Ethereum “miners,” using expensive hardware, race to be the first to validate transactions on the Ethereum blockchain. The victor generates a new load of the cryptocurrency ether, which can be sold for cash.
Originally these operations were small-time; miners revved up cheap graphic-processing units, or GPUs, in their rooms, generating a dribble of steady income. But over the past few years expensive, specialized “application-specific integrated circuit,” or ASIC, rigs have threatened to dominate Ethereum mining, which relies on an old hashing algorithm known as Ethash.
The result is that GPU miners, who run cheaper hardware, would be squeezed out by the large corporate mining farms that can afford these ASICs. The “Ethereum Improvement Proposal” known as ProgPoW is supposed to change this and level the playing field, as well as decentralize the network away from the largely China-based ASIC operations.
But ProgPoW's detractors say the upgrade wouldn't solve this and is functionally useless. They believe ProgPoW's only effect will be the introduction of NVIDIA-optimized code onto the Ethereum network, allowing NVIDIA—and its partner, Core Scientific—to profit.
Here's the evidence, as the accusers tell it: First, NVIDIA and Core Scientific have several deals involving GPUs. One, brokered in January, involves Core Scientific's hosting of a ton of NVIDIA AI devices. Another involves the sharing of ProgPoW-ready tech, according to Alexander Levin, the president of mining firm ASICshack (and a hardcore ProgPoW critic). If NVIDIA GPUs were to suddenly surge in popularity, both companies would profit; NVIDIA from raw sales, Core Scientific from GPU hosting.
But Minehan is the key. It is through her, the accusers say, that the companies have direct influence over ProgPoW's development. Levin alleges that her pseudonymous collaborators, “Def” and “Else,” are in the pay of NVIDIA. A data dump purports to incriminate Minehan as an NVIDIA patsy, though its veracity, admittedly, is impossible to establish. (Minehan asserts the info was doctored: indeed, one supposed screenshot purports, unconvincingly, to show Minehan corresponding with a shadowy benefactor literally called "NVIDIA Corporation.")
And to top it all off, analysts have described ProgPoW as better suited for NVIDIA's GPUs over its rival, AMD. Why else, if not for a grand conspiracy, would the upgrade disadvantage a growing competitor?
It “seems likely that ProgPOW is just a Trojan horse for Core Scientific to take advantage of hardware they already have setup,” said Ameen Soleimani, the CEO of crypto adult entertainment platform Spankchain and a hardcore critic of ProgPoW. “Which means it could be even worse in terms of centralization than not doing it.”
NVIDIA refused to discuss the matter. “We’ll decline to comment on this end,” a spokesman, Robert Sherbin, told Decrypt. Asked if that could be viewed as an admission of guilt, he said: “Declining to comment is simply that—not commenting on the matter.” Core Scientific Chief Administrative Officer Matthew Bishop, meanwhile, denied the ProgPoW-Core Scientific connection outright, professing himself unsure of what ProgPoW even was: the Ethereum algorithm change "hasn’t been a topic of discussion” among his clients, he said.
Minehan said that though it’s true Core Scientific and NVIDIA are involved in several deals, they are explicitly for artificial intelligence GPUs. Those devices, she says, would generate only a dribble of profit from crypto mining. She also acknowledged that Core Scientific and NVIDIA could benefit from an uptick in GPU sales caused by ProgPoW. But that’s their business. “What is wrong with companies profiting?” she said.
And the allegation that Def and Else are NVIDIA Volta engineers? Nonsense, said Minehan, who claims that any evidence to the contrary was faked to smear her. Further, she said that ProgPoW does not favor NVIDIA’s chips over AMD’s: both companies audited the code.
(AMD has not responded for comment.)
Neither does Minehan dispute that she has a long-standing relationship with NVIDIA. “I have worked very closely with NVIDIA and AMD for the majority of my professional life,” she said. “Why? I’m an expert in blockchain technology, and an expert in cryptocurrency mining. NVIDIA keeps an eye very closely on the GPU market.”
