Let’s talk about ASIC mining

#5788

Grin has two variants. One is asic friendly one is not (touted to be)

#5789

why not? it’s quite possible with controlling the supply-demand chain especially when you’ve only few distributors/retailers in a country.

#5790

Ohh, look…US has 965Msol/s against 1507Msol/s…

#5791

Edit: related comments from MoneroCrusher who performed the network analysis:

3 Likes
#5792

Here the detailed analysis and approach how they found it out, very interesting by the way …

Estiminated 5400 Asics there …

2 Likes
#5793

ASICs are a powerful force in this game as they can, un-intuitively, guard against 51% attacks.

#5794

imagine such comment here in 2017…

zooko, what have you done?

2 Likes
#5796

Are ASICs dead???

1 Like
#5797

No, they are not dead. Monero attempted to be ASIC resistant, forked, and for short time it was. Now there are reports that ASIC’s are back on that coin at a greater percentage than before. The difference now is that who ever has them will remain silent about it.

So making Zcash ASIC resistant, unless a complete revamp of how things are done, is pointless.

6 Likes
#5798

That is of course an assumption that it’s secret ASIC…it could be worse, it could be FPGA’s. Good luck bricking those off.

3 Likes
#5799

if asic is 50%, or 100% more efficient that GPU, it’s one situation.

other situation when asics that are 1000% more efficient are announced and dev-team of “asic-resistant” coin does absolutely nothing about it, zero reaction.

we should speak about how insane are zcash dev’s.

1 Like
#5800

The mining space changes so rapidly that it’s hard to predict what the next innovation will be. There is a clear demand to make a device that is both flexible, adaptable and faster than common GPUs. Even small percentage advantages might be worth it for those entities at scale. It’s a toss up whether or not those entities will choose to sell to the public. My thought is that they won’t, finding it more profitable to keep innovations secret and competing with other well funded entities doing the exact same thing. Public mining days might be numbered, which is a sad proposition.

That said, I think it will probably take 5-10 more years for the mining industry to fully mature to a state of predictability. At that time we will be in an entirely different class of hardware focused on extremely low power usage and highly efficient cooling designs. This will most likely shift to two-phase immersion cooling with novec-like fluids (these are really cool btw). As non-mining data centers move to these more efficient technologies, to get higher density per sf, the prices will decrease dramatically for other industries like the crypto mining space. Advancements happening are both fascinating, exciting and also at times worrisome considering our goals of a fairer coin distribution alongside a highly decentralized and secure protocol.

#5801

"The exploit allows an attacker to do basically anything, including modifying the payout address of an exploited miner. "

1 Like
#5802

Maybe devs should include by now a new attack vector, (mass) infected Asics …

#5803

Once @BITMAINtechcomplies with the GPL licenses for the firmware I will disclose the vulnerability to them so that they can fix it. pic.twitter.com/zwsAaPQjRL

— James Hilliard (@james_hilliard) February 12, 2019"

Extremely unprofessional of James Hilliard. When I read that it instantly erased any respect for him as a security researcher. I understand his position with respect to GPL, but there is a better way to go about it in a non-public way.

Also, something is slightly off about you posting this @bitcartel. If I’m not mistaken, you are a BU member, which is pro on-chain scaling of Bitcoin. I’m not sure why you would promote James Hilliard who fought that scaling initiative before the Bitcoin split.

#5804

I posted a link to CCN.

If you take a look, TheBlockCrypto, BitcoinMagazine and CoinGeek also mention James Hilliard, because, well, he discovered the vulnerability.

It would be impossible to write a story about the issue without mentioning him!

5 Likes
#5805

So innosilicon now has a ASIC “GRIN”

Grin does not implement a flashy new consensus mechanism like proof of stake to achieve consensus. Instead, it goes back to the bread and butter of PoW using the Cuckoo Cycle algorithm.

Cuckoo style PoW was selected to mitigate against the Bitcoin-style “hardware arms-race” by making it ASIC resistant. Cuckoo Cycle is a memory-bound algorithm, making it viable for CPUs and increasing its decentralization.

The difficulty of mining in the network is based on the current hash power and is designed to average a fast block time of around 60 seconds. You can find extensive information on Cuckoo Cycle PoW in the white paper by John Tromp and Grin mining on the Grin Github.

1 Like
#5806

It used to be considered ASIC resistant, but I’ve realized for some time now that ASICs sporting tons of SRAM have a huge efficiency advantage over GPUs. Thus, Grin’s primary PoW of cuckatoo31+ is made as ASIC friendly as a memory bound PoW can be.

Even Grin’s ASIC resistant secondary PoW, cuckaroo29, required some special measures to disincentivize lean mining, and can only maintain its resistance by frequent tweaks (Monero style).

4 Likes
#5807

In Central Alberta I am on a floating power rate. Generally most electricity providers here offer a floating rate that goes from $0.03 to $0.08 but averaging around 3 to 4 cents.

There is a flat rates around 5 to 7 cents per KWH but I do not know anyone who uses that and most my economic friends insist the floating rates are cheaper on average.

#5808