Advanced Micro Devices (AMD -1.83%) has underperformed the broader semiconductor sector in the past year. Shares of the chipmaker have slid almost 20% during this period, but it looks like the company's fortunes could turn around in the coming year.

HSBC analyst Frank Lee has doubled his price target on AMD stock to $200, which points toward 37% gains from current levels. Lee says that AMD has managed to close in on Nvidia's (NVDA -0.46%) lead in the artificial intelligence (AI) accelerator market with its latest series of processors, and that could help the former record a big bump in its revenue next year.

Let's take a closer look at the reasons why HSBC has become so bullish on AMD stock, and check if AMD could indeed hit the $200 mark going forward.

Person in wheelchair sitting in front of two monitors and typing.

Image source: Getty Images.

AMD's data center revenue could jump significantly thanks to its latest chips

AMD launched its MI350X and MI355X data center graphics processing units (GPUs) last month. The company says that these new processors are three times more powerful than the previous MI300X accelerators, and they even outperform Nvidia's Blackwell GPUs in both AI training and inference workloads.

AMD has equipped these processors with a whopping 288 gigabytes (GB) of high-bandwidth memory (HBM), which is 1.6 times higher than Nvidia's GB200 and B200 AI graphics cards. The serious performance gains and competitiveness that AMD claims to have achieved with its latest GPUs are translating into more business for the company, according to HSBC.

AMD says that it is partnering with multiple AI giants such as Meta Platforms, Microsoft, Oracle, OpenAI, and xAI, among others, to build and deploy AI data centers powered by its processors. In fact, AMD claims that "seven of the 10 largest model builders and AI companies are running production workloads" using its AI accelerators.

This improving adoption of AMD's AI chips can be attributed to a big bump in computing power along with an improvement in energy efficiency, which enables its customers to reduce operating costs. Even better, AMD has reportedly priced its latest AI accelerators competitively so that it can eat into Nvidia's share.

According to HSBC analysts, AMD's MI355 AI GPU is 30% cheaper than Nvidia's B200. Not surprisingly, they are expecting the price-to-performance advantage of AMD's latest chips to translate into healthy data center revenue growth. The company is now expected to achieve $15 billion in AI GPU revenue in 2026, which is higher than the consensus estimate of $9.6 billion on Wall Street.

It's worth noting that AMD ended 2024 with an estimated data center GPU revenue of over $5 billion. So, AMD could potentially triple this business in just two years. Moreover, AMD management is forecasting its data center GPU revenue to reach an annual run rate of "tens of billions of dollars" in the long run, and it looks like its new chips could help it achieve that target.

The projected earnings growth points toward solid upside

Consensus estimates are projecting a 17% increase in AMD's earnings this year, followed by a much stronger jump of 47% in 2026 to $5.71 per share. That won't be surprising, considering the estimated jump of 3x in its data center GPU revenue. AMD has an additional catalyst in the form of the client CPU (central processing unit) market, where it is winning more share, apart from capitalizing on the secular growth of the AI personal computer market.

All this suggests that AMD could indeed hit Wall Street's consensus earnings target in 2026, though there's a chance that it may be able to do better than that. This improved earnings power could translate into more upside on the stock market. AMD is currently trading at 39 times forward earnings, which is a discount to the U.S. technology sector's average of 51.

Even if it trades at a discounted 35 times earnings and achieves $5.71 per share in earnings in 2026, its stock price could hit $200. So, investors can start accumulating this semiconductor stock, as it seems set for a turnaround thanks to its improving stature in AI chips.