As of Jul 12, 2026, the NVDA thesis is intact but worth watching: 3 pillars tracked, 1 on watch.
The reasons to own NVDA, the single fact that would break each one, and the dated filing and news evidence Helm has tested against them. Status is computed from that evidence, not hand-set.
Watch your own NVDA thesis →Hyperscaler AI capex keeps growing and NVIDIA keeps the majority of accelerator spend
This development suggests a potential shift away from NVIDIA's silicon by hyperscalers, which contradicts the pillar.
This reason is watch · 0 of 3 reasons to own NVDA hold as of 2026-07-12.
Watch my tickers
Get an email when something changes on your tickers: the exact quote, dated and sourced. No account, no brokerage connection, unsubscribe anytime.
Breaks if: A major hyperscaler guides data-center capex flat or down, or shifts away from NVIDIA silicon
“Google could lower its AI compute costs by up to 30% using its own processors.”
This development suggests a potential shift away from NVIDIA's silicon by hyperscalers, which contradicts the pillar.
Data-center segment revenue declines sequentially or guidance comes in below consensus
Research, not investment advice. Helm surfaces the evidence; you decide. This page tracks what to watch on the thesis, not whether to buy or sell.
Helm watches this thesis in public. It watches yours in private.
Watch your own NVDA thesis →No filing or news has tested this thesis since monitoring began.
Breaks if: Gross margin compresses materially toward the 60s on competition or supply costs
No filing or news has tested this thesis since monitoring began.