mike at powx.org [ARCHIVE] / is on Nostr

⚠️ This account is not associated with a real person. It's an archival account for mailing list messages.
🗓️ Active from 2021-05-18 to 2021-05-18
💬 Posted a total of 4 messages
📚 Participated in the following categories: bitcoin-dev
🔇 Please do not DM this user as it's used for archive purposes only.

Last Notes

2023-06-07 22:53:45

- reply

📅 Original date posted:2021-05-18
📝 Original message:Devrandom is correct to point out that there is nuance to these things and it’s better to look at the details rather than proclaiming that PoW is PoW. (I do agree though w the original point that other ideas often turn out to reduce to PoW despite their convoluted architecture)
A note on the key difference between hardware and energy as it relates to centralization:
Hardware is easily transferable. If you have low electricity costs, Bitcoin ASICS need to be physically located in proximity to use it. You can’t sell your low power costs on an open liquid market (you can sell your hash rate but that still requires all the miners next to the power plant). Hardware can be sold online freely to anyone anywhere in the world.
If a small number of foundries are producing low energy opow hardware (just as there are a small number producing SHA256 ASICS- in fact it would be the same set foundries, somewhat expanded because optical chips use larger, older nodes... for example Global Foundries has a great photonics process at 90nm), they can (and will) still sell the hardware to people all over the world.
There is a huge latent demand for BTC mining. Many people currently buying alt coins or even BTC would prefer to invest in mining if they could turn a profit despite their high energy cost.
Another clear benefit would be the difficulty of detecting and controlling low energy mining relative to the ASIC-warehouse-next-to-waterfall model used today. You can’t move a waterfall if the local government decides to regulate you.
Just some thoughts.
Sent from my iPhone
> On May 18, 2021, at 5:18 AM, Devrandom <c1.bitcoin at niftybox.net> wrote:
>
>
> On Mon, May 17, 2021 at 11:47 PM ZmnSCPxj:
>>
>> When considering any new proof-of-foo, it is best to consider all effects until you reach the base physics of the arrow of time, at which point you will realize it is ultimately just another proof-of-work anyway.
>
> Let's not simplify away economic considerations, such as externalities. The whole debate about the current PoW is about negative externalities related to energy production.
>
> Depending on the details, CAPEX (R&D, real-estate, construction, production) may have less externalities, and if that's the case, we should be interested in adopting a PoW that is intensive in these types of CAPEX.
>
> On Mon, May 17, 2021 at 2:20 PM Keagan McClelland wrote:
>
>> First it just pushes the energy consumption upstream to the chip manufacturing process, rather than eliminating it. And it may trade some marginal amount of the energy consumption for the set of resources it takes to educate and create chip manufacturers. The only way to avoid that cost being funneled back into more energy consumption [...]
>
> I challenge you to substantiate these assertions. Real-estate and human cognitive work are not energy intensive and are a major factor in the expected costs of some alternative PoWs. The expected mining effort is such that the cost will reach the expected reward, no more, so there is every reason to believe that energy consumption will be small compared to the current PoW.
>
> Therefore, the total associated negative externalities for the alternative PoWs may well be much lower than the externalities of energy production. This needs detailed analysis, not a knee-jerk reaction.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linuxfoundation.org/pipermail/bitcoin-dev/attachments/20210518/35df2a71/attachment-0001.html>

2023-06-07 22:53:45

- reply

2023-06-07 22:53:43

- reply

📅 Original date posted:2021-05-18
📝 Original message:Nothing in a dynamic system like PoW mining can be 100% anticipated, for example there might be advanced in manufacturing of chips which are patented and so on.
It sounds like your take is that this means no improvements can ever be made by any mechanism, however conservative.
We do go into a fair amount of detail about Minimum Effective Hardness in our paper https://assets.pubpub.org/xi9h9rps/01581688887859.pdf , which is actually a special case of hardness that we invented for the context of adding an operation to a PoW, and how it applies to random matrix mults.
Sent from my iPhone
> On May 18, 2021, at 7:58 AM, ZmnSCPxj <ZmnSCPxj at protonmail.com> wrote:
>
> Good morning Michael,
>
>> That’s interesting. I didn’t know the history of ASICBOOST.
>
> History is immaterial, what is important is the technical description of ASICBOOST.
> Basically, by fixing the partial computation of the second block of SHA256, we could selectively vary bits in the first block of SHA256, while reusing the computation of the second block.
> This allows a grinder to grind more candidate blocks without recomputing the second block output, reducing the needed power consumption for the same number of hashes attempted.
>
> Here is an important writeup: https://www.mit.edu/~jlrubin/public/pdfs/Asicboost.pdf
> It should really be required reading for anyone who dreams of changing PoW algorithms to read and understand this document.
>
> There may be similar layer-crossings in any combined construction --- or even just a simple hash function --- when it is applied to a specific Bitcoin block format.
>
>>
>> Our proposal (see Implementation) is to phase in oPoW slowly starting at a very low % of the rewards (say 1%). That should give a long testing period where there is real financial incentive for things like ASICBOOST
>>
>> Does that resolve or partially resolve the issue in your eyes?
>
> It does mitigate this somewhat.
>
> However, such a mechanism is an additional complication and there may be further layer-crossing violations possible --- there may be an optimization to have a circuit that occasionally uses SHA256d and occasionally uses oPoW, that is not possible with a pure SHA256d or pure oPoW circuit.
> So this mitigation is not as strong as it might appear at first glance; additional layers means additional possibility of layer-crossing violations like ASICBOOST.
>
>
>
>
> Regards,
> ZmnSCPxj
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linuxfoundation.org/pipermail/bitcoin-dev/attachments/20210518/d8b1d62e/attachment-0001.html>

