Nvidia (NVDA 1.45%) stock has soared by 870% since the start of 2023, catapulting its market capitalization to a whopping $3.5 trillion. Demand continues to exceed supply for the company's graphics processing units (GPUs) for the data center, which are the most powerful chips in the world for developing artificial intelligence (AI) models.
Nvidia CEO Jensen Huang says new AI reasoning models, which spend more time thinking to produce accurate responses, require up to 1,000 times more computing capacity than traditional large language models (LLMs). As a result, he thinks data center spending will reach $1 trillion per year by 2028 as companies race to meet that demand.
Nvidia is gearing up for the strongest year in its history thanks to its latest GPU architectures, which offer more performance than anything else on the market right now. Nvidia stock is trading at around $142 as of this writing (June 10), but here's why I think it's on the way to $200 before year-end.
Image source: Nvidia.
Nvidia is racing to launch new chips to meet demand
Nvidia's Hopper architecture was at the foundation of its H100 GPU, which was the top-selling AI data center chip in 2023 and for most of 2024. It was superseded late last year by the Blackwell architecture, with GPUs like the GB200 producing up to 40 times more performance than the H100 in specific configurations. This is ideal for inference workloads, which is the process by which AI models analyze data to produce responses for the end user.
Inference demand is soaring because of the reasoning models I mentioned, which include OpenAI's GPT-o3 and o4, Anthropic's Claude 4, Meta Platforms' Llama 4, DeepSeek's R2, and more.
Previous LLMs were great at generating one-shot responses that created a fast and convenient user experience, but they sometimes produced inaccurate information. Reasoning models refine their responses in the background before delivering them to the user, which takes longer and consumes significantly more tokens (words, punctuation, and symbols) -- between 100 and 1,000 times more, according to Jensen Huang.
Therefore, it's possible Blackwell's 40-fold performance leap over Hopper still isn't enough for some reasoning models, which could lead to a sluggish user experience. Nvidia recently launched Blackwell Ultra, which offers a slightly better 50-fold improvement over Hopper, but developers will have to wait until next year for the next major jump in performance.
That's because Nvidia will start shipping a lineup of GPUs based on its new Rubin architecture in 2026, which could be 3.3 times faster than Blackwell Ultra -- or around 165 times faster than Hopper. This constant chase for performance is why Huang thinks annual data center spending is heading toward $1 trillion in 2028, and the company with the most powerful chips is likely to capture the most market share.
Wall Street's consensus estimate (provided by Yahoo! Finance) suggests Nvidia could generate a record $200 billion in revenue during its fiscal year 2026 (which ends on Jan. 30, 2027). Based on the company's recent financial results, the data center segment will account for around 89% of that revenue, which is why AI and GPU sales are the main focus for investors right now.
Nvidia stock has a pathway to $200 this year
Nvidia generated $3.19 in earnings per share (EPS) over the last four quarters, which places its stock at a price-to-earnings (P/E) ratio of 46. That is comfortably below its 10-year average of 59.9, suggesting the stock might be undervalued right now.
Plus, Wall Street expects Nvidia to deliver up to $4.27 in EPS during fiscal 2026, placing its stock at a forward P/E ratio of 33.4:
NVDA PE Ratio data by YCharts
That means Nvidia stock would have to climb by 38% over the next nine months just to maintain its current P/E ratio of 46, and it would have to soar by 79% to trade in line with its 10-year average P/E ratio of 59.9. That would place Nvidia stock somewhere between $196 and $254 in the early stages of 2026.
However, the stock market is a forward-looking machine, so investors usually start pricing in future earnings ahead of time. That's why I predict Nvidia stock will surpass $200 before 2025 is over. If Jensen Huang provides a bullish update on next year's Rubin GPUs, Wall Street analysts might even raise their EPS estimates, which would be another upside catalyst for the stock.
But no matter what happens in the remainder of this year, investors should stay focused on the long term, because if data center operators spend $1 trillion annually on AI infrastructure by 2028, as Huang expects, $200 might actually be a bargain for Nvidia stock when we look back on this moment in the future.
