Nvidia (NVDA 6.18%) stock has been on fire over the past six months. Shares of the semiconductor giant have gained 131% in what may seem an unjustified rally given the challenges in the personal computer space and weakness in data centers, but there is a big reason why the market has been unconcerned about these red flags.

And now, the only Wall Street firm that had recently had a negative rating on the stock has finally joined the bulls. Frank Lee of banking and financial services giant HSBC was the last one among the 48 analysts covering Nvidia with a negative rating on the stock. But on Tuesday, he upgraded Nvidia to buy from reduce. Even better, he doubled his price target to $355 from $175, which would amount to a 28% jump from current levels.

Are Lee and other bullish analysts right to be so enthusiastic now? Let's take a look.

Nvidia enjoys terrific pricing power

Lee wrote in a research note that he is "shocked" by Nvidia's pricing power in artificial intelligence (AI) chips. HSBC's analysis reveals that AI chips command a heavy premium, selling for 10 to 20 times more than Nvidia's gaming chips. As a result, the chipmaker won't need to raise its sales volume as much as HSBC thought in order to drive meaningful growth in earnings and valuation.

What's more, HSBC adds that Nvidia is going to be the dominant player in the generative AI chip market this year with a 90% share, leaving peers like AMD and Intel behind. A closer look at the developments in the generative AI market will tell us just why HSBC has upgraded the stock.

Nvidia's graphics cards have been instrumental in powering the development of ChatGPT, the OpenAI and Microsoft-backed chatbot that has become hugely popular in recent months. Market research firm TrendForce estimates that ChatGPT could eventually need 30,000 Nvidia GPUs (graphics processing units). As TrendForce points out, each Nvidia A100 GPU that's capable of training AI models is priced between $10,000 and $15,000. Meanwhile, the Hopper H100 chip -- the A100's successor -- is up to 9 times more powerful in AI training and reportedly costs upward of $32,000.

For comparison, Nvidia's flagship RTX 4090 gaming graphics card is priced at $1,599, which explains why HSBC is focused on the fact that the company's AI hardware is way more expensive than its PC offerings. Given that HSBC expects Nvidia to dominate this market, the chipmaker could witness a solid bump in revenue and margins.

Nvidia revealed last month that demand for its H100 cards, which it claims are the most powerful AI GPUs, is growing rapidly thanks to the adoption of generative AI. Giants like Oracle, Amazon, Microsoft, and Meta Platforms are deploying H100 GPUs to power their generative AI infrastructure, and this is just the beginning as AI technology is currently in an early growth phase.

Grand View Research forecasts that the generative AI market will generate more than $109 billion in annual revenue by 2030, clocking a compound annual growth rate (CAGR) of 34% through the end of the decade. This should drive demand for chips capable of supporting AI infrastructure; Allied Market Research expects the demand for AI chips to increase at a CAGR of 37% through 2031.

Not surprisingly, Bank of America Global Research analyst Vivek Arya estimates that the adoption of generative AI could boost Nvidia's revenue at an annualized pace of 25% through 2027. Citigroup, meanwhile, estimates that ChatGPT alone could add between $3 billion and $11 billion to Nvidia's revenue over the next year. That would give the company's top line a sizable boost given that its trailing 12-month revenue stands at $27 billion.

Is the stock worth buying now?

However, new investors will have to pay a handsome premium to buy Nvidia as a play on the generative AI boom. The semiconductor stock trades at a whopping 159 times earnings following its recent rally. However, its forward earnings multiple of 61 reflects that Nvidia's bottom line is expected to improve significantly. More specifically, Nvidia's earnings could increase by 36% this year and 33% in the next fiscal year.

Additionally, if Nvidia does clock 25% annualized growth over the next five years as per BofA estimates, its revenue would increase to $82 billion at the end of the forecast period, as compared to $27 billion in the last fiscal year. At that forecast revenue level, if the company in five years is trading at its average price-to-sales ratio of 17, then its market capitalization would be almost $1.4 trillion.

Nvidia currently has a market cap of $683 billion, which means that it's reasonable to expect the stock to more than double based on current forecasts. So investors who are looking for growth stocks can consider buying Nvidia despite its rich multiple as AI-related catalysts could help it sustain its hot rally.