenue guidance of $21.7 billion is a headline figure that deserves careful framing. The company supplies custom AI silicon (XPUs) and high-speed networking gear to hyperscalers, and its guidance visibility tends to be better than most in the sector because of the contract-driven nature of its custom chip business. A $21.7 billion quarterly run rate for AI chips alone, if sustained, would put Broadcom’s AI semiconductor segment on an annualized pace approaching $87 billion, a number that would have seemed implausible even two years ago.

Shares rose approximately 3.5% on the news. The market’s reaction was positive but not euphoric, which may reflect that sophisticated investors are already pricing in continued AI capex strength. The more interesting question going into Q4 will be whether Broadcom’s hyperscaler customers show any signs of digestion or pause after the aggressive build cycle of 2025 and 2026.

TSMC earnings and Micron financing: what to watch

TSMC reports Q3 results shortly, and the Seeking Alpha preview flags three things investors are focused on: the revenue and margin print itself, any forward commentary on AI-driven demand trends, and what management says about capital expenditure plans. TSMC sits at the center of the AI chip supply chain, manufacturing for Nvidia, AMD, Apple, and a growing roster of custom silicon customers. Its guidance language often sets the tone for the entire semiconductor sector.

Separately, a new analysis on Seeking Alpha highlights what it calls a “hidden financing advantage” for Micron (MU), pointing to blowout revenue, margins, and free cash flow driven by AI memory demand. Micron’s HBM (high-bandwidth memory) business has been a direct beneficiary of Nvidia’s GPU ramp, as each AI accelerator requires significant HBM capacity. The financing angle in the piece suggests Micron may have structural advantages in funding its next capital investment cycle, though the source is an opinion piece and investors should weigh that framing accordingly.

Power politics and safety debate enter the picture

Two less market-moving but worth-watching stories emerged this week. President Trump publicly defended Ohio data centers, framing AI infrastructure buildout as a national competitiveness issue while promising measures to address local community concerns about power costs. The political cover from the White House matters for permitting and utility rate negotiations that data center operators face at the state level.

On the safety front, a former OpenAI engineer warned publicly that major AI developers are not being careful enough, and called for safety frameworks comparable to nuclear power or aviation. This follows a pattern of AI safety criticism from ex-employees of frontier labs. For investors, the relevance is primarily regulatory: louder safety debates tend to increase the probability of legislative action that could affect the pace or terms of AI deployment, though no specific regulatory proposal is currently attached to these comments.

---

This article is for informational purposes only and does not constitute investment advice. Always do your own research before making any investment decision.