What’s Up With ProgPoW? Benchmarks & Addressing the Criticisms

ProgPoW is a new algorithm that was proposed to replace Ethash and therefore make Ethereum ASIC-resistant. The idea is being debated since August 2018 and apparently, we are coming to the point when we as a community have to make the final decision – do we want ProgPoW to be implemented or not?

The idea was also discussed during January 4 ETH Core devs meeting and it seems that the majority of the community and the developers team members are leaning towards the adoption of this new algorithm.

In this post, I will briefly go over what ProgPoW is and then will address the criticism this idea is being receiving. Again, as a community, it’s our duty to be aware of what is going on and to actively either support or oppose this new proposal.

If you are already familiar with ProgPoW, you can jump directly to any of the following criticisms:

What is ProgPoW

ProgPoW is a modification of the Ethash algorithm that is specifically tailored to be mined with GPUs. But isn’t Ethash a GPU-friendly algorithm? Apparently, it is not. The problem with Ethash is that the GPU that mines cannot use all of its resources. According to the developers of ProgPoW, “The SM’s […] consume most of the GPU’s die area. They run at less than 30{520ebc7d04367f10067c097a94259816cd1e6df76ec40212834858ca2e1af940} utilization.”

ProgPoW is designed to bring those 30{520ebc7d04367f10067c097a94259816cd1e6df76ec40212834858ca2e1af940} all the way up to 90{520ebc7d04367f10067c097a94259816cd1e6df76ec40212834858ca2e1af940}, allowing GPUs to unleash their almost full potential when mining. Even though ProgPoW is a modified version of Ethash, the differences between the two algorithms are pretty subtle and they do not compromise security in any way.

Why does ProgPoW exist?

Many of the Ethereum community members are not happy with the way ASICs are taking over Ethereum. Hell, even investors start to turn their faces away from Ethereum now that it can be mineable by ASICs, according to Business Insider.

For now, Ethereum ASICs aren’t that powerful, however they might become pretty beefy, pretty soon. All an ASIC needs to mine Ethash is a lot of memory bandwidth with a small compute core. Linzhi announced that they are developing a new miner that will outperform (more than 7x performance at 1/8th of the power usage) all existing Ethereum ASICs and I am sure that others (especially Bitmain and Innosilicon) aren’t far behind either.

ProgPoW was designed to prevent ASICs from taking over Ethereum mining hashrate. If Ethereum will switch from Ethash to ProgPoW, it will become pretty much ASIC-resistant, though not on a conventional way.

What I mean is that by being fully optimized to mine ProgPow, GPUs become better at it than any existing or future ASICs. In other words, while it is possible to develop and ASIC for ProgPoW, it’s pretty much impossible (at least on paper) to make it significantly more efficient than a GPU.. And since GPUs are more accessible and more flexible, people will more likely stick to their GPUs when mining Ethereum, even with ASICs being available.

Besides, thanks to the structure of ProgPoW, GPU overclocking is no longer necessary. You can read about how it works here. GPU underclocking is possible and I am sharing my benchmarks down below.

Will ProgPoW Be Incompatible with Existing Ethash ASICs?

ProgPoW is incompatible with existing ASICs for Ethash. Even though ProgPoW is very similar to Ethash in many ways, any ASIC manufacturer will have to have a completely different approach if they will ever want to design an ASIC for this new algorithtm.

One of the reasons is that ProgPoW adds a sequence of random math instructions and random cache reads to make a much larger mix state. The bigger is the mix state and the more random, the harder it is to build fixed function hardware. While definitely not impossible, it will take a lot more effort to make an ASIC for ProgPoW. And most importantly, those new ASICs will have to have a structure very similar to a GPU, with similar manufacturing costs and performance.

Regarding FPGAs, ProgPoW should be able to be resistant to those too. Unlike ASICs, FPGAs can be reprogrammed to mine virtually any algorithm, however changing the bitstream takes time and since the algorithm change will be happening every 50 blocks (every 12 minutes), using an FPGA to mine ProgPoW should result pretty problematic and and inefficient, or even impossible (especially after specs 0.9.3 and on).

ProgPoW Team

First and foremost, nobody from the Ethereum core team is officially involved in this project. ProgPoW is a community initiative and as such, it has its own activists.

