SEC charges Nvidia with inadequate disclosures about impact of cryptomining
sec.gov
sec.gov
Hey, all you blockchain skeptics, here’s what you are always asking about: a practical use for blockchain tech.
Having equipment sitting around unused is way more expensive than most people care to think about.
A heating coil might not be terribly valuable in July, but come November it will have roughly the same value it had the previous February. Bitcoin mining hardware that's switched off for five months is losing value at an alarming rate, and very quickly won't be a net reduction in costs.
Very quickly a heat pump becomes cheaper to operate, and cheaper to purchase. And if you amortize the cost over the life expectancy, that threshold is very, very low. People who have recent memory of living paycheck to paycheck are constantly screwed by the latter, and that Venn diagram overlaps heavily with several other circles that make crypto super attractive to some people and super ridiculous to others.
I went down this avenue of compute as radiator long ago, and I could never get the math to work, unless I moved somewhere that was cold most of the year, and that sure as fuck is not going to happen.
I recently replaced the 3 ton hvac unit on my small 1500 sq ft single family home with a heat pump and all the bids were in that range.
Also if you want to control every single room including bathrooms and walk in closets it gets tricky.
If they were going to heat using an electric space heater, it's a wash.
If they were going to heat using natural gas, but the electricity comes from nuclear, it's lower carbon.
If they were going to heat using natural gas, and the emissions are higher. BUT, if they were going to run the mining rig anyway, and they've merely turned off the gas heat, the emissions are overall lower.
Heat pumps work by moving heat, they usually move more heat than input energy.
Basically, a refrigerator is a heat pump. The input energy runs the heat pump, which it uses to move heat out of the refrigerator. Typically, for every watt of electricity, more than a watt of electricity is moved out of the refrigerator.
They work by compressing gasses into fluids, and then letting the fluid expand back into a gas. Basically, when a fluid evaporates, it absorbs heat as potential energy. The energy can be harvested by compressing the gas at a high enough pressure that it condenses into a fluid. Do this in a loop, and you can move heat.
No laws of physics violated!
Suppose you built a shed around the outdoor unit of a central air conditioning system and let it come up to max ambient temperature. Then right next to it you built a shed with a resistive heater which consumes the exact same wattage as the air conditioner. The first shed will be much warmer because you're not creating heat so much as you're moving heat. If you increase the resistive heater's wattage by about 2.5x then the sheds will be about the same temperature.
* Don't actually do this and expect the system to survive.
Unfortunately they care. For now only poor face real consequences of climate change. They are to poor to mitigate its effects.
No you don't, when permafrost melts it causes ground to move. It is really bad for everything build on or in it. Something like houses sinking and falling apart. In some cases it cause big holes in the ground and they got even bigger over time.
Commercial solutions have been in development for a few years. Here’s one example: https://www.wisemining.io/
Edit: after re-reading your comment I realize you also aren’t aware that miners can be throttled. Powering it down is not the only alternative to consuming 3kW.
The only thing I did find was that apparently Russians are taking the components from kitchen appliances to use for other things, which seems entirely unrelated.
Ultimately the heat output has nothing to do with how the tech is used (like bitcoin). If Bitcoin (or other cryptocurrencies) stopped holding value people wouldn't heat their homes like this because it would cost more than traditional heating methods (as someone pointed out above about heat pumps.)
You forget exactly how efficient a coal-fired powerplant actually is. Love or hate them, they have 100+ years of technological innovation to squeeze every watt out of coal. Your fireplace at home is not nearly as efficient. So an electric heater powered by the grid is almost certainly more efficient than burning your own coal at home in an inefficient stove.
Efficiency wise, the comparison between a heat pump and a furnace depends on the outside temperature. Heat pumps lose their efficiency for high ∆T, so for winter in a Russia, a furnace is likely the better option
One day we may heat our homes with distributed computing like BOINC.
Consider the house that marginally has electrical resistive wall heaters (Like the bay area) and an excess of spare computers sitting around (also like the bay area) ... It's marginally better for the house nerd to leave their PCs on using BOINC than run the wall heater.
When temperatures drop below zero degrees, ice will build up on the outdoor unit of any heat pump. How the heat pump reacts to this determines how effective it will be in providing heat to your home. To remove the ice build-up the heat pump will need to go into Defrost Mode. During this time the heat pump will not be delivering heat into your home. HyperCore’s Defrost Logic has been fine-tuned to extend the period in-between defrost periods and optimise its heating performance.
Mitsubishi Electric offers heat pump systems with Hyper-Heating INVERTER® (H2i®) technology which can provide up to 100 percent of heating capacity at 5° F and continue operation down to -13° F even wiThe Mitsubishi Mr Slim heat pumps that we install are able to warm your home in temperatures as low as -15°C and -25°C thout auxiliary heat.
https://www.mitsubishicomfort.com/articles/what-is-a-heat-pu...
https://www.mitsubishi-electric.co.nz/materials/aircon/broch...
