In the 3rd quarter of January 2023 (August-October), the company's net sales decreased by 17% year-on-year to $5.931 billion. Non-GAAP adjusted EPS also decreased by 50% to $0.58. While net sales exceeded market expectations of $5.77 billion, adjusted EPS fell below market expectations of $0.69. Sales were impacted by GPU shipment constraints due to inventory adjustments, leading to a significant 51% decrease in the gaming division ($1.57 billion) impacting overall revenue. Also, the Professional Visualization division saw a 65% decrease in revenue ($0.2 billion) due to reduced sales from inventory adjustments. On the other hand, the Data Center division saw a 31% increase in revenue ($3.83 billion) reflecting strong performance from major U.S. cloud providers, and the AI self-driving solutions sales contributed to an 86% revenue increase for the autonomous driving division ($0.25 billion). Despite some offsetting from lower demand for data centers in China and warranty payments ($70 million), the gross margin decreased by 10.9 points due to inventory costs of $70.2 billion. Increased personnel and compensation costs from raises, as well as rising infrastructure costs in data centers, led to a 30% increase in adjusted operating expenses, resulting in a 55% decrease in adjusted operating profit ($1.536 billion). During the period, the company implemented a $3.75 billion shareholder return through share buybacks and cash dividends.
Guidance (midpoint) for the 4th quarter of January 2023 (November-January) indicates sales of $6 billion, up to ±2% year-on-year. Adjusted gross margin is expected to be ±50bp at 66%. Adjusted operating expenses are forecasted to be $1.78 billion (up 11.3%), and a single-digit increase in adjusted operating expenses is expected over the next few quarters. The company highlighted that growth engines continue to be data centers and autonomous driving, with a recovery in the gaming division, leading to a gradual growth trajectory.
The future of artificial intelligence (AI) seems to be a litmus test for the company's data centers as it moves towards further growth. Just as seen with the big hit of OpenAI's 'Chat GPT,' AI requires high-performance GPUs from the company to process vast amounts of data instantly. With already 500,000 NVIDIA GPUs implemented for 'Chat GPT,' following 'Chat GPT,' Google, Microsoft, and even Opera have announced AI chatbot deployments, bringing attention to the expanding demand for the company's AI GPUs. As we approach the performance announcement for the 4th quarter of January (November-January) on February 22nd (local time), it will be interesting to see if there are references to the outlook for AI-specific GPUs.