One of ProgPoW most active teams is IfDefElse on Github. Their team allegedly consists of three people, two of which decide to keep their identities secret. Besides their activity on github, their role is to find or hire an external auditor for ProgPoW.

The third and only known IfDefElse member is Kristy-Leigh Minehan. Going under the nickname “Miss If”, Kristy-Leigh Minehan is the chief technology officer of Core Scientific, founder of the Mineority Group and also ex Genesis Mining. She is also known for participating in the creation the famous ETHlargement tool that increases Ethereum mining performance on certain GPUs.

The Critics

For a lot of people, the idea of bringing Ethereum back to GPU miners seems pretty attractive. In fact, not only GPU rig owners like that idea – apparently, investors seem to associate ASICs with mining centralization, whereas GPU mining allows anyone with a computer to help the network to stay decentralized. ASICs are expensive and fewer people invest in them, while GPUs are popular and used by common folk.

Ethereum Cat Herders, the community assigned to evaluate/audit ProgPoW by ETH Core Devs, made a poll on twitter to see whether the crypto community supports the idea. While the results are in favor of ProgPoW, I will not take that poll into account, simply because Twitter polls can be easily manipulated.

We are keen to gauge a rough consensus from the community on ProgPoW and as such will be asking some questions over the next week or so to get some additional data. To kick things off…

Do you think ProgPoW should be implemented? Please reply below to give additional details.

— Ethereum Cat Herders (@EthCatHerders) February 6, 2019

Instead, I will try to explore some of the most common concerns people post in the comments and discuss on the forums regarding this new algorithm and the way it might affect Ethereum.

Critique #1. ProgPoW is a “front” for Nvidia and AMD interests because Kristy-Leigh Minehan has ties to them.

According to her own words, Kristy-Leigh Minehan has close ties with the two GPU manufacturers. She also openly stated that both companies reviewed the code of this new algorithm. It would make sense to suppose that she started to lead the development of this project sponsored by Nvidia and AMD.

Mining had a great impact on both GPU manufacturers. Just take a look a year back when the price for mining-capable GPUs was about to reach the stars. Gamers were mad at miners for a reason – neither GPU manufacturer could satiate the demand for GPUs. Back in 2017, Ethereum used to be one of the main coins to mine on a GPU, mostly because Ethereum is big, popular and it’s way easier to sell your mined ETH.

It would make sense to suppose that now that GPU mining is slowing down, so are Nvidia and AMD sales, which is why they have planned to bring Ethereum mining back.

Personally, I find this critique pretty reasonable. However, I don’t mind Ms. Minehan being sponsored as long as the result of her work will benefit the network. After all, we miners don’t support the network for free either. So my main question is: will ProgPoW benefit Ethereum? Some people think it won’t and here is why:

Critique #2. ProgPoW will force ASIC manufacturers to keep their specialized hardware secret.

Simply put, if the community will agree to declare war on ASICs, ASIC manufacturers will have to officially declare that they gave up on developing miners for Ethereum. Because if they announce that they found a way to make an ultra efficient ASIC for ProgPow, Ethereum devs might make another change to the algorithm to kill those new ASICs. It happened with other coins (such as Monero) so it might happen with Ethereum too.

As we know, ASIC manufacturers make a living by mining with the ASICs they manufacture. Some giants go as far as using the ASICs they make to mine on their farms for few months before selling them (which spawned a lot of critique and controversy). Others openly keep most of the ASICs they make to themselves.

Mining centralization is one of the biggest fears of any crypto. Mining centralization defeats the main purpose of the cryptocurrency, making it worthless.

Now, if a coin is openly ASIC-friendly, does that keep it from being centralized by one company? Not really.

A good example is the recent SIA Coin incident. Originally, SIA Coin was designed to be ASIC-friendly, which motivated big manufacturers to develop and release their miners. Innosilicon was the one to win the race by creating the SIAMaster. The miner was so efficient and profitable that the company decided to not sell it to the public and keep the units in their farms instead (they sold a few batches but that’s about it). This allowed Innosilicon to quickly take the lead and own up to 43{520ebc7d04367f10067c097a94259816cd1e6df76ec40212834858ca2e1af940} of the SIA network hashrate, which was then stopped by a hard fork.