I usually use GPUs for heating during winter as the power has to be used anyway, so why not get ethereum for it.
I don't think you understand how heat pumps work. Heat pumps have significantly greater than 100% "efficiency". They don't turn electricity into heat. They use electricity to move preexisting heat, and it turns out that's far more efficient in terms of Joules of heat delivered to your home per Joule of energy spent. In fact, for any given Joule of energy spent, you can generally move two to three Joules from outside your home to inside of your home.
If the source of heat you're pumping is coming from resistive heating you're using electricity for, you're only getting 1 Joule of heat for every Joule of electricity. Adding heat pumps to this system doesn't help you.
> I usually use GPUs for heating during winter as the power has to be used anyway, so why not get ethereum for it.
Because owning and operating a heat pump is almost certainly cheaper than the costs of owning and operating a mining rig, even offset by the value of the cryptocurrency you generate. You'd almost certainly be better off by heating your home with a heat pump and using the energy savings to buy that same cryptocurrency.
The exceptions to this are if you are in a location where energy is extremely cheap, or perhaps if you generate more than your household usage of electricity (including resistive heating) through solar or wind but aren't able to sell that electricity back to the grid.
A heat pump doesn't have to take heat from the outside. By using heat from the extracted air, you can use whatever you want to generate heat inside and get multiple times the effective heating as it's reused. Use wood, electrical floor heating or GPUs. It doesn't matter. The energy from the extracted air is transferred to the fresh supply air. The exhaust air will be freezing.
Also, electricity has been expensive the last two years, but mining has still been profitable, considering I already have some GPUs in my workstation and home server.
You don't earn money by using a heat pump, you do(or at least did) by mining. By mining and using a heat pump on that energy I increase the usage of that energy. Win win win.
If you're already spending 1J of energy to get 1J of heat, a heat pump is not going to turn that 1J of heat into 2-3J of heat, nor is it going to recover any of the Joules you've spent to generate that heat. So sure, you can use a heat pump to move that heat around a space. But doing so just spends more energy and decreases the overall efficiency of the system.
The principle of a heat pump getting such efficiency numbers is entirely predicated upon the notion that you're able to move that heat from somewhere it already exists "for free" in sufficient bulk.
> You don't earn money by using a heat pump, you do(or at least did) by mining. By mining and using a heat pump on that energy I increase the usage of that energy. Win win win.
You fundamentally misunderstand the economics of this situation. Mining costs money in the form of hardware and electricity. In exchange you can potentially extract some amount of revenue. If your alternative was to use that energy to generate one Joule of heat for every Joule of energy spent, you might as well mine to get a rebate.
But the sum of mining revenue minus mining costs are almost certainly less than the costs of simply operating a heat pump instead. Again, you'd be better off using a heat pump to heat your home and using the money saved on energy to simply buy $CRYPTO at market prices.
> If you're already spending 1J of energy to get 1J of heat, a heat pump is not going to turn that 1J of heat into 2-3J of heat, nor is it going to recover any of the Joules you've spent to generate that heat.
It's not magic, or complicated. The energy spent is used multiple times, as I said. You can save energy using a heat pump both by using it for initial heating, or transferring existing heat that would otherwise be thrown away.
So no, I'm not misunderstanding anything. If you think it's not possible then you too can get this "magic" using e.g. Nibe F750 + SAM40. Check their documentation, it even has charts for everything!
You insist you know about heat pumps, but your comments indicate you don't understand the fundamental principles.
It's an interesting topic, if you find the time.
Unless I'm misunderstanding part of this proposed setup, I'm pretty sure this doesn't work. The higher efficiency of heat pumps comes from the fact that the outdoors is an effectively infinite (for the purposes of a house) source of temperature differential. You can only move as much heat as exists, so you can't use a heat pump to multiply a finite heat source.
A pro with that solution is that the extracted air has a high temperature all year, making it more efficient that using a heat pump to extract heat from outside that might be -30 celcius. A downside is that the amount of air is limited, so if you need more heat then you have to supplement it with something else.
You get it, but it seems there are a few others that don't realize that a heat pump can also be used to increase efficiency by reducing the heat lost. While I do get that it's unknown to most people, some people here perfectly illustrates the Dunning-Kruger effect...
Check out e.g. Nibe F750 with SAM40.
I'm using the heat pump to heat water using the extracted air. That water is then used for floor heating, ventilation air heating and ofc. hot water.
https://blog.haschek.at/2021/how-i-heat-my-home-by-mining.ht...
https://data.worldbank.org/indicator/EN.ATM.CO2E.PC?end=2018...
