Skip to main content

Nvidia Fourth-Quarter Earnings Live Blog

Real Money's Eric Jhonsa covers Nvidia's fourth-quarter earnings report and conference call with management.
  • Author:
  • Publish date:

With Nvidia still seeing all the demand it can handle from gamers and cloud giants, investor expectations are high ahead of the GPU leader’s latest earnings report.

Among analysts polled by FactSet, the consensus is for Nvidia to report January quarter (fiscal fourth quarter) revenue of $7.42 billion (up 48% annually), GAAP EPS of $1.00 and non-GAAP EPS of $1.23.

Nvidia also provides quarterly sales guidance in its reports. The FactSet revenue consensus for Nvidia’s April quarter stands at $7.31 billion (up 29%).

Eric Jhonsa, Real Money’s tech columnist, will be live-blogging Nvidia’s report, which is expected at 4:20 P.M. Eastern Time, along with an earnings call scheduled for 5:30 P.M.. Please refresh your browser for updates.

6:42 PM ET: Nvidia's call has ended. Shares are down 1.8% after-hours to $260.25 after Nvidia posted another beat-and-raise quarter. FQ4 sales and non-GAAP EPS respectively beat by $220M and $0.09, and FQ1 sales guidance is $790M above consensus at the midpoint.

Gaming and Data Center segment sales respectively rose 37% and 71% Y/Y in spite of supply constraints, while Pro Visualization sales rose 109% Automotive sales fell 14%, but Nvidia forecast the segment will see strong growth later this year as design wins ramp.

On its call, CFO Colette Kress said FQ1's Q/Q growth will "primarily" be driven by Data Center. And CEO Jensen Huang said (following the recent termination of the Arm deal) that Nvidia's Grace Arm server CPU is still on track to launch next year, while adding that Grace will be succeeded by follow-on Arm CPUs.

Thanks for joining us.

6:34 PM ET: Huang with some closing remarks. He talks up Nvidia's recent performance and product launches, while taking some time to once more go over Nvidia Drive and Omniverse's potential. Says he can't wait to share more about new developments at next month's GTC conference.

6:31 PM ET: A question about long-term margin expectations, and the impact that software could have.

Kress: Our margins already benefit from our software investments. And as our standalone software businesses grow, that should be accretive to our margins over the long run.

6:29 PM ET: A question about Data Center expectations for this year, and whether upcoming Intel/AMD CPU launches and the arrival of DDR5 memory could provide a boost.

Huang: There are multiple server performance bottlenecks. We're the largest consumer of the fastest memory. We also work to address networking bottlenecks. All of this is meant to reduce CPU bottlenecks. We're constrained on throughput with PCIe Gen 4, but won't be with PCIe Gen 5.

6:25 PM ET: A question about whether networking sales could outgrow total Data Center sales this year. And one about driving synergies with Mellanox.

Huang: If not for our collaboration with Mellanox, our DGX SuperPod and Meta's new AI supercomputer wouldn't be possible. I'm thrilled with the acquisition and the work that the Mellanox team is doing. Mellanox roadmaps have been accelerated. BlueField's roadmap has probably been accelerated by a year.

Regarding networking growth, Huang says that if not for supply constraints, the business would be out-growing cloud service provider (CSP) growth. Notes demand for high-speed networking remains very high.

6:21 PM ET: A question about Data Center growth. How much is from units vs. pricing vs. mix?

Kress: It's early in the quarter to determine the contributors for Data Center growth. Gaming should see sequential growth as well, and so might other segments.

6:19 PM ET: A question about RTX 3050 sales and where Nvidia is in the Ampere cycle.

Huang: RTX is a home run. It reinvented modern computer graphics, via ray-tracing and DLSS. We're about a third of the way through upgrading an installed base that is growing. Steam usage has grown by 50% over the last 2 years. Epic's game store is seeing huge growth. Hybrid work is another growth driver, fueling workstation GPU demand. Also, many creative workloads are shifting to 3D. 3D graphics are now used for all forms of digital content creation. We're in the early innings of RTX.

6:15 PM ET: A question about AI adoption rates among hyperscaler and enterprise clients.

Huang: The number of AI applications is growing incredibly fast. AI chatbots will be pervasive in the future, across websites, retail stores, etc. We're still in the early days of AI adoption. Almost all future software will be written with the help of AI or by AI. 

6:10 PM ET: A question about Nvidia's software revenue.

Huang: Software is a key part of our accelerated computing efforts. We've always been a software-driven business. We license enterprise software on a per-server-node basis. We believe every server will eventually be running AI software. Omniverse is based on connections. There are 40M designers and creators worldwide, and there will be hundreds of millions of robots. The Omniverse business model is per connection per year. With Nvidia Drive, we share in the economics of the software and services it helps deliver. The lifetime opportunity with cars is quite high.

6:06 PM ET: A question about the big Q/Q increase seen in inventory purchase and supply obligations.