My take away from this case (among others) is that just because a coin is openly ASIC-friendly it does not mean that everyone will play fair.

Will ASIC manufacturers try to make ASICs for the new ProgPoW algorithm? Definitely, yes. Will they keep those units secret? Of course. Do they already do that with Ethash? Very likely.

The case with Innosilicon and SIA shows us that is an ASIC manufacturer wants to play dirty and put their own profit at the expense of decentralization, they will do it no matter what.

Note that I’m not hating on Innosilicon or on any other ASIC manufacturers. Business is business. However, I conclude that keeping an algorithm openly ASIC-friendly does not save it from being abused by giants. Back in the day, Innosilicon wasn’t bragging to the public that they owned the lion’s share of SIA network hashrate. It took some work and research for people to realize what’s up and ring the bell.

Critique #3. ProgPoW is a distraction from POS.

Ethereum 2.0 has been WIP for a long time now and every time a new update is delayed, people get even more impatient to see the POS implementation. Ethereum devs face a lot of challenges and they have pretty large to-do lists, so why make them even larger by requesting the ProgPoW change?

While this point seems valid, it’s good to remember that developers working on implementing ProgPOW are not the same developers working on ETH 2.0.

Who are they? We don’t really know, though they might as well be some of the people who work on Parity or Geth. Both projects already spent quite a bit of time to implement ProgPOW in their clients: https://github.com/paritytech/parity-ethereum/pull/9762. Open Ethereum Pool (open-source mining pool) also joined the party and is now supporting the ProgPoW algo.

Moreover, ProgPOW testnet is already live and as Alison Berreman from Ethnews pointed out, “Ethereum developments are notorious for arriving later than expected, but the short timeline suggests that concerns that ProgPoW is a distraction may be somewhat hollow – and that time for debate has ended.”

If it was a distraction, it probably no longer is and no matter whether we will agree or not to implement ProgPoW, this will not affect the development of Ethereum 2.0. Or at least not as much as some people suggest it will.

Critique #4. ProgPoW only exists to give GPU Miners some sort of privilege at the expense of ASIC owners.

Basically, why even bothering with the fork if its only purpose is to favor a certain group within the mining community?

As a decentralized cryptocurrency, Ethereum should not take any sides, I totally support that. Making such big changes just to please the GPU mining community is not the way to go.

However, there is also one fact that most people seem to ignore. In the original whitepaper, Ethereum was designed to be an ASIC-resistant, GPU-mineable coin, which is why it was based on Ethash.

It’s not about favoring a certain group of people, it’s more about staying true to the original plan, the original concept of what Ethereum is. Here’s another interesting fact.

Back in the day when Bitmain announced their first Ethash ASIC and the community started to discuss ways to prevent that, Vitalik Buterin’s answer was “we will not focus on making Ethereum ASIC-resistant because we will be switching to POS soon anyway and we honestly don’t have time to split our efforts and work on changing our PoW algo.”

That was said… a while ago, Ethereum 2.0 was delayed few times since then and we will probably still have PoW for a bit (some estimate for 2-3 years from now on). Besides, as we discussed in Critique #3, ProgPoW is already here, almost ready to be implemented. It was developed by an independent team, so it’s not really a distraction now.

Critique #5. Why killing ASICs if they make the network more secure?

We have Ethereum Classic as an example – it wasn’t needed a lot of hashpower to successfully attack the network.

While this point is perfectly valid, here is the catch. The higher the network hashrate is, the more expensive (and less likely) it becomes for a hostile force to perform a 51{520ebc7d04367f10067c097a94259816cd1e6df76ec40212834858ca2e1af940} attack on it.

Ethereum’s network hashrate is about 300x times higher than ETC’s. And paradoxically, it might get even bigger once ASICs are banned (as we can see in 2018 when Ethereum GPU mining was flourishing).

The Chinese giant Linzhi, among others, is actively developing a more powerful ASIC to mine Ethash. They have spent $4 million dollars so far and if all goes as expected, the new model can potentially revolutionize Ethereum mining – according to the words of the company spokesman anyway.