"Energy Matters" says that EU energy production was 26% nuclear in 2015, but only 12% for consumption[1], so I'm not sure how to look at that comparison. Either way, that's probably mostly France (~80%[2]). US around 8-9%[1].
TL;DR: Yes, most western countries probably have less nuclear, but it's a low bar.
[0] https://en.wikipedia.org/wiki/Electricity_sector_in_Russia#M...
[1] http://euanmearns.com/primary-energy-in-the-european-union-a...
[2] https://en.wikipedia.org/wiki/Electricity_sector_in_France#M...
Surveillance is an issue, but for the most part, it's not warrantless in our current system. Cryptocurrencies, excluding things like Monero and Zcash actually make it considerably easier to do warrantless surveillance.
If I'm forced to choose between some instances of illegal warrantless surveillance, and the complete inability to trace financial crimes, I'll take the instances of illegal surveillance, with the hope we'll try to fix those issues.
When the war started in neighbouring Ukraine (I live in Romania) the local market started seeing graphics cards again which, previously, had been marked by local official resellers as "not in stock for the near future". This is for business procurement, so not even retail.
I know I'm in the minority, but my GPUs are used almost exclusively for password cracking. I've got ASICs for the crypto mining.
I was not doing that type of work so I'm not sure what type of software was used but I think it had multiple open source programs available.
Did you know that if you're on a Windows network, random machines all around you will just try to authenticate to you, and disclose the hash of the user on the system? Plug in on a Ethernet jack in the lobby, or conference room, get onto the wifi, compromise any machine on the network and run responder or inveigh, and you just get dumps of user passwords, and if you can crack any of them, its a pretty quick path up to Domain Admin on the network, which then can dump everyone's hashes.
Of course, the energy consumed by any even semi-serious mining operation far exceeds a home's heating needs.
[1] YMMV in permafrost regions, and the hardware is too expensive for many people.
And there's smaller mining operations, the economies of scale aren't that great, more the difficulty of finding GPUs. You can lend it to neighbors and give them a cut of the proceeds, you could do the whole town. So in fact ordinary mining operations waste the heat, this would not.
Plus power there is cheap, I would guess nuclear.
Considering how cryptocurrencies seem to attract the get-rich-quick crowd, it’s impressive to see the levels of motivated reasoning they are capable of, since neither motivation nor reasoning alone seem to be among their usual strengths.
Such a beautiful idea, if only the computation was actually useful.
Literally hundreds of alternatives to Proof of Work, especially PoW with some enhanced utility (calculating large twin primes, folding proteins, file storage, GIS, etc) have been tested in the real world, but it turns out that a straight up "find the nonce" PoW is the best form of democracy we can build that doesn't require some global registry of every human on earth. Anyone with access to electricity can participate in ensuring the global security of the network, and be fairly compensated for it.
Proof of Stake can work alright in certain market conditions, but those didn't exist on any blockchain until relatively recently. Look at the utter chaos going on with Terra validators. The validating token goes to approximately zero, so bad actors can cheaply buy up the staking asset and hijack validation, further destabilizing the network, allowing the next attackers to buy in even cheaper.
It's a hard problem, and Satoshi's solution is an incredibly elegant one. If you've got a better idea than PoW, then do share. Whoever figures it out stands to make billions of dollars. You should know that it's mostly considered to be a solved problem at this point, and nearly everyone has moved on to other hard problems in the space.
Blockchains are not useless. Append only, verifiable data structures have countless applications that again, can be used to solve actual problems.
Systems which combine these to create a “trillion+ economy” that’s sole external affect appears to be inducing an obsessive overuse of the word “fiat” while consuming incomprehensible amounts of the worlds resources—not just power, but hardware, and importantly, the focus of many intelligent people—that is a venture of questionable use.
The calculation itself, once removed from bitcoin ecosystem, is quite useless.
> Literally hundreds of alternatives to Proof of Work, especially PoW with some enhanced utility (calculating large twin primes, folding proteins, file storage, GIS, etc) have been tested in the real world, but it turns out that a straight up "find the nonce" PoW is the best form of democracy we can build that doesn't require some global registry of every human on earth.
How is "find the nonce" PoW better for democracy than "fold proteins" or "find a chain of primes" PoW? In what way does calculating primes or folding proteins require global registry of every human on earth?
> Anyone with access to electricity can participate in ensuring the global security of the network, and be fairly compensated for it.
What about internet and expensive hardware, don't you need those also? What would you need besides those to do "fold protein" or "find chain of primes" PoW.
I would love to be proven wrong, to see any reference that grc provides algorithmic benefit to boinc, as in "getting rewarded with grc is proof the provided solution for the requested scientific computation is correct", or even a little thing like storing the boinc participation statistics, but those are features of the boinc network independently from grc.