Huang: We expanded our supply chain footprint a lot this (fiscal) year. This was partly in preparation for some exciting product launches. Orin is going to be a very large business for us going forward. This year will feature some major product launches.

6:04 PM ET: A question about Nvidia's plans for Grace.

Huang: We have multiple Arm projects going on, including connected devices and robotics processors. Orin sales will inflect this year. Grace will be succeeded by follow-on Arm CPUs. Hyperscalers are embracing Arm CPUs. Grace "is just the first example" for us. Our focus is on enabling accelerated computing.

6:01 PM ET: A question about how much wafer supply limited FQ4 Data Center growth.

Kress: We still have supply constraints for some businesses, such as in Networking. We expect supply to improve each quarter this year.

Huang: We have more Data Center demand than supply. Data Center covers a variety of GPU and networking products. We saw constraints pretty much across the entire range of products. Our guidance assumes an improved supply base.

5:57 PM ET: First question about Nvidia's strong Data Center growth expectations for FQ1.

Kress: Our FQ1 guidance can include an acceleration for Data Center growth. We'll see growth in Gaming and possibly other segments as well.

Huang reiterates that Data Center is growth is being fueled by a number of end-markets and workloads. Natural language models, recommender systems, cloud gaming, etc.

5:54 PM ET: The Q&A session is starting.

5:54 PM ET: Now back to Kress to discuss Nvidia's guidance. Notably, she says Q/Q growth in FQ1 is expected to be driven "primarily" by Data Center.

She also says FY23 non-GAAP opex is expected to grow at a similar % to FY22, which saw 27% opex growth.

5:52 PM ET: Huang is talking again, this time about the termination of the Arm deal following regulatory pushback. Says Nvidia "gave it our best shot," but opposition was strong. Reiterates that Nvidia plans to launch its Grace Arm server CPU in 2023.

5:50 PM ET: Kress: Networking saw strong demand across end-markets. Supply remains a constraint.

5:48 PM ET: Kress goes over some recent AI benchmark records and design wins obtained by Nvidia's server GPUs. Also talks about the broadening of Nvidia's server OEM and system integrator ecosystem.

5:45 PM ET: Kress is talking again. She notes Nvidia's Data Center sales to hyperscalers more than doubled Y/Y, and that its AI inference-related sales more than tripled. Demand from traditional enterprises (referred to as "vertical industries" clients) is also said to be strong.

5:43 PM ET: Kress hands the mic over to Jensen Huang, who's going for the deal Nvidia announced this morning with Jaguar Land Rover. The deal will result in all new Jaguar Land Rover vehicles shipping with Nvidia's Drive platform starting in 2025, and will yield both hardware and software revenue streams.

5:41 PM ET: Regarding Automotive, Kress says the business will return to Q/Q growth in FQ1, with more significant growth afterwards as Nvidia Drive design wins ramp.

5:40 PM ET: Kress notes the Omniverse 3D collaboration platform (recently entered general availability) doesn't merely drive software revenue, but act as another driver for GPU demand.

5:39 PM ET: Kress: Strong Q/Q Pro Visualization growth was driven by a shift towards higher-end workstations and Ampere demand.

5:38 PM ET: Kress reiterates that Nvidia has limited visibility into how many of its gaming GPUs are going to crypto miners, and that nearly all of its desktop GPU shipments now involve limited hash rate (LHR) parts.

5:36 PM ET: Kress: More than 150 notebook design wins were announced for RTX 30-series GPUs. Channel availability for our Gaming products remains low.

5:35 PM ET: Kress: Gaming saw record desktop and notebook GPU sales, with strong high-end demand. 

5:34 PM ET: Kress is talking.

5:33 PM ET: IR chief Simona Jankowski is going over the safe-harbor statement. Typically, Nvidia's call features prepared remarks from CFO Colette Kress, after which Kress and CEO Jensen Huang take questions from analysts.

5:32 PM ET: The call is starting.

5:27 PM ET: Here's the webcast link, for those looking to tune in.

5:26 PM ET: Hi, I'm back to cover Nvidia's call, which should start in a few minutes.

5:03 PM ET: I'm taking a short break, but will be back to cover Nvidia's earnings call, which is set for 5:30 PM ET. Shares are currently up 0.4% after-hours after Nvidia beat FQ4 estimates and issued FQ1 sales guidance that was well above consensus.

5:02 PM ET: In spite of its big prepayments, Nvidia ended FQ4 with $21.2B in cash and marketable securities (up $1.9B Q/Q). Debt remained at $10.9B.

4:59 PM ET: Nvidia's inventories rose 43% Y/Y to $2.61B. That's significant growth, but it still trailed FQ4 revenue growth of 53%.

4:57 PM ET: GAAP R&D spend totaled $1.47B (+28%). Judging by the product announcements made during Nvidia's November GTC conference, a large portion of that spend is now software-related.

4:54 PM ET: Nvidia continues spending aggressively: GAAP operating expenses rose 23% Y/Y to $2.03B, while non-GAAP opex rose 22% to $1.45B. Both figures were slightly ahead of guidance.

