NVIDIA: The Next AI Chip Battle May Be Bigger Than Earnings

NVIDIA’s upcoming earnings are being judged by a much tougher standard than simply beating estimates. The market already expects strong numbers. The real question is whether the next phase of AI infrastructure growth can justify a new valuation cycle.

For several quarters, NVIDIA has delivered results above expectations, yet the stock has struggled to sustain the upside afterward. That tells us something important: earnings growth is no longer enough by itself. Investors are now looking beyond the current quarter toward Rubin, margins, competition, financing structures and the durability of AI demand.

Morgan Stanley’s outlook remains constructive. Its estimates point to revenue of roughly $91.2B for the July quarter and $102.3B for October, while the bigger upside is expected to emerge as Rubin scales into 2027.

Rubin Could Change the Game

Rubin is becoming the key strategic variable.

The transition from Blackwell to Rubin does not currently appear to be creating the kind of demand gap seen in previous product cycles. Blackwell demand remains extremely strong, while inference workloads are expanding beyond frontier AI labs into enterprises and sovereign customers.

That matters because the AI infrastructure market is broadening. Demand is no longer concentrated in a small group of hyperscalers. More companies and governments are building dedicated AI capacity, potentially extending NVIDIA’s growth runway well beyond the current generation.

The Margin Debate Is Critical

One of the biggest risks is not revenue — it is profitability.

Higher DRAM, wafer, packaging and substrate costs could keep pressure on gross margins. The market expects eventual margin recovery, but Morgan Stanley is more cautious.

This creates an interesting setup: NVIDIA could generate significantly more revenue than currently expected while earning lower margins than consensus assumes.

In other words, the next debate may shift from “Can NVIDIA grow?” to “How efficiently can NVIDIA monetize that growth?”

Competition Is Still the Bigger Long-Term Question

AMD, ASIC designers and hyperscalers developing custom chips remain important threats. But Rubin’s improved AI-factory economics could strengthen NVIDIA’s position if the product performs as expected.

The bigger issue is whether competitors can take meaningful share once supply constraints ease.

Right now, the evidence is still incomplete. Rubin needs several quarters of real-world deployment before the market can properly judge its impact on ASICs and AMD.

Valuation Creates an Interesting Asymmetry

Morgan Stanley’s scenarios show a wide range:

Base case: $288
Bull case: $330
Bear case: $160

The spread highlights how dependent NVIDIA’s future valuation is on AI demand, margins, competition and execution.

Yet NVIDIA is no longer trading with the extreme premium investors once associated with its dominance. That makes the risk-reward profile more interesting — provided AI infrastructure spending remains durable.

The biggest takeaway is simple:

NVIDIA’s next chapter may not be about another earnings beat.

It may be about whether Rubin proves that the AI semiconductor cycle still has multiple years of structural growth ahead.

If Rubin delivers at scale, NVIDIA’s competitive moat could widen again. If margins deteriorate and custom chips accelerate, the market may finally begin pricing a different AI-chip hierarchy.

The earnings number matters.

But the real signal will be what management says about Rubin, margins, demand, competition and 2027 growth.
#我的七夕交易分享
@Gate_Square
$NVDA