Now, how powerful those new units will be? And what other companies are working on a new ASIC design for Ethash? What if a company will come up with such a powerful ASIC that will allow the manufacturer to quickly take over the lion’s share of the network hashrate? I mean, it happened before with other coins such as SIA (which is not a small coin), so it could, in theory, happen to Ethereum too. And let’s not forget that Ethereum is a way juicier target than SIA.

My takeaway from this is: on one hand, ASICs can potentially increase the security of a coin, but this mostly works if the coin is small enough (like ETC). However, ASICs can be also potentially dangerous to a coin that is in the process of transitioning from GPU to ASIC mining. Yeah, Ethereum ASICs are still in their baby state and aren’t that powerful compared to GPUs, however this can change pretty soon.

Critique #6. ProgPoW might cause unnecessary hard forks.

Hard forks are unhealthy simply because they shard the community. Over the past month, a pretty big share of the community invested into Ethereum ASICs. Ethereum ASICs offer a way better value for the money than conventional GPUs, which makes them pretty attractive.

ProgPoW will make those ASICs incompatible with Ethereum, which will indeed cause a lot of complaints and anger. Similar cases happened to other coins that decided to ditch ASICs – once Monero hard forked to brick Monero ASICs, a hard fork called Monero Classic was created. Same thing happened to SIA – Bitmain and Innosilicon ASIC owners went on to create SIA Classic.

It’s not healthy for a community to split like that, which is an argument people use against ProgPoW.

While their point is completely valid, let’s also not forget about the upcoming Ethereum 2.0 hard fork. There is a different point of view which I cannot disagree with either.

Basically, even though the Ethereum mining community might split with the introduction of ProgPoW, it might make things less painful in the long run.

You see, once Ethereum becomes POS, none of the existing ASICs will be able to mine it. Yeah, there are indeed other coins based on the Ethash algorithm, however they aren’t as big or as profitable as Ethereum.

If ProgPoW will not happen, those people who invested their money on Ethash ASICs will have to mine something else once Ethereum 2.0 is released. We are talking about hundreds of thousands ASICs that will need an application. This means that thousands of people will stop mining Ethereum and will either fork Ethereum or go mine a different existing coin.

Now if ProgPoW will bring GPU mining back to Ethereum, the day of the fork will be way less painful for the mining community. People will simply switch their GPUs to mine something else (like Grin or Ravencoin), or maybe even lend their GPU farms to Golem, who knows.

Critique #7. ProgPoW Will Not Stop ASICs

Some people are concerned that ProgPoW will still be mineable with ASICs. Bitmain, among others, is known for breaking into so-called “ASIC-resistant” algos – they did it many times (Dash, Ethereum, etc) and they will certainly keep doing that as long as it’s profitable.

Truth is, it is indeed possible to create ASICs for ProgPoW and its developers are the first ones to aknowledge that. So what’s the catch?

According to the official technical specifications, the new algorithm is tailored to be the most efficiently mineable by GPUs. While an ASIC can be made to mine it, “This specialized ASIC would look suspiciously similar to existing commodity GPUs. It would only be marginally smaller and would have similar power performance.”

Basically, ProgPoW is tailored to be mined by GPUs and any device that wants to mine it efficiently has to be pretty much a GPU. Bitmain and others can manufacture GPUs indeed, however then they would be competing with Nvidia and AMD, which means that they would not have such an unfair advantage over common folks like you and I.

How Much Will ProgPoW Benefit GPU Mining?

GPU mining is and will be alive with or without ProgPoW – there are plenty ASIC-resistant coins to mine and Ethereum isn’t as pricey as it used to be in early 2018, which means it’s not as profitable to mine.

Let’s also not forget about the 3 to 2 ETH per block reward decrease in the upcoming Constantinople Fork. Ethereum is steadily moving to POS and it will get there eventually, leaving both ASIC and GPU mining behind in the process.

Will ProgPoW make Ethereum the GPU miner fest it used to be? I doubt it, unless Ethereum prices will skyrocket once again. But will they?

According to Coinswitch, the Constantinople Fork might make investors trust Ethereum a little bit more, which will positively affect the market.

Another factor that could potentially affect positively the price of the coin is the ban of ASICs. Quoting Business Insider, the announcement of the first ASIC for Ethereum made investors weaken faith in the crypto. Getting Ethereum from ASIC-friendly back to ASIC-reistant might or might not make more people invest into the coin.