For FQ1, Nvidia is guiding for GAAP opex of $2.19B (excluding the Arm write-off) and non-GAAP opex of $1.6B.

4:50 PM ET: The 14% Y/Y Auto revenue drop is attributed to the ongoing decline in Nvidia's "legacy" infotainment SoC sales, along with auto production constraints, partly offset by self-driving program ramps.

4:48 PM ET: As one would expect, FQ4 Data Center growth (+71% Y/Y) was said to be driven by Ampere-architecture GPU sales "across both training and inference for cloud computing and AI workloads such as natural language processing and deep recommender models."

Gaming growth was likewise said to be driven by Ampere GPU demand, with game console SoC sales (primarily to Nintendo) dropping Q/Q due to seasonality.

4:44 PM ET: FQ4 free cash flow was $2.74B, up from $1.28B a year earlier.

For the whole of FY22, FCF was $8.05B, up from FY21's $4.68B. Excluding cash flows related to stock compensation, FY22 FCF was around $6B.

4:41 PM ET: Nvidia notes its FQ1 GAAP EPS will include a $1.36B write-off for the prepayment it made to Arm in 9/20, as a result of the deal being terminated earlier this month.

4:38 PM ET: Nvidia continues making giant prepayments to help secure wafer capacity: The company's "outstanding inventory purchase and long-term supply obligations" totaled $9B at the end of FQ4, up from $6.9B at the end of FQ3 and just $2.54B a year earlier.

Foundry partner TSMC, which set a massive 2022 capex budget last month, is believed to be getting a large chunk of these prepayments.

4:35 PM ET: Within the OEM & Other segment, sales of CMP mining GPUs fell to just $24M from FQ3's $105M. This fits with Nvidia's guidance for CMP sales to be "negligible" during the quarter.

4:33 PM ET: Nvidia continue fluctuating after-hours: Shares are now down just 0.4%. Will be interesting to see how the stock trades as the call progresses.

4:32 PM ET: Pro Visualization revenue (driven by workstation GPUs) rose 109% Y/Y to $643M, easily beating a $559M consensus. 

Auto revenue fell 14% to $125M, missing a $140M consensus. OEM & Other revenue (it includes sales of low-end PC GPUs, mining GPUs and edge AI products) rose 25% to $192M, topping a $178M consensus.

4:29 PM ET: Here's the earnings release. And here's the CFO commentary.

4:27 PM ET: FQ4 GAAP EPS was $1.18, above a $1.00 consensus.

GAAP and non-GAAP gross margins respectively came in at 65.4% (up 230 bps Y/Y) and 67% (up 150 bps), close to Nvidia's guidance. For FQ1, Nvidia is guiding for a GAAP GM of 65.2% and a non-GAAP GM of 67%.

4:25 PM ET: Nvidia is now down 2.3% AH to $258.99. The numbers look solid (particularly the FQ1 guide), but as mentioned before, pre-earnings expectations were pretty high.

4:24 PM ET: Gaming revenue rose 6% Q/Q and 37% Y/Y to $3.42B, topping a $3.36B consensus.

Data Center revenue rose 11% Q/Q and 71% Y/Y to $3.26B, above a $3.18B consensus.

4:22 PM ET: Shares are roughly flat after-hours for now.

4:21 PM ET: Nvidia guides for FQ1 revenue of $8.1B, plus/minus 2%. That's soundly above a $7.31B consensus.

4:20 PM ET: Results are out. FQ4 revenue of $7.64B beats a $7.42B consensus. Non-GAAP EPS of $1.32 beats a $1.23 consensus.

4:15 PM ET: Consensus sales estimates for Nvidia's business segments:

Gaming - $3.36B (+34% Y/Y)
Data Center - $3.18B (+67%)
Pro Visualization - $559M (+82%)
Auto - $140M (-3%)
OEM & IP - $178M (+16%)

The quarter's Gaming and Data Center sales might be more a function of supply rather than demand, given that the latter has been running well ahead of the former for gaming and server GPUs, as well as for some networking products.

4:08 PM ET: As a reminder, Nvidia's report is expected to arrive at 4:20 PM ET. The call is set to kick off at 5:30.

4:06 PM ET: Nvidia closed nearly unchanged on Wednesday ahead of its report. Shares are down (amid a broader tech selloff) since the company posted a strong FQ3 report on 11/17, but remain up 73% over the last 12 months. Those gains leave Nvidia sporting a $652B market cap heading into its report.

4:02 PM ET: The FactSet consensus is for Nvidia to report January quarter revenue of $7.42B and non-GAAP EPS of $1.23. The April quarter revenue consensus is at $7.31B.

Given the demand strength that Nvidia is seeing across its biggest end-markets, informal expectations are probably higher.

4:00 PM ET: Hi, this is Eric Jhonsa. I'll be live-blogging Nvidia's earnings report and call.