That's Geek to me
The accusations, however, do not end there: Last week Bob Summerwill, executive director of the Ethereum Classic Cooperative, highlighted links between Core Scientific and Calvin Ayre, resulting in Minehan’s disinvitation from an ETC conference.
Both Core Scientific CEO Kevin Turner and Calvin Ayre beneficiary Craig Wright, Summerwill pointed out, hold advisory positions at the same company: Squire mining, a former precious metals mining company that pivoted to cryptocurrency mining after a reverse-merger into Ayre’s CoinGeek in March 2019 that involved a $22.5 million share purchase. A few months later, Squire inked a binding letter of intent with Core Scientific, to host 41,166 Bitcoin SV compatible ASIC devices.
Was the invisible hand of Calvin Ayre pulling the strings, attempting to bring the ProgPoW operation under CoinGeek’s purview?
Summerwill believes so. Indeed, he is convinced that CoinGeek, a large miner of the cryptocurrency BSV, is set to be NVIDIA’s primary purchaser once ProgPoW comes through and eliminates the ASICS. He also asserts, as evidence, Minehan’s many informal connections with CoinGeek and BSV.
CoinGeek denies this outright. “I can say with some authority that this is conspiracy theory nonsense,” said Ed Pownall, a spokesperson. Minehan also denies it, saying CoinGeek’s deal concerns only Bitcoin miners; hardly ideal for a takeover of the Ethereum network. Minehan also forwarded a private message from BSV backer and self-proclaimed Bitcoin inventor Craig Wright, in which Wright described ProgPoW as “dumb as fuck.”
Wright did not respond for comment.
Minehan's critics also look askance on her work history. Levin, of ASICshack, points out that Minehan used to run a curious, now-defunct conglomerate of mining firms called Mineority, until she left to join Core Scientific in August, 2018. Levin and others claim that Mineority was a costly failure: promised payments never came through, customer support tickets grew into an unmanageable backlog, and business roadmaps became cul de sacs, according to him and several people who spoke to us.
Minehan says the accusations are false: “We had no legal employees,” she said. “You have all these people saying they worked for Mineority, or were proper legal employees, but won’t be able to show any employment. They refused to be employees. A lot of the individuals never wanted to sign a contract.”
Now, as tensions flare, ProgPoW’s fate grows uncertain. Though two audits of the code have nulled some basic technical concerns, Tim Beiko, who helps organize developer meetings, notes that while it is has been "accepted” by the community, additional concerns around the ‘political’ aspect imperil it.
Online pundits have rushed to the same conclusion. Nick Johnson, a developer on the Ethereum Name Service, wrote on Twitter Thursday: “At this point I think ProgPoW has no chance of going through, not because it's a bad idea, but because people have successfully sown enough Fear, Uncertainty and Doubt that the objective arguments for and against it have become irrelevant.”
Meanwhile Minehan’s departure from Core Scientific, although reassuring some, has left others even more suspicious: Why quit your job for an EIP, one person asked in the comments beneath Minehan’s departing announcement. Indeed, what does Minehan stand to gain from leaving a high-flying C-suite job to pursue a “tentatively approved proposal”?
Minehan says she left because “Internet mobs” were harming Core Scientific’s business. “It’s very easy to spin up the Internet, and attack people who have no idea what ProgPoW is,” she said. And while Core Scientific does stand to benefit from ProgPoW, Minehan believes that her perceived conflict of interest is now neutralized. “There’s no benefit for me personally,” she said. Although she does run a handful of personal ASICs and GPUs, she said.
But perhaps Ethereum inventor Vitalik Buterin put it best: "I think the main harm of ProgPoW is that it just wastes people's mental energy on something that ultimately is not a big deal because we are switching to Proof of Stake," he told Decrypt. Proof of stake, of course, is an entirely new algorithm that doesn't involve miners, mining hardware, or NVIDIA.
Updated to clarify that Minehan does not, in fact, have equity in Core Scientific.