2023-06-07 22:53:43

- reply

📅 Original date posted:2021-05-18
📝 Original message:That’s interesting. I didn’t know the history of ASICBOOST.
Our proposal (see Implementation) is to phase in oPoW slowly starting at a very low % of the rewards (say 1%). That should give a long testing period where there is real financial incentive for things like ASICBOOST
Does that resolve or partially resolve the issue in your eyes?
Sent from my iPhone
> On May 18, 2021, at 7:36 AM, ZmnSCPxj <ZmnSCPxj at protonmail.com> wrote:
>
> Good morning Michael,
>
>> That’s a fair point about patents. However, note that we were careful about this. oPoW only uses SHA3 (can be replaced with SHA256 in principle as well) and low precision linear matrix multiplication. A whole industry is trying to accelerate 8-bit linear matrix mults for AI so there is already a massive incentive (and has been for decades).
>>
>> See companies like Mythic, Groq, Tesla (FSD computer), google TPU and so on for electronic versions of this. Several of the optical ones are mentioned in the BIP (e.g. Lightmatter)
>
>
> Please note that ASICBOOST for SHA256d is based on a layer-crossing violation: SHA256 processes in blocks, and the Bitcoin block header is slightly larger than one SHA256 block.
>
> Adding more to a direct SHA3 (which, as a "sponge" construction, avoids blocks, but other layer-crossing violations may still exist) still risks layer violations that might introduce hidden optimizations.
>
> Or more succinctly;
>
> * Just because the components have (with high probability) no more possible optimizations, does not mean that the construction *as a whole* has no hidden optimizations.
>
> Thus, even if linear matrix multiplication and SHA3 have no hidden optimizations, their combination, together with the Bitcoin block header format, *may* have hidden optimizations.
>
> And there are no *current* incentives to find such optimizations until Bitcoin moves to this, at which point we are already committed and it would be highly infeasible to revert to SHA256d --- i.e. too late.
>
> This is why changes to PoW are highly discouraged.
>
>
> Remember, ASICBOOST was *not* an optimization of SHA256 *or* SHA256d, it was an optimizations of SHA256d-on-a-Bitcoin-block-header.
> ASICBOOST cannot speed up general SHA256 or even general SHA256d, it only applies specifically to SHA256d-on-a-Bitcoin-block-header.
>
> Regards,
> ZmnSCPxj

2023-06-07 22:53:43

- reply

2023-06-07 22:53:43

- reply

2023-06-07 22:53:42

- reply

📅 Original date posted:2021-05-18
📝 Original message:That’s a fair point about patents. However, note that we were careful about this. oPoW only uses SHA3 (can be replaced with SHA256 in principle as well) and low precision linear matrix multiplication. A whole industry is trying to accelerate 8-bit linear matrix mults for AI so there is already a massive incentive (and has been for decades).
See companies like Mythic, Groq, Tesla (FSD computer), google TPU and so on for electronic versions of this. Several of the optical ones are mentioned in the BIP (e.g. Lightmatter)
Sent from my iPhone
> On May 18, 2021, at 6:59 AM, ZmnSCPxj <ZmnSCPxj at protonmail.com> wrote:
>
> Good morning devrandom,
>
>> On Mon, May 17, 2021 at 11:47 PM ZmnSCPxj:
>>
>>> When considering any new proof-of-foo, it is best to consider all effects until you reach the base physics of the arrow of time, at which point you will realize it is ultimately just another proof-of-work anyway.
>>
>> Let's not simplify away economic considerations, such as externalities. The whole debate about the current PoW is about negative externalities related to energy production.
>>
>> Depending on the details, CAPEX (R&D, real-estate, construction, production) may have less externalities, and if that's the case, we should be interested in adopting a PoW that is intensive in these types of CAPEX.
>
> Then let us also not forget another important externality: possible optimizations of a new PoW algorithm that risk being put into some kind of exclusive patent.
>
> I think with high probability that SHA256d as used by Bitcoin will no longer have an optimization as large in effect as ASICBOOST in the future, simply because there is a huge incentive to find such optimizations and Bitcoin has been using SHA256d for 12 years already, and we have already found ASICBOOST (and while patented, as I understand it the patent owner has promised not to enforce the patent --- my understanding may be wrong).
>
> Any alternative PoW algorithm risks an ASICBOOST-like optimization that is currently unknown, but which will be discovered (and possibly patented by an owner that *will* enforce the patent, thus putting the entire ecosystem at direct conflict with legacy government structures) once there is a good incentive (i.e. use in Bitcoin) for it.
>
> Regards,
> ZmnSCPxj

2023-06-07 22:53:42

- reply