This year's increase of over 200% still has "cost-effectiveness"! Citigroup is optimistic about Nvidia's monopoly on AI chips, and a stock price of $600 is not a dream.

As the biggest beneficiary of this AI boom, NVIDIA's stock price has doubled this year. However, some analysts believe that the demand for GPUs is still high, and there is still significant room for growth in NVIDIA's stock price.

On Monday, NVIDIA's stock price rose by 2.18%. Citigroup reiterated its "buy" rating on NVIDIA in its research report and raised the target price from $420 to $520.

Citigroup analyst Atif Malik said, "This forecast is based on NVIDIA having 'multiple' data center orders worth over $2 billion and the continued rise in demand. He predicts that NVIDIA's P/E ratio will reach 35 times."

Under the assumption of a "bull market" with a P/E ratio of 40 times, he expects NVIDIA's target price to reach $600.

Citigroup stated that there are three key driving factors for NVIDIA's stock model:

1. Continuous growth in GPU demand.
2. A deeper understanding of custom ASICs and competition with AMD GPUs in data centers.
3. Reports released by multiple departments within Citigroup over the weekend regarding the impact of generative AI.

Due to its expertise in GPUs, NVIDIA has become the undisputed leader in AI chips, and its stock price has risen by over 200% this year.

However, competitors like AMD are also trying to seize this multi-billion-dollar opportunity by offering chips to customers seeking potential alternatives.

According to Citigroup, based on NVIDIA's advantages in optimizing GPU computing software, network product portfolio, and excellent hardware, NVIDIA's dominant position will continue.

Malik believes that NVIDIA has a "substantial advantage over AMD in terms of AI performance," which will help NVIDIA maintain a market share of around 90% in the AI accelerator market, which is expected to reach $150 billion by 2027:

It is clear to us that the AI accelerator market will grow at an astonishing rate. And NVIDIA seems prepared to continue to dominate the competition as it has been leading this market.

Meta Platforms' PyTorch 2.0 machine learning framework can help AMD narrow the gap between its software and NVIDIA's, but NVIDIA has accumulated years of experience in GPU optimization through its software framework CUDA and neural network library cuDNN. It may take several generations of GPU and software improvements from competitors to match NVIDIA.

We remain optimistic about NVIDIA's year-over-year growth in data center sales this year. Malik raised his earnings per share expectations for NVIDIA's fiscal years 2024, 2025, and 2026 by 6%, 38%, and 30% respectively.

Currently, analysts generally expect NVIDIA's earnings per share for the last quarter to be $2.06 and revenue to be $11.04 billion.