What grc provides to the scientific community is a weird incentive for monkey-brains: monkey brain sees grc number go up, monkey brain releases happy hormones. It is a mirror of the participation statistic on the blockchain, not because that makes sense, but because blockchain. Sure there is some theory that some monkey may give a monkey a banana for making their number go down and its number go up, but that transaction involves no scientific computation and is purely speculative.
This is a complete, steaming, and self-serving pile of bullshit. It isn't at all democratic and you need far more than just "electricity" to participate.
Better solution: PoS.
Do you know who is using already a highly tuned modern fast currency system based on PoS?
You do. I do. Everyone else does.
It's called us dollar, euro and other stable fiats.
And yes inflation problem doesn't go away just because you use Bitcoin.
We should evaluate things against what the actual, real world alternative would be, not just against the best possible world.
(That said, I should acknowledge that a large part of that electricity comes from burning natural gas at an efficiency loss, so resistive heating is still worse than the natural gas furnace.)
But, to restate my previous comment: if you have a natural gas furnace and a GPU, but no heat pump, and are deciding which to use for heat, it doesn't matter how efficient the heat pump you don't have is.
If you own your own home and have the capital to get a heat pump, great, do it. But a lot of people rent. They can't just swap out their furnace for a heat pump.
As an extra benefit the heat pumps have lower maintenance, risk.
Ahh, now I see where there might be confusion. Heat pumps that operate on electricity are increasingly common in the US. Most of them are air-to-air, but some are ground-source. What I'm claiming are rare-to-non-existent are residential heat pumps that operate directly off of natural gas, using an ammonia absorption process, with no electricity involved.
Arguably 'freemint' was saying that it's more efficient to generate electricity centrally using natural gas and then distribute this electricity to power heat pumps than to burn natural gas for heat at each site. Yes, likely. My surprise was that I thought he was claiming that non-electric natural gas powered heat pumps were currently available to residential consumers. I knew this style exists (see the link I gave above) but I've never seen one in operation.
It is just a more tax efficient way to distribute earnings.
It's bad for society to have companies spend money on their own shares, and dividends have the appealing property that they allow a holder to both continue to hold, and bank some gains for further investments.
I think we should get rid of the tax advantage for capital gains though, the mechanism through which a company distributes capital shouldn't determine tax rate. Also I think dividends make it psychologically easier for people to live off the interest.
If you have 10000 in stock, and there's 20% inflation and the price of your stock goes up to 12000, you haven't made any real gains (no added purchasing power) but you have accrued 2000 in taxable gains.
Instead of adjusting returns for inflation, we just reduced the marginal tax rate for capital gains.
Unless the insiders act illegally. If an insider wants to act illegally they can do so in any number of ways other than knowing that a buy back is coming.
I am curious if there is evidence for this. I don’t see how distributing capital that could be used for growth pressures share prices up.
People not selling back the shares are trading the current profits for a bigger share of future profits, which may never exist. So I am not sure on a risk adjusted discounted cash flow model there is any justification for increased price pressure.
I always presumed that the main benefit of buybacks v dividends was being able to time capital gains for tax purposes.
But this is all speculation too, no evidence behind my post.
The other reason is that there are very few stocks that are being valued by the market on any sort of model. That goes for cash flow models as well as other things like PE ratios, etc. We are in the tinkerbell regime, prices went up over the last 10 years because people are clapping louder. Obviously that is a controversial take but I am not predicting a reversal, just saying that the market overall is in a weird place.
But the fact itself adds demand for the shares on the market, affecting the share price.
Off the market, an offer does become public immediately, that is a necessary component of doing it at all. The sales do not. Directly is the wrong word, there's no more direct way to affect price than a bid/ask/sale of a share. The key is that the information about how much of the stock will actually get bought at what price is delayed on its way to becoming a price signal in the market.
Here's an old-ish but good paper about it, showing how this delay can be exploited by those with inside information. Not a loophole but rather another thing to watch out for. https://chicagounbound.uchicago.edu/cgi/viewcontent.cgi?arti...
Say I have a company worth a $1,000,000 dollars that has a 1,000 shares each worth $1,000. Company comes into a giant one off windfall. They use that windfall to buyback half the shares. There are now there are only 500 shares that own a $1,000,000 dollar company so each share is worth $2,000. Or put another way .2% of future profits is worth more than .1% of future profits.
> I always presumed that the main benefit of buybacks v dividends was being able to time capital gains for tax purposes.
It's not about timing, it's simpler. Buybacks are taxed as capital gains, and dividends as income.
Also, with the windfall the $1,000,000 company is no longer worth that much, it is worth $1,000,000 + the windfall, so you will be buying back at a much higher rate.
But you're right about qualified dividends. I had no idea qualified dividends were taxed at the same rate as capital gains. I'd always hear that was why companies do stock buybacks.
