Nvidia continued to report strong numbers in its first quarterly results this year, despite failing to supply graphics cards to vendors on time due to a shortage of chips.
During an earnings call on Thursday, CEO Jensen Huang, boasted that there was “growth across every platform”, and that the “datacenter business achieved another record and gaming remained strong.”
He put it down to Fortnite and PlayerUnknown's Battlegrounds (PUBG), both hugely popular online shooter games particularly addictive amongst millions of young teens. As for the datacenter, Huang’s favourite thing to bang on about is AI and how Nvidia’s GPUs are needed to handle the heavy workloads in training neural networks. There’s no stopping him gushing about it.
“At the heart of our opportunity is the incredible growth of computing demand of AI, just as traditional computing has slowed. The GPU computing approach we have pioneered is ideal for filling this vacuum. And our invention of the Tensor Core GPU has further enhanced our strong position to power the AI era,” he said.
Here are the main figures you’ll need to know for Q1 of 2018:
- Revenues were $3.21bn, up 66 per cent on the year-ago quarter.
- Net income was $1.24 bn, up 145 per cent year-on-year.
- Earnings per diluted share for the quarter were $1.98, an increase of 20 cents in the last quarter, and a boost of 151 per cent from this time last year.
At this year’s GPU Technology Conference (GTC), Huang admitted that there was a shortage in high-end chips likes its Volta series marketed at companies and academics dealing with heavy AI workloads. But its cheaper GeForce series, aimed at the gaming industry for rendering realistic graphics, have been hit harder by people snapping them up to mine cryptocurrencies like Ethereum and BitCoin.
Last time we checked it was about two to four weeks wait for Volta chips, and a 12 to 16 week slog for GeForces. THis has meant that the price for GeForce cards has shot up hundreds of dollars compared to last year.
Huang said he hoped to see the price inflation decrease and return to more sensible levels so that gamers could benefit from them. Nvidia have also cancelled its controversial GeForce Partner Program (GPP) citing “unnecessary rumours” about bullish tactics and anti-competitiveness.
He also batted away questions about competition from Google’s newly announced TPU3 chip, also used for training and inference in neural networks. He insisted that Google was still behind the Volta architecture.
Unlike Google’s TPUs, which are only available on its cloud, the Volta is programmable and flexible, whether it was for medical imaging, computer graphics or deep learning. ®