NVIDIA (NVDA)
The undisputed king of the AI chips data centres need to train and run their models.
Is NVIDIA a good stock for a UK beginner?
The honest version: The undisputed king of the AI chips data centres need to train and run their models.
Over about 2 years to 2026-07-31. This is the share price only - reinvesting the dividends would add to it. And it's the USD return - as a UK investor your actual £ return also moves with the exchange rate. Past performance is not a guide to the future, and it could just as easily have fallen.
Based on beta - how much the price swings versus the whole market. Bumpier isn't bad; it just means a rougher ride, which matters more the sooner you might need the money.
Prices move on results, news and the mood of the whole market - no single headline explains a day, and a quiet week is usually just noise, not a signal.
AI adoption broadens across industries and the company sustains its technology lead.
AI spending proves cyclical or overbuilt, or competition erodes pricing power over time.
What does NVIDIA do?
Nvidia designs the graphics chips (GPUs) and specialised AI accelerator chips that data centres lean on to train and run AI models, plus gaming and other computing. It's a money machine right now: a 63% net margin: How much of each £1 of sales becomes profit after all costs. Higher = more profitable per sale. (it keeps 63p of every £1 of sales as profit), a 114% return on equity: How much profit the company makes for each £1 shareholders have put in. Higher usually means a more efficient business., and 85% revenue growth: How fast the company's sales grew versus a year ago. as everyone races to build AI kit. The one thing worth watching -> it's priced like a star, with a P/E: Price-to-earnings: the share price divided by yearly profit per share. Lower can mean cheaper; higher often means investors expect fast growth. of 32.3 (about 16.5 on next year's forecasts) that's elevated versus the market, sky-high growth and quality scores (G95, Q95) but a low value score (V33) - the classic premium-priced grower.
On our factor screen it looks strongest on growth and quality, and weakest on value.
- ✓Pays a dividend - about 0.5% a year
- ✓Growing - revenue up about 85% over the year
- ✓Very profitable - turns about 63% of sales into profit
- ✓Low debt - a sturdier balance sheet
- ✓Strong return on shareholder money (ROE 114%)
- Quality screens high (93/100)
- Growth screens high (93/100)
- Very high net margin and revenue growth versus most large companies
- Forward P/E well below the trailing P/E, implying expected further earnings growth
- Leading position in a fast-growing AI accelerator chip category
- Customer concentration among a small number of large cloud/AI companies
- Potential slowdown in AI infrastructure capital spending
- Rising competition from other chip designers and customers' in-house chip efforts
What do NVIDIA's numbers mean?
How much money does NVIDIA make?
Revenue and profit by quarter, and how much of each sale turns into profit.
Does NVIDIA pay a dividend?
Yes - NVIDIA currently pays a dividend of about 0.5% a year (the yearly payout as a share of the price). A dividend is a slice of profit handed to shareholders; the yield moves as the price moves, and a company can cut or stop it.
When does NVIDIA report earnings, and how did recent quarters go?
NVIDIA is next scheduled to report on about 2026-08-26 - dates can move, and we don't predict results; this just tells you when to look.
Each quarter a company reports its results against what analysts expected. ‘Beating’ or ‘missing’ is about that expectation, not whether the business is doing well in absolute terms.
| Reported | Expected EPS: The earnings per share analysts expected for the quarter - the average of their forecasts. 'Beating' or 'missing' is measured against this number. | Actual EPS | vs expected |
|---|---|---|---|
| 2026-05-20 | $1.77 | $1.87 | Beat +6% |
| 2026-02-25 | $1.54 | $1.62 | Beat +5% |
| 2025-11-19 | $1.26 | $1.30 | Beat +3% |
| 2025-08-27 | $1.01 | $1.05 | Beat +4% |
| 2025-05-28 | $0.75 | $0.81 | Beat +8% |
| 2025-02-26 | $0.85 | $0.89 | Beat +5% |
Across the last 6 quarters here, NVIDIA came in ahead of what analysts expected 6 times. One quarter is noise, not a trend.
See who else reports over the next two weeks →
Reported vs expected earnings per share (EPS) from published results; the expectation is the analyst consensus, not our view. Report dates are estimates that can move.
More in Technology
Micron Technology
Western Digital
Gen Digital
Dell Technologies
Teradyne
Fidelity National Information Services
Hewlett Packard Enterprise
NetApp, Inc.What are the scenarios for NVIDIA?
An illustrative range for the year ahead, with the assumption behind each case — not a prediction or a price target.
What are the pros and cons of NVIDIA?
How many points the write-up makes each way — a balance check, not a score or verdict.
- Very high net margin and revenue growth versus most large companies
- Forward P/E well below the trailing P/E, implying expected further earnings growth
- Leading position in a fast-growing AI accelerator chip category
- High quality and growth factor scores (Q95, G95)
- Valuation remains elevated versus the broader market on a trailing basis
- Low value-factor score (V33) suggests pricing for continued strong growth rather than a bargain screen
- Very low dividend yield limits any income component
- Heavy reliance on continued AI capital-spending growth
- Customer concentration among a small number of large cloud/AI companies
- Potential slowdown in AI infrastructure capital spending
- Rising competition from other chip designers and customers' in-house chip efforts
- Export-control and geopolitical restrictions affecting chip sales to some markets
- High valuation leaves less room for disappointment
The write-up's own warning lights — if these start happening, the case above changes.
- A meaningful deceleration in data-centre/AI revenue growth relative to current forward estimates
- Margin compression from pricing competition or rising input costs
- Evidence of major customers materially cutting AI infrastructure budgets
Built from public filings & market prices, checked for an education-only tone, and never a price target. How we make these →