Thanks for teaching me something new. You're right it's about allowing investors to time their taxable events.
But look at a simple example, your example. Company Worth $1,000,000. 1,000 shares, each worth $1,000.
Windfall Event: They get $1,000,000 of unexpected income. Now the company was whatever it was before + $1,000,000.
Naturally, that means the company is worth $2,000,000. So the stock price is $2,000.
Well, we buy back half the shares, but it is 500 shares at $2,000 a share. This costs $1,000,000.
Now each share is worth $1,000 because the company has spent half its capital.
Say windfall is 1 mil.
Post windfall market cap is 2 mil. (1 mil for company + 1 mil in cash)
Company buys 500 shares at 2,000. So the company is now worth 1 mil again but there are 500 shares. 1 mil / 500 shares is 2k per share.
The SEC’s order finds that NVIDIA violated Section 17(a)(2) and (3) of the Securities Act of 1933 and the disclosure provisions of the Securities Exchange Act of 1934. The order also finds that NVIDIA failed to maintain adequate disclosure controls and procedures. Without admitting or denying the SEC’s findings, NVIDIA agreed to a cease-and-desist order and to pay a $5.5 million penalty."
Cost of doing business.
Remember about the SEC whistleblower program, if you had inside proof of intent the fine would be a lot higher, and you'd get a share of the fine!
Granted, this might not have been true in 2018, which appears to be what the SEC's charge is focused on. Nonetheless, I don't think the huge amount of demand from crypto for Nvidia's products should be construed as a bad thing; there's plenty of other demand.
Besides, there is a difference. Because whenever crypto collapses (and it has twice already in the past), a lot of the existing mining GPUs end up on the secondary market. If this happens now, it will have a material impact on the prices of GPUs going forwards.
I don't think nvidia even did this, they marketed to gamers and reported the correct number of sales. What people do with the card once its sold is not up to nvidia its not information that an investor can reasonable assume nvidia will report on.
If I invest in a steel mine they market their steel to car manufactures and report the correct number of sales. If they also happen to sell to gun manufactures I can't then sue said company for not reporting that.
It’s not about that. It’s about:
“NVIDIA had information, however, that this increase in gaming sales was driven in significant part by cryptomining.”
Which they did not disclose.
> Which they did not disclose.
What's annoying to me as an NVDA investor is that I still don't have that information. Or maybe it can be found somewhere else?
The problem is that NVIDIA leadership knew/believed that a proportion X of their GPUs were being bought for crypto mining (thus tying demand with crypto prices), while they reported to investors a different number, Y, with Y < X. This is material information that they simply lied about, which is illegal for a publicly traded company. In fact, the lawsuit could have happened even if Y > X.
If NVIDIA had reported the same numbers they did, and people had been in fact using them at the same rate that they did, but NVIDIA hadn't known about it, then nothing illegal would have happened. If NVIDIA had later found out and disclosed this to investors, again no lawsuit (though perhaps share prices would have changed, one way or another).
The SEC lists some facts suggesting nvidia thought they were ending up mining (https://www.sec.gov/litigation/admin/2022/33-11060.pdf):
"NVIDIA launched a product line of cryptomining processors, known as “CMP,” which the company marketed to large cryptomining operations ... Based on known CMP sales, the company identified cryptomining as a significant element of the OEM GPU sales"
"NVIDIA also received information indicating that cryptomining was a significant factor in year-over-year growth in NVIDIA’s Gaming GPUs revenue. Some of the company’s sales personnel, in particular in China, reported what they believed to be significant increases in demand for Gaming GPUs as a result of cryptomining"
These were the 2 most concrete points to me - management was aware miners drove up GPU prices for gamers and saw it as another big market to capture, and sales attributed increased demand to miners. The rest of it basically says crypto grew, nvidia sales grew, and nvidia thought sales grwe because of crypto.
Honestly, I'm not sure it's possible even knowing these factors to know who really bought the GPUs or where they really ended up, but idk all the SEC probably cares about is nvidia not disclosing material seeming information.
Some even suspect that crypto mining is better for the card than gaming as it avoids the constant ramp-up/ramp-down of the clock/activity by keeping it pegged at 100%. In turn, that would lower thermal stress.
The difference in data here is obviously scale, google has -way- more CPUs than GPUs so the absolute counts of failures will be different.
If a given piece of silicon is hosing up a GEMM (matrix multiply), in graphics scenarios this may be invisible to the human eye as it could potentially just introduce artifacts in a scene rendering that could be entirely ephemeral to the frame.
In the case of crypto mining though, it's completely possible (probable?) that there are GPUs that can't possibly ever calculate a proper SHA3 hash (see the paper on AES instructions that fail in symmetric ways).
I think another comparison is with hard drives, which are also used by some cryptocurrency schemes, and do degrade faster with intensive use.
