For the past six months, Nvidia (NASDAQ: NVDA) has done something unusual: it has gone nowhere.
After a staggering rally of more than 1,100% since early 2023, the AI chip leader has added just 1% over the last six months, settling into a persistent sideways trading pattern. The reason? Lingering investor concerns about whether artificial intelligence spending can maintain its torrid pace—and whether custom chips are starting to eat into Nvidia’s dominant market share.
But following the company’s recent GTC conference and its fourth-quarter earnings call, CEO Jensen Huang has delivered a forceful response. His core message: demand for AI computing is far from peaking, and Nvidia’s addressable market could triple or quadruple in the years ahead.
Huang addressed the sustainability question head-on during the February earnings call. “Compute demand is growing exponentially—the agentic AI inflection point has arrived,” he told analysts. The shift from simple probabilistic models like GPT-3 to complex reasoning systems like GPT-5 requires exponentially more computing power for both training and inference. That means more Nvidia GPUs.
More importantly, Huang laid out a two-stage roadmap for AI’s evolution. The first wave—agentic AI—is already here. The next wave, physical AI, will bring intelligence to robots, autonomous vehicles, and industrial systems. “That’s a giant opportunity,” he said.
His forecast is striking: by 2030, global data center spending will reach $3 trillion to $4 trillion annually. For context, the top five hyperscalers are expected to spend roughly $700 billion on capital expenditures in 2026, with total global spending around $1 trillion. That implies a market that triples or quadruples by the end of the decade, representing a compound annual growth rate of 32% to 41%.
According to McKinsey and Bernstein, GPUs and networking equipment account for more than 50% of data center spending. In both categories, Nvidia is the dominant supplier. The data center segment alone could represent a multi-trillion-dollar opportunity for the company.
When it comes to the threat from custom chips, Huang points to a metric competitors can’t easily dismiss: inference tokens per watt. For cloud providers, profitability depends on how many tokens they can process or generate for each watt of power consumed.
Nvidia’s advantage lies in its system-level approach. The company doesn’t just sell GPUs—it delivers complete data center solutions spanning GPUs, CPUs, and high-speed networking, all optimized to work together. That integration delivers the highest performance and energy efficiency in the industry. “Nvidia produces the lowest cost per token, and data centers running on Nvidia generate the highest revenues,” Huang said.
There’s also the software moat. Nvidia’s CUDA ecosystem and development tools are the industry standard, and any alternative chip requires developers to rebuild their software stacks from scratch. That switching cost is prohibitive. For the foreseeable future, Nvidia remains the default architecture for AI infrastructure.
The market has taken notice. Wedbush analyst Dan Ives described Nvidia’s backlog as “stunning,” adding that it will force a wholesale reset of earnings estimates. Of the 53 analysts covering Nvidia, 96% rate it a Buy. None recommend selling. The consensus price target implies roughly 50% upside from current levels, and the most bullish target of $400 suggests the stock could double.
Institutional investors have been accumulating shares for five consecutive quarters, with a buy-to-sell ratio of 3:1 in recent periods. That steady accumulation has built a solid floor under the stock, even as broader markets remain uncertain.
The six-month consolidation has formed a textbook bull flag on the chart. The flagpole ran from $90 to $180, followed by a sideways drift. A break above flag resistance near $196 would project a target of at least $270—roughly 50% upside. A stronger move could reach $360, aligning with the most optimistic analyst targets.
Valuation tells a similar story. Nvidia trades at about 22 times forward earnings, roughly in line with the S&P 500—a multiple that reflects none of the company’s superior growth profile. Based on 2030 earnings estimates, the P/E drops to 14. By 2035, it falls to just 9. Even assuming more modest long-term growth, current levels offer a substantial margin of safety. If earnings continue to surprise to the upside, the market will have no choice but to reprice the stock.
Six months of consolidation have allowed Nvidia to digest its historic run while building energy for the next move. Huang’s latest comments not only address concerns about AI spending sustainability—they point to a market several times larger than today’s. With analyst estimates rising, institutions buying, technicals tightening, and valuation reasonable, the pieces are in place for Nvidia to break out of its trading range.
The next catalyst could be first-quarter earnings in late May, or simply the market’s growing recognition of what’s ahead. Either way, for patient investors who have waited through the sideways grind, the reward may be close at hand.
