NVDA vs AVGO — which stock performed better?

NVIDIA Corporation (NVDA) versus Broadcom Inc. (AVGO), Technology, on 2026-07-06.

Result

NVIDIA Corporation (NVDA) performed better on 2026-07-06. NVDA moved 0.550%, AVGO moved 0.520%.

Why these two companies

Nvidia and Broadcom are the two dominant forces in AI semiconductor infrastructure, but they are pursuing very different strategies: Nvidia sells the world's most powerful general-purpose GPUs that every hyperscaler is racing to acquire, while Broadcom designs custom AI chips (XPUs) tailored specifically for individual hyperscalers like Google and OpenAI. The debate is whether the GPU standard will remain entrenched or whether custom silicon will gradually displace it — and both companies are printing record revenues right now as the AI buildout accelerates.

The research

Nvidia and Broadcom represent two distinct but complementary bets on the AI infrastructure supercycle. Nvidia's dominance is built on its CUDA software ecosystem and the H100/H200/Blackwell GPU architecture — a combination so deeply embedded in AI training workflows that switching costs are enormous. Broadcom's strength lies in custom silicon: designing Application-Specific Integrated Circuits (ASICs) for hyperscalers who want chips optimised for their specific AI workloads at a lower cost per inference than general-purpose GPUs. Both companies are growing at extraordinary rates, but the market is asking whether Nvidia's lead is sustainable as more customers build their own chips with Broadcom's help.

Nvidia's Q1 FY2027 results (quarter ended April 2026) were staggering by any measure. Revenue hit $81.6 billion, up 85% year-over-year, with Data Center revenue surging 92% to $75.2 billion — now representing over 90% of total company revenue. GAAP EPS came in at $1.87, and the company approved a $50 billion share buyback. The Blackwell GPU architecture is ramping at record pace, with CEO Jensen Huang describing demand as 'insane.' Nvidia's profit margin of 55.6% and return on equity of over 100% reflect a company with near-monopoly pricing power in its core market. The Street has a consensus Strong Buy with an average price target of approximately $268, implying meaningful upside from current levels near $170.

Broadcom's story in 2026 is one of strategic expansion. The company's TTM revenue of approximately $75 billion grew 48% year-over-year, driven by three engines: custom AI XPUs (Application-Specific Integrated Circuits), VMware infrastructure software (growing at 43% post-acquisition), and networking chips that connect GPU clusters inside data centres. The landmark news in late June 2026 was OpenAI's selection of Broadcom as the partner to build 'Jalapeño' — OpenAI's first custom inference chip — joining Google (TPUs), Apple, and SpaceX as hyperscalers building bespoke silicon with Broadcom. Analysts project Broadcom's FY2027 revenue at $172 billion, a 62% jump from current levels, as the custom silicon pipeline fills. The consensus is Strong Buy with an average price target near $524, representing approximately 45% upside from the current price near $347.

The competitive tension between the two companies is nuanced. Broadcom does not directly compete with Nvidia on general-purpose GPUs — its XPUs are designed for inference and specific training tasks, not the broad flexibility of Nvidia's platform. However, every dollar a hyperscaler spends on a Broadcom custom chip is a dollar not spent on Nvidia GPUs, and the custom silicon market is growing faster than any single customer can absorb. Nvidia's moat is its software ecosystem: CUDA, cuDNN, and the broader developer toolchain mean that AI researchers default to Nvidia hardware. Broadcom's moat is its relationships: once a hyperscaler commits to a custom chip design cycle (typically 18–24 months), the switching costs are enormous. Both moats are real; the question is which is more durable.

The risk profiles differ meaningfully. Nvidia's primary risk is concentration: 90%+ of revenue from a single segment (Data Centre) and significant exposure to export restrictions on China sales, which have already cost the company billions in revenue. Broadcom's primary risk is customer concentration: Google, Apple, and now OpenAI represent a large share of its custom silicon revenue, and any decision by a hyperscaler to bring chip design fully in-house (as Google has partially done) could reduce Broadcom's addressable market. Broadcom also carries significant debt from its VMware acquisition (debt/equity of 166%), though its $27 billion in annual free cash flow provides ample coverage. Both companies report Q2 earnings in late July/early August, and the Street will be watching Nvidia's Blackwell ramp and Broadcom's XPU order pipeline closely.