[a]: It's (hopefully) common knowledge in the tech community that purchasing used hard drives is a Bad Idea(TM)
If their data is correct, it should follow that these exact issues will happen on the small transistor process GPUs as well.
Blech, I. A. and Sello, H. (1966). The Failure of Thin Alluminum Current-Carrying Strips on Oxidized Silicon. In Proc. Symposium on PoF in Electronics, 496-505.
Power supplies on the other hand... those don't seem to last as long, but we open those up and they are hand soldered by someone in China.
Given their conclusion that this is the result of manufacturing tiny process transistors, it follows that this can occur in GPUs as well.
Using GPUs for mining doesn't destroy them any more than using GPUs to render p0rn videos doesn't destroy them.
I imagine the bearings in the fan might have more wear, however, but that might be more easily fixed.
When the GPU mining bubble pops, there’s a secondary effect as miners unload their used GPUs at fire sale prices on the secondary market. This means that the demand reduction is even worse than it might seem.
A few million is not a punishment it is encouraging Nvidia to do nasty things in the future since they know they can get away with it.
> A few million is not a punishment it is encouraging Nvidia to do nasty things in the future since they know they can get away with it.
Ok so tell me precisely how much NVDA benefited by not disclosing how much sales were to crypto?
I'm a NVDA stock holder and I just don't see what the fuss is. It's been impossible to buy a Nvidia card over the last 3 years due to scalpers and crypto farmers. I guess I thought everyone kinda knew that?
Do we know that actually? There was already outsized demand (demand > supply) for Nvdia GPUs to the extent that consumers literally couldn't buy GPUs for ~2 years (still is a problem) without buying from a scalper. A few things to consider:
(1) In most markets where demand outpaces supply, new entrants emerge. That is almost impossible to do in GPUs due to compatibility/support. So we've had pent up demand in that market.
(2) Consumers waited for supply to catch up to make the purchase even though they were ready to purchase.
(3) AFAIK there was no surcharge implemented by NVDA and this value was captured by scalpers. In fact, the price for performance ratio for the latest series of NVDA GPUs was so good that even if crypto wasn't a thing many consumers were ready to replace their old GPU.
(4) Crypto focused demand is a leading indicator of sales for that channel. With the shift to PoS, the overall demand for crypto rigs could shutter overnight.
My theory (as an NVDA investor) is that sales wouldn't falter because they were already selling products as fast as physically possible due to supply constraints. The question then therefore is what is the forecast 2-5 years out? NVDA is IMO heavily diversified in markets that have long term plays that easily could substitute any dip in consumer GPUs due to less demand from crypto.
TL;DR - I think people are overthinking this and if you don't have any idea of how to calculate what a meaningful penalty should be (i.e. "well the company was set to gain X amount of profit due to their infraction") then it's hard for me to take the argument seriously that $5M "isn't enough".
https://www.engadget.com/nvidia-sec-settlement-crypto-mining...
Its old news by a week or week and a half
What's also baffling to me is how their share price is taking a nosedive because of this news. Any investor should know about this. And $5 million fine doesn't sound like enough to hurt the company that much.
But if it's true that they misled investors, does that mean that investors can get their money back or something? It seems like any investor who wasn't aware of this is now screwed twice.
See for example case 18-cv-7669 by the "Ironworkers Local 580 Joint Fund" from late 2018, which was later merged with a class action lawsuit.
Nvidia talked about crypto in their 2018 quarterly filings but they allegedly downplayed the effect on the company.
Analysts upgraded their rating, growth prediction of 17%, shares trading at record price. (for the time, not comparable to today)
In the last quarter of 2018 crypto crashed, nvidia revenue dropped by 7%, and the share price by 29%
My view is that the banks did hugely mess up and did not face enough consequences — and the correct approach is to massively enhance enforcement. (In many ways that actually has happened.) But moreover they preyed on stupidity, just massive massive stupidity that will always be there to contend with. Fraud always looks better than real investments to the marks. People have to overlook the shady parts, and to make them do that a product must offer unrealistic returns. Millions and millions of people are not smart enough to tell when that’s happening to them. As long as the SEC and its ilk let these run, people get hurt and everyone suffers, either personally, or from the inefficiency of fraud generally or the collapse of interlinked financial systems when the fraud gets big enough. Dealing with this is a huge task and often unpopular. If you tell millions of Americans they can’t invest in crypto scams, many of them will be royally pissed off, because they were promised riches! That’s even worse than whack-a-mole — letting frauds off the leash for one minute leaves your hands tied unless you are top of your game and have a lot of built up trust.
That’s why people keep pointing to the lack of arrests as a huge failure of 2008. As mistakes by government go it was much worse than the bailout, which was rather necessary to limit the damage. It meant people thought the government was in on it, or that they didn’t care. Frankly I think Bitcoin would never have taken off if there were a slew of high profile arrests. It would not have seemed necessary.
