Markets are clearly looking for Nvidia to once more comfortably beat analyst expectations when it reports on Wednesday afternoon. Currently, the consensus among analysts polled by FactSet is for Nvidia to post October quarter (fiscal third quarter) revenue of $4.41 billion (up 46% annually due to both strong organic growth and the Mellanox acquisition) and non-GAAP EPS of $2.58.
For the January quarter -- Nvidia usually provides quarterly sales guidance in its reports -- the revenue consensus stands at $4.4 billion (up 42%).
I’ll be live-blogging Nvidia’s earnings report, which is typically released 20 minutes after the market’s close, and an earnings call set for 5 P.M. Eastern Time. Here are some things for investors to watch.
1. Gaming Momentum
Desktop graphics cards, gaming notebooks and game consoles all continue flying off the shelves, and Nvidia’s Gaming segment is benefiting in a variety of ways. The consensus is for the segment’s revenue to be up 24% annually in the October quarter to $2.06 billion, and 31% in the January quarter to $1.96 billion.
Desktop GPU sales got a boost last quarter from the September launch of Nvidia’s first Ampere-architecture gaming GPUs (the GeForce RTX 3080 and 3090). All signs suggest demand also remains strong for Nvidia’s Turing-architecture notebook gaming GPUs, as well as for Nintendo’s Switch console, which is powered by Nvidia’s Tegra X1 SoC.
2. Demand from Cloud Giants
Following very aggressive first-half data center spending, some Internet/cloud giants (the proverbial hyperscalers) are taking a breather in the second half of the year, judging by recent disclosures from firms such as Intel (INTC) - Get Report, Micron (MU) - Get Report and Western Digital (WDC) - Get Report. However, Nvidia CEO Jensen Huang indicated on Nvidia’s Aug. 19 earnings call that his firm would continue seeing strong hyperscaler demand in the second half, aided by ever-growing AI investments and an ongoing shipment ramp for Nvidia’s flagship A100 server GPU.
Along with server GPU sales, Nvidia’s business with hyperscalers now also includes sales of Mellanox’s high-speed interconnect hardware. On Monday, Nvidia announced a memory upgrade for the A100, along with 400-gig Mellanox InfiniBand products.
The consensus for Nvidia’s Data Center segment -- it covers sales of server GPUs for AI/HPC workloads, Mellanox’s products and Nvidia’s DGX servers to cloud and non-cloud clients -- to see revenue grow 148% in the October quarter to $1.8 billion, aided by both organic growth and the Mellanox acquisition. January quarter revenue is expected to be up 100% to $1.93 billion.
3. Demand from Traditional Enterprises
While forecasting continued hyperscaler strength during the October quarter, CFO Colette Kress said on the August call that her company has “a little bit more of a mixed outlook” for Data Center segment sales to traditional enterprises (referred to by Nvidia as “vertical industries” clients) due to macro uncertainty. More recently, Q3 tech earnings reports have pointed to ongoing enterprise hardware weakness.
Nvidia’s DGX server sales largely involve traditional enterprises, and they also account for a large portion of the sales recorded by Nvidia’s Professional Visualization (workstation GPU) business.
Following a 30% July quarter drop, the consensus is for Professional Visualization revenue to be down 34% in the October quarter to $215 million, and 31% in the January quarter to $228 million. Going forward, the arrival of Nvidia’s Ampere-based RTX A6000 workstation GPU and A40 server GPU, which are respectively set to launch in mid-December and in early 2021, could give the segment a boost.
4. Other Businesses
The consensus is for Nvidia’s Automotive segment revenue -- it covers both sales of infotainment processors and engagements for Nvidia’s Drive ADAS/autonomous driving platform -- to be down 32% annually to $110 million. With many other automotive chip suppliers having reported seeing demand rebound sharply in Q3 as auto plant shutdowns ended, there could be some upside to that consensus figure.
Nvidia’s OEM & IP segment, which among other things covers sales of non-gaming PC GPUs and Nvidia’s Jetson embedded computing boards, is expected to see revenue grow 12% to $160 million. A strong notebook market and Nvidia’s Jetson design win momentum could drive upside here.
5. Comments About Product Shortages
In early October, Huang said Nvidia expects the RTX 3080 and 3090 to remain in short supply through year’s end. For the time being, both products are still out of stock at major retailers (and generally selling for big aftermarket premiums), and the same holds for the cheaper and more recently-launched RTX 3070.
In addition, Taiwan Semiconductor (TSM) - Get Report, which produces most of Nvidia’s GPUs outside of the Samsung-manufactured RTX 3000 line, remains capacity-constrained for its most advanced manufacturing processes. As a result, it’s worth paying attention to any earnings call comments Nvidia shares about product shortages or chip wafer supplies.