Josh Brown Says the Nvidia Dip Worth Buying May Not Be in Nvidia at All

AI & Compute

AMD’s Helios puts Nvidia’s AI buyer leverage to the test

Josh Brown argues that AMD’s Helios systems could give AI buyers leverage over Nvidia, but AMD’s 191-times trailing earnings multiple already prices in substantial share gains.

By
Nathan Brooks
Filed
Channel
AI & Compute
Read
4 min read

By Nathan Brooks

Nvidia generated $115.84 billion in fiscal third-quarter 2027 revenue, up 103.2% from a year earlier, as Data Center sales reached $107.60 billion. The result exceeded a $108.73 billion consensus. Non-GAAP earnings per share reached $2.68, above the $2.47 estimate.

That result did not settle the stock debate. On CNBC’s Halftime Report on November 17, 2026, shareholder Josh Brown said: “The dip worth buying may not actually be in Nvidia.” CNBC had introduced Brown, a program regular, as a prominent Nvidia shareholder one week earlier.

What changes in buyer leverage?

For years, buyers of AI accelerators had to court Nvidia to secure GPU allocations. Brown’s alternative is AMD’s Helios rack-scale systems, which reach some customers through Hewlett Packard Enterprise, or HPE.

He called this the third of Nvidia’s problems and the one most likely to interest viewers. AMD and HPE could benefit from disruption at Nvidia, he argued, because buyers with two qualified suppliers can negotiate more effectively.

His evidence came from Anthropic’s latest filing, as Brown described it. He said Anthropic brought up a Helios system using AMD chips through HPE in three days. 24/7 Wall St. did not independently verify his account of the filing.

Brown tied the claim to valuation. He argued that a threatened Nvidia supply grip since ChatGPT’s launch could explain its single-digit forward multiple, whether or not Jensen Huang returned. In his view, that development would give large customers more bargaining power even if Nvidia’s financial results remained strong.

What does Nvidia’s guidance show?

The forward numbers remained sharper still. Fourth-quarter guidance called for $130.0 billion, plus or minus 3%, with gross margin of 71% to 72%. The forecast assumes no Data Center compute revenue from China.

Interim CEO Colette Kress said Vera Rubin had begun strongly, the roadmap remained on track and Nvidia’s ambition had never been greater. The filing also flags risks from the leadership transition and an ongoing board investigation. Nvidia had hired an outside firm to search for Huang’s replacement.

Three direct customers accounted for 14%, 11% and 11% of revenue, according to the filing. That concentration increases large buyers’ negotiating power whenever a second qualified supplier exists.

The stock rose 1.8% after hours to $158.26. Even after that gain, it remained down 26.09% over the prior month and 14.94% year to date.

How far has AMD actually come?

AMD shares had gained 137.22% year to date. Its reported numbers and guidance supplied the stronger operating case:

  • Third-quarter revenue reached $13.38 billion, up 44.7% year over year.
  • Data Center revenue reached $8.55 billion, up 97%, while segment operating margin expanded to 34% from 25%.
  • Fourth-quarter revenue guidance was approximately $17 billion, plus or minus 3%, implying about 66% year-over-year growth.

According to AMD’s investor-relations materials, Anthropic, Meta, Microsoft, OpenAI and Oracle run workloads on the first Helios systems. CEO Lisa Su said: “We expect growth to accelerate further in the fourth quarter as we scale Helios and Venice production.” Her statement concerns planned production scaling, not confirmed shipment volume.

Brown’s three-day deployment claim adds a narrower, unverified point. It nevertheless fits AMD’s commercial pitch: Helios gives buyers a route around Nvidia’s allocation system, while HPE provides an established channel into some enterprise deployments.

Does the rotation case justify AMD’s valuation?

Valuation is where the rotation gets less tidy. Nvidia trades at 31 times trailing earnings with a 2.56% free cash flow yield. AMD trades at 191 times trailing earnings with a 0.81% yield.

AMD starts from a much smaller Data Center base, so each new gigawatt contract can move its results further than the same contract would move Nvidia’s. Investors are already paying a substantial premium for continued AMD share gains.

Brown’s diagnosis holds. Nvidia’s customer concentration, leadership uncertainty and the growing number of AI customers using AMD all point toward greater buyer leverage. Yet the stock-picking conclusion is weaker: spreading AI exposure across both companies makes more sense than selling one to buy the other. The procurement case is stronger than the stock case.

What should investors watch?

Three tests matter:

  • AMD must approach its roughly $17 billion fourth-quarter revenue guide as Helios and Venice production scale.
  • Nvidia must keep gross margin within its guided 71% to 72% range while leadership changes proceed.
  • More AI labs must report running production workloads on Helios rather than remaining pilot users.

The next test is straightforward: AMD must approach its roughly $17 billion revenue guide as Helios scales, while Nvidia holds gross margin at 71% to 72%. If both occur, Brown’s alternative becomes a more credible competitive hedge; if Helios adoption stalls, the procurement argument remains more solid than the rotation trade.

Source: 247wallst.com

Share this article:

More from Nathan Brooks

Nathan Brooks

Show full bio

Senior reporter covering people, launches and day-to-day practice in Semiconductors, chip manufacturing, hardware, AI compute at Chip Dispatch.

1,076 articles

Related articles

« Previous articleNext article »