That's only if your basic view of the financial markets is of something legitimate and socially beneficial. I would say that is mostly not the case; and in the past, the US government at times needed to recognize this at least partly, on pain of mass upheaval and system collapse, putting stronger regulation in place, e.g.:
https://www.federalreservehistory.org/essays/glass-steagall-...
(which isn't that strong either)
This is all before the complex derivative markets formed which gave rise to the 2008 crash. The housing situation in the US even now seems rather tenuous and volatile, whether through some financial crash or a different form of crisis:
https://slate.com/business/2021/06/blackrock-invitation-hous...
Securities Exchange Act of 1934 Securities Act of 1933
I think this is the problem you have investors buying stock in a company they have no concept what that company does they just know its a hot stock because the price is currently pumping. They go all in with no due diligence assuming the price will continue to pump for no reason and when it doesn't and the supply demand level out they act like they where scammed like the company should have told them when to buy and sell.
But: the stock market investors that are moving the really big bags of cash rarely have hours and days to spend trying to hunt down GPUs for their gaming rigs. So they wouldn't see these obvious hints.
https://web.archive.org/web/20220126135425if_/https://cdna.p...
EDIT : Never mind, still 2x prices, but seems to start coming down ?
https://cdn.wccftech.com/wp-content/uploads/2018/04/Asus-RX-...
[1] https://www.vijaypradeep.com/blog/2017-04-28-ethereums-memor...
I tried to restrict to US buyers to try to get it in the hands of someone who just wanted to play games.
They went to US addresses as my shipping restrictions and other exclusions required, but I unexpectedly got dinged for “International Fees” by eBay.
So I dug in as to why. When I checked the shipping address in Google Maps, they were warehouses. Googling again, known freight forwarders.
From what I can ascertain, one likely went to Qatar, and the other ended up in Uzbekistan. The accounts that bought them were quite old, not new, and had extremely high quantities of feedback. Based on the absolutely extreme auction end prices ($1540 for a 3070!), I wasn’t surprised in the end. The buyers were clearly extremely well organized and had exceptionally professional communication.
Their stock price doesn't tank on this news, it's trending down along with BTC and other crypto assets. The market isn't stupid, it's pricing into NVDA price the risk of cryptos going into a prolonged bear market.
Investors will probably never get their money back. First of all, NVDA stock is up a ton since 2018, and second of all, shareholders are considered full risk-on and should diversify. Class actions are possible but highly improbable to work. The company has been buying their stock back too which show their commitment to reward shareholders as well.
Secondly, from an investors point of view, GPU cryptomining demand is seen as a lot more volatile than PC gaming demand. A crypto crash or the planned ethereum switch to proof of stake could reduce this demand substantially. If a larger fraction of NVIDIA's sales (and thus revenue) is due to cryptomining, it's less likely to be similar next year (not only because cryptomining might buy less new cards, but also because they might flood the market with used cards, reducing demand for new cards from PC gamers, which happened during the last big crypto crash). This is the main reason the SEC is not happy about it. When reporting to investors NVIDIA also has an incentive to downplay crypto-related demand.
edit: this is an old story see previous discussion https://news.ycombinator.com/item?id=31284952
> the sale of its graphics processing units (GPUs) designed and marketed for gaming
SEC recognises that these cards where marketed to gamers and even recognise that the parts of NVIDIA marketed toward crypto mining was reported as crypto mining.
It seems like nvidia did business as usual and the real problem is those filthy gamers that aren't gaming.
If NVIDIA reported its gaming cards to SEC as crypto cards, now that the crypto market is dead and those cards are being used for gaming would SEC still consider NVIDIA to be in the wrong for their undisclosed impact on gaming?
Seems like SEC is mad at gamers for making money instead of gaming and they are taking it out on NVIDIA. Im just glad some one is standing up to these filthy gamers.
> “NVIDIA’s disclosure failures deprived investors of critical information to evaluate the company’s business in a key market,
1) Weird of you to add everything after your first sentence without indicating you edited your comment.
2) I have no idea what you're talking about. This isn't about nVidia reporting their cards as crypto cards. This is about nVidia not reporting what portion of their revenue is subject to volatility in the cryptomarket. SEC isn't mad about anything. It sounds like you're mad, tbh, and I'm not sure why.
Nvidia doesn't know what customers use their cards for and it especially doesn't know the volatility of the market, no one does. If nvidia knew market volatility before hand they wouldn't need to make gpu they could just trade the market more efficiently than every one else.
You and SEC are talking about pure speculation as if it was fact, nvidia didn't speculate in their reporting they gave accurate numbers as they should. Until you can show otherwise its all just speculation.