Why the Lower Hashrates?

As you will notice in the benchmarks below, the mining hashrate of ProgPoW is avour 50{520ebc7d04367f10067c097a94259816cd1e6df76ec40212834858ca2e1af940} lower than that of Ethash. A GPU that hashes 20Mh/s on Ethash will hash around 11 Mh/s on ProgPoW. The reason is simple – ProgPoW is a different algorithm, which is why the hashrate is measured differently too. By switching to ProgPoW, your GPU will not mine slower and you will not earn less ETH. It’s just the network hashrate numbers will change both for GPU and ASICs, that’s all.

Testing ProgPoW

Since the testnet is live, I have decided to test the new algorithm myself. According to the developers, ProgPoW should work equally well with both AMD and Nvidia cards. In fact, in the official presentation, Krity-Leigh Minehan mentioned that one of their main aims was to avoid giving an advantage to either manufacturer.

Nvidia GPUs ProgPow Hashrate Benchmarks

OS: Windows 10 PRO

Drivers: 419.17

Miner Used: https://github.com/minerideseama/Ethereum-ProgPow-test/releases

Block: 7280000

(here you can make a donation for devs 0xa7e593bde6b5900262cf94e4d75fb040f7ff4727 and here for the guys who made this possible for windows 0xd9331260fb214F9Dd00c90873FF6B0c5ad2A60dE)

Compiled version: ethminer 0.18.0-alpha.3+commit.260c47d2
Source: https://github.com/AndreaLanfranchi/ethminer
Miner compiled on WIN10 64 with CUDA 10.

If you already done your own test, please comment below your results and also submit them here: https://progpow.pro/submitGPU.php

Here are my results with Ethminer:

AMD GPUs ProgPow Hashrate Benchmarks

OS: Windows 10 PRO

Drivers: Blockchain compute

Miner: https://github.com/minerideseama/Ethereum-ProgPow-test/releases (here you can make a donation for devs 0xa7e593bde6b5900262cf94e4d75fb040f7ff4727 and here for the guys who made this possible for windows 0xd9331260fb214F9Dd00c90873FF6B0c5ad2A60dE)

Block: 7280000

Compiled version: ethminer 0.18.0-alpha.3+commit.260c47d2
Source: https://github.com/AndreaLanfranchi/ethminer
Miner compiled on WIN10 64 with CUDA 10.

If you already done your own test, please comment below your results and also submit them here: https://progpow.pro/submitGPU.php

Voting For or Against ProgPow

Hopefully it was clear enough that I am not promoting any side here and I invite you to use your own discretion. Once you pick your side, I invite you to actively support the idea you are standing for.

If you are a supporter, the easiest way to for you to vote for ProgPoW is to use Ethermine, or any other pool that support it. The pool is adding a special PPYE(S)” tag to its blocks. If you are not a supporter, make sure to share your concerns and arguments on the forums and reddit. As a community, we have to listen to each other, but we also have to share our logic with others – sharing is caring, as they say.

Another way to help the side you pick is by casting your vote on http://progpowcarbonvote.com/ . All you will need to do is to send a zero ether transaction to an address, mentioned in the website.

Last but not least, you can also support the development of ProgPoW financially by sponsoring a dedicated team of developers. Follow this link https://gitcoin.co/grants/54/progpow-full-stack-integration for more info: “This is necessary for the critical research to be completed and implemented in a way that ensures the safety and security of the entire Ethereum ecosystem.”

In Conclusion

There is a lot of noise surrounding ProgPoW, which makes it a bit hard to understand what is really going on. There is a lot of support, as well as a lot of hate and even trolling going on in the forums. I hope this review was somewhat helpful to you to better understand the situation.

Let me know your thoughts in the comments below. I also invite you to share any important information you know – it might be helpful to anyone reading this.

Thank you for reading. As always, your comments, suggestions and questions are welcome.

Subscribe and stay tuned for further updates!

The post What’s Up With ProgPoW? Benchmarks & Addressing the Criticisms appeared first on 1st Mining Rig.

SOURCE: 1st Mining Rig – Read entire story here.