There was a jibe circulating when OpenAI’s game-changing large language chatbot ChatGPT was just entering the zeitgeist that in the space of two weeks, it already had more proven use cases than cryptocurrency and blockchain technology.
It’s hard to disagree. A bit like mice, it’s said that you’re never more than six feet away from a ChatGPT user, whether that’s a journalist, coder or Nick Cave imitator.
Blockchain users and crypto enthusiasts, on the other hand, tend to be stuck in their own circular feedback loop, self-perpetuated because they all follow the same 50 Twitter accounts.
Periodically a Sam Bankman-Fried or Do Kwon will pop up in the news to remind people that crypto is still a big deal, but for the most part, it remains an esoteric market made up of magical internet money beans.
Yet if there is one observable use case for crypto, it’s that it has made some corporations and individuals a s**t tonne of money.
Take Nvidia, the US technology multinational whose high-spec graphics processing units became the preferred unit for non-bitcoin cryptocurrency mining operations.
Exactly how much Nvidia has netted from GPU sales to crypto miners is impossible to know, because the firm never acknowledged mining as a separate line on the balance sheet, bundling it instead within the gaming segment.
But rest assured it was no pittance; the great GPU shortage of 2021 was partially attributable to the surge in demand for crypto mining rigs amid a rally on crypto prices.
The anti-crypto, crypto-adjacent company
Despite this buying bonanza, Nvidia sought to actually limit its crypto-mining customer base by implementing ‘Lite Hash Rate’ (LHR) technology into its GPUs, which reduced the cards’ ability to mine for ether and other non-bitcoin cryptos (Bitcoin requires a specific computer called an ASIC to mine).
Miners, the enterprising bunch that they are, quickly managed to exploit a workaround that unlocked the GPUs’ full mining performance once again.
Subsequent GPUs then used a reinforced LHR protocol, while Nvidia launched the mining-specific CMP (Cryptocurrency Mining Processor) card with little to no fanfare in 2021.
In what was likely an extremely handsome margin, CMPs were reportedly selling for as high as US$4,300 at the time.
According to Nvidia’s 2022 annual results, CMP sales brought in US$550mln in revenues, which was merely 2% of total group revenues, but given how mainstay GPUs remained popular for mining, crypto-derived revenues were probably much higher.
The opaqueness of Nvidia’s crypto-derived revenues actually became a legal issue when, in May 2022, it was fined US$5.5mln by the US Securities and Exchange Commission (SEC) for failing to disclose that cryptocurrency mining comprised a “significant element of its material revenue”.
“NVIDIA’s disclosure failures deprived investors of critical information to evaluate the company’s business in a key market,” Kristina Littman, chief of the SEC Enforcement Division’s crypto assets and cyber unit, said at the time.
AI for an eye
Throughout Nvidia’s historic aversion to acknowledging its profit-generating crypto mining customer base, the group has never really been forthright about its disdain for the sector.
Until, in a recent interview with The Guardian, chief technology officer Michael Kagan exclaimed: Crypto doesn’t “bring anything useful for society… I never believed that (crypto) is something that will do something good for humanity.”
On the other hand: “With ChatGPT, everybody can now create his own machine, his own programme: you just tell it what to do, and it will. And if it doesn’t work the way you want it to, you tell it ‘I want something different’.”
It’s a unique position for a company to be in: Intentionally dissuading a highly lucrative customer base from buying your products.
An argument can be made that it’s an ethical choice due to crypto mining’s environmental impact, but Kagan’s statements seem to suggest Nvidia just simply doesn’t ‘vibe with’ the crypto mining sector and would prefer distributing its GPUs to more deserving customers.
Such as OpenAI, which used around 10,000 Nvidia GPUs to build ChatGPT.
That says a lot about where Nvidia sees its financial and philosophical future heading.