What proof do you have that the SEC is speculating about NVIDIA having this information?
As a gamer, part of me gets annoyed that crypto-miners make it more difficult to get gaming GPUS. On top of that, the massive energy consumption caused by crypto mining feels excessive.
But the freedom-loving part of me feels uncomfortable with solutions meant to prevent gaming GPUs from being used for crypto mining (let me use my hardware however I want).
Is there some sort of better, arm-chair-expert compromise?
Well rise of mining is somewhat a misunderstanding, because since I've been a frequent user of cryptocurrency since early 2011, from my point of view that's when GPU mining started but I digress.
But now it is obvious to me. It is better to have in your earnings gpu sales because you can always claim that you expect a steady source of income. If they had 60% of their revenue explicitly from a volatile market like crypto mining, then their stock would not look attractive.
Their performance was also luck luster compared to the consumer rtx cards.
So what the SEC is fining them for is lying to investors. If nvidia didn't know how much of their sales were to cryptomining, then there would be no fine. If they did know, and did share that in their reports, there would be no fine.
This is not a fine for "You didn't know how your product was being used" but "you knew it was being used for crypto but reported to investors that it being was gaming"
But then lo and behold he traveled to Spain, a country which recognizes nobility, and tried arrogancing his way into a party (like "what do you mean I'm not on the list!?"). The press asked who the fuck is he, asked his parents, his parents weren't his parents, his parents had no idea who the fuck he was. Said as much publicly. "I have no fucking idea who this asshole is" said his father.
And then he was on TV in 2017 or 2018 as the world-class bullshitter of all time, but with much contempt. He had hours of talk-shows just for him, his law school professor going in front of the camera saying what he did was illegal because he pretended to have authority. He was a scammer, in particular he posed as a human-rights lawyer targeting victims of torture to get their secrets and extort them. Tried to do it to me, but you can't con an honest John.
So what happened, going back to your point? He brainwashed himself, told himself he was a prince 10 million times. But lied to himself literally 10 million times.
I think it's Miguel von und zu Liechtenstein, I misspelled it earlier, it's a tricky one. Said his name was Mikhail legally but when he flashed both copies of his Chilean ID it was Miguel, not Mikhail.
https://www.gamba.cl/2018/05/la-decadente-historia-de-miguel...
It was also in the papers, like Publimetro, and El Mercurio. Also on a morning show, https://www.youtube.com/results?search_query=miguel+de+liech..., TVN I remember. 30 minute show about him, which I have no problem with it's righteous justice and juicy gossip at the exact same time, win-win. If only all television were that good in every sense.
That's how I found out he was full of shit, I saw his face on a television at a cornershop and it all made sense. But stayed in character to a literally insane degree, and was very consistent about it. He even convinced high-ranking politicians, the President, he went all out. Shot the moon. And it sort of worked out, hey! Isn't that remarkable, actually pulled it off haha, good one.
He did his damndest to get everyone's hatred, too, particularly with that foundation trying to outlaw hate speech and instigation of violence, as a front to commit extortion and virtue signaling. Yeah because then he got to talk to people giving him early warning of pitchforks, could do man-in-the-middle attacks by plagiarizing his victims pleas to elicit the pity that was not rightfully his, and gating access to government help to clean his victims out.
Human-rights corruption, like stealing candy from a baby!
You are calling victims of tortures liars?
And one could argue that gaming is also rendering / computing / solving problems.
It's about what NVidia told their stockholders in reports.
Rather than: "This years growth is fueled by sales to the cryptominer market" (which is a pretty dang volatile market to sell to as a crash will floor nvidia's market with cheap used cards rather than selling overinflated new ones...)
They lied and said: "oh yeah no, totes all those gamers buying cards for sick rigs", and the pc gamer market is much more stable, it's predictable, you know roughly when / how much of the market will be buying new to replace a card that's now too old/limited.
tl;dr they lied about who they were selling to, because the volatility (no guarantee of that market still existing in a year or two's time) would make shareholders understandably nervous.
It seems like they didn't lie they sold gpu and reported the correct number of gpu sold. The only claim nVidia made was that it marketed these cards to gamers which the SEC acknowledges they did, this isn't a lie.
> In two of its Forms 10-Q for its fiscal year 2018, NVIDIA reported material growth in revenue within its gaming business. NVIDIA had information, however, that this increase in gaming sales was driven in significant part by cryptomining. Despite this, NVIDIA did not disclose in its Forms 10-Q, as it was required to do, these significant earnings and cash flow fluctuations related to a volatile business for investors to ascertain the likelihood that past performance was indicative of future performance.
The issue isn't the number of GPU sold, noone's saying they lied about how many they sold or their profit. They lied about the market conditions for that quarter, to make the growth seem "stable", not from a volatile market that could dry up in 2 years or 2 months...