![\"Image](\"https:\/\/www.newsbeep.com\/il\/wp-content\/uploads\/2026\/07\/1784080999_983_.jpeg\"\/)
<\/p>\nThe artificial intelligence (AI) infrastructure build-out is still just getting started, and many of the stocks directly involved in it are just getting started, too. Most analysts and companies are expecting that the hot phase of this build-out may start winding down in 2030. Still, it could easily extend for years beyond that if the AI firms start seeing tangible returns on investment and AI technology provides enough value to warrant serious subscription fees.<\/p>\n
That could open the door for even more gains and make many stocks on the hardware side great investments — including these five.<\/p>\n
<\/p>\n
Image source: Getty Images.<\/p>\n
How much upside is left in the build-out?<\/p>\n
For 2026, the big four AI hyperscalers project that they will spend more than $650 billion on data center capital expenditures<\/a>. That doesn’t include the spending of other companies involved in AI, like OpenAI or Anthropic. It also fails to factor in what companies in China and other international markets are spending. Even if the total figure for AI capex globally is closer to $1 trillion right now, Nvidia (NVDA<\/a> +4.08%) expects far more.<\/p>\n Today’s Change<\/p>\n (4.08%) $8.31<\/p>\n Current Price<\/p>\n $211.84<\/p>\n Key Data Points<\/p>\n Market Cap<\/p>\n $4.9TMarket cap calculated using publicly traded shares outstanding only. Does not include unlisted, private, or dual-class non-traded shares. Implied market cap may vary.<\/p>\n Day’s Range<\/p>\n $203.80 – $212.53<\/p>\n 52wk Range<\/p>\n $164.07 – $236.54<\/p>\n Volume<\/p>\n 5M<\/p>\n Avg Vol<\/p>\n 158.7M<\/p>\n Gross Margin<\/p>\n 74.15%<\/p>\n Dividend Yield<\/p>\n 0.14%<\/p>\n Nvidia is a great voice to listen to in this area because it has the largest market share of AI computing units by far. It has repeatedly told investors that it expects global data center capital expenditures to reach $3 trillion to $4 trillion annually by 2030.<\/p>\n While that may sound far-fetched, the reality is that Nvidia’s position in the chip space gives it unusually clear insight into these trends. In the short term, Nvidia also expects AI hyperscalers to spend more than $1 trillion next year, a huge rise over this year’s figure.<\/p>\n That bodes well for most companies involved in the space.<\/p>\n The five will have a fantastic run over the next few years<\/p>\n Nvidia easily tops the list of AI stocks worth buying and holding, as it has a great future and has already established itself as the AI leader. The market isn’t pricing a ton of hoped-for growth into its stock yet, either. It trades at just 23.5 times forward earnings. <\/p>\n NVDA PE Ratio (Forward)<\/a> data by YCharts.<\/a><\/p>\n Next up, I think Broadcom (AVGO<\/a> +1.37%) is a great pick, as it’s tackling a different segment of the computing market. Instead of making GPUs like Nvidia, which can provide massive parallel processing power for a variety of computing applications, Broadcom partners with AI hyperscalers to develop custom AI chips called application-specific integrated circuits. As the name implies, they’re built to handle only specific types of workloads, but in their narrow specialized areas, they outperform GPUs on a cost basis.\u00a0<\/p>\n Broadcom<\/a> expects its AI semiconductor revenue to top $100 billion in 2027, which will be huge growth. <\/p>\n Nvidia and Broadcom only design the chips; they don’t manufacture them. That work is mostly done by Taiwan Semiconductor<\/a> (TSM<\/a> 0.24%), the world’s top chip fabricator. As long as AI spending increases, TSMC will remain an excellent stock pick. It’s also building out more production capacity for chips, because it sees greater future demand.<\/p>\n Taiwan Semiconductor Manufacturing<\/p>\n Today’s Change<\/p>\n (-0.24%) $-1.00<\/p>\n Current Price<\/p>\n $420.58<\/p>\n Key Data Points<\/p>\n Market Cap<\/p>\n $2.2TMarket cap calculated using publicly traded shares outstanding only. Does not include unlisted, private, or dual-class non-traded shares. Implied market cap may vary.<\/p>\n Day’s Range<\/p>\n $418.88 – $431.36<\/p>\n 52wk Range<\/p>\n $223.70 – $479.00<\/p>\n Volume<\/p>\n 494.5K<\/p>\n Avg Vol<\/p>\n 14.1M<\/p>\n Gross Margin<\/p>\n 60.72%<\/p>\n Dividend Yield<\/p>\n 0.83%<\/p>\n Micron (MU<\/a> +5.07%) and Sandisk (SNDK<\/a> +5.01%) design and manufacture memory chips, another important component of AI computing systems. While producers like TSMC have largely been able to meet demand for logic chips, the handful of players in the memory industry have not. As a result, there has been a serious supply crunch.<\/p>\n![\"Nvidia](\"https:\/\/www.newsbeep.com\/il\/wp-content\/uploads\/2026\/07\/1784081000_820_.png\"\/)
<\/p>\n![\"NVDA](\"https:\/\/www.newsbeep.com\/il\/wp-content\/uploads\/2026\/07\/d04fbdaec433f291a99e8a2883fdecae.png\"\/)
<\/a><\/p>\n![\"Taiwan](\"https:\/\/www.newsbeep.com\/il\/wp-content\/uploads\/2026\/07\/1784081001_580_.png\"\/)
<\/p>\n