Nvidia's data center business segment demonstrates renewed strong growth momentum.

NVIDIA reported renewed growth last week, with its Data Center segment growing 43% to a record $968M in the last quarter, while gaming rose 56% to $1.49 billion. The market responded positively, sending NVIDIA stock up over 6%, with many analysts revising their price targets higher. This turnaround indicates that demand for AI hardware did not meaningfully taper off as some had feared; the slowdown experienced the previous year was painful to watch, but the large customers had purchased substantial capacity and needed time to digest it before resuming their growth in AI hardware. NVIDIA further reduced demand for additional hardware by increasing the performance of its V100 GPU by a factor of four over the last year through software optimizations.

The market demand for training, which was the primary growth driver for NVIDIA AI chips, has begun to shift toward acceleration for complex inference workloads, such as those found in "conversational AI"—where spoken queries feed neural networks to produce spoken and even translated answers. NVIDIA's T4 inference processor is now deployed by many cloud service providers, including Amazon and Google, both of whom have developed their own inference chips but still rely on NVIDIA for the heavy lifting, likely due to the breadth and depth of NVIDIA's software and ecosystem.

The size and complexity of deep neural networks themselves are doubling every 3.5 months, according to open.ai.org, which furthers the need for training and inference acceleration. Many newer applications, such as conversational AI, actually require multiple neural networks to work together to produce accurate results. Intel's recent acquisition of Habana Labs and the UK startup Graphcore's specific focus on designing solutions for these networks of networks both indicate the industry-wide need for inference acceleration for complex models. The computation required by AI applications will continue to grow dramatically.

Since Intel and most AI startups are only just beginning to ship their new AI silicon, these trends almost exclusively benefit NVIDIA through most of 2020. Many startups and their investors can now breathe a sigh of relief, as the growth assumptions in their business cases no longer seem so outlandish given NVIDIA's results. Looking forward, NVIDIA is expected to announce its next generation GPU later in 2020, likely at GTC or by SuperComputing 2020 in November. NVIDIA CFO Colette Kress guided the street to expect approximately $3B in total revenue in the next quarter, roughly flat from Q4, despite an expected $100M negative impact from the coronavirus. While NVIDIA currently enjoys clear sailing, the competitive landscape will become more crowded as dozens of companies large and small begin shipping domain-specific AI hardware chips as the year progresses.