Bitcoin Forum

Hi 4all!
Let me introduce my project of some additional mining equipment  :)

What's it?
This is a mining switch (https://photos.google.com/share/AF1QipOZd3-lsD1M6voOjkEnlbIQ92Al52vfIwa7u5VZI0KdMpICMsXojJZqcC1nY4eOsA?key=bkZyN2NWMThuVzR1Q1JoRHJfZTJOb3FQOUx2SGRn) for ETH (BTC currently being tested).
If you find a block, it goes to the solo pool. Solo pools operate the same way as usual pools, with the only difference being that block reward is not distributed among all miners. The entire reward in a solo pool goes to the miner who finds the block.
For example, this guy (https://flexpool.io/0xfb260b9Fc0b40F982b12F97Fd6A3047F6d4B5cD8) found a Ethereum block 12211885 using about two video cards (86.7 MH/s). The entire reward was 30.22654 ETH. But he got only 0.0005 ETH of it...

What it's for?
During mining you continuously search for blocks (or "golden nonce (https://www.blockchain-council.org/blockchain/what-is-a-golden-nonce-and-what-is-its-usage-in-blockchain/)"). To reduce payout variance, miners group into pools and share their income proportionally to their hashpower.
To determine each miner's hashpower, pools use a share system: when you are mining and find a block header of high difficulty (https://learnmeabitcoin.com/beginners/difficulty) but not high enough to be above the current target, you send it to the pool. These difficult-but-not-enough headers are called "shares".
Since finding them is way more frequent than blocks (miners often send several of them per second), no matter how small your hashpower is, it will still be taken in account at payout time even if you're unlucky. The more shares you send, the more hashpower you have, the bigger the share of the pool's income you will get. They are not broadcast to the bitcoin network, since they would be immediately rejected there and are of no value to anyone but the miner that found them and the pool software. But sometimes you find a block!

How does it work?
The ShareSwitch (https://photos.google.com/share/AF1QipOZd3-lsD1M6voOjkEnlbIQ92Al52vfIwa7u5VZI0KdMpICMsXojJZqcC1nY4eOsA?key=bkZyN2NWMThuVzR1Q1JoRHJfZTJOb3FQOUx2SGRn) continuously analyzes the traffic between the miner and the pool. To do this, the device creates a clone of the miner and both pools inside itself. This must be done because network traffic sniffing/decryption in real time is too difficult and we have to create endpoints.
As a result, the miner thinks that it is connected directly to the pool, the pool(s) thinks that it is connected directly to the miner. In the middle, the device's core routes the data. If valid hash* is found, it will be forwarded to the solo pool. You can also set the percentage of shares sent to the solo pool and the usual pool. This will avoid being banned for inaction.

   *most probably valid hash based on the global difficulty level.

What advantage?
Something similar has long been done with a proxy. This device has next advantages:
1. Very small delay. Typically about 1 ms at full load (4 miners connected).
2. High reliability. This works nearby and is controlled by the owner, unlike a proxy somewhere far away.
3. This cannot be identified from the pool side. Nothing changes for the pool.
4. Low power consumption (no more than 5W).
5. Advanced logic based on FPGA and dedicated controller Atheros AR9331.
6. Stylish green heatsink inside.

I posted some photos and diagrams in my GooglePhotos (https://photos.google.com/share/AF1QipOZd3-lsD1M6voOjkEnlbIQ92Al52vfIwa7u5VZI0KdMpICMsXojJZqcC1nY4eOsA?key=bkZyN2NWMThuVzR1Q1JoRHJfZTJOb3FQOUx2SGRn).

Updates will be soon  :)

UPD. April 26, 2021. WEB control board was added. Optimization of power consumption.
UPD. May 15, 2021. Automatic pool test added. This checks for key vulnerabilities and predicts success.
UPD. May 20, 2021. Website launched: https://shareswitch.online
UPD. June 7, 2021. The production of enclosures has been prepared. Cutting and marking files have been prepared.
UPD. June 28, 2021. The prototype has finished testing. Negotiations are underway to make a batch.
UPD. Aug 02, 2021. We have finally purchased all the components and the devices have gone into production! The order is available on our website (https://shareswitch.online).

I am glad to inform that we have finally purchased all the components and the devices have gone into production!
As promised, 5000 pieces were ordered. This allowed us to significantly reduce the cost of each unit!
The order is available on our website https://shareswitch.online
Youtube: https://youtu.be/SB6VW-VImzQ
Questions: sales@shareswitch.online

UPD. Aug 20, 2021. The whole batch was shipped!
UPD. Sep 15, 2021. First update is available.
UPD. Feb 14, 2022. Bridge feature added.

Hmmm...  Why the link to "app . goo . gl" being banned?..  ???  https://photos.app.[Suspicious link removed]/BTfusP9QydBLaZSS8

Because it's a scam and likely malware infected.

What you describe is impossible, each stratum connection only accepts hash produced
from the unique data it sent. It's invalid for any other stratum connection.

In fact what you are describing is a man in the middle attack.

https://en.wikipedia.org/wiki/Man-in-the-middle_attack

If the etherbase (for pplns and solo pools) is the same, then everything else - demon, task - is the same too. This is easy to check in explorer.
Next, you must specify which version (and implementation) of the EthereumStratum do you mean. Ethereum is a patchwork quilt and your declaration looks too bold. EthereumStratum != BTCStratum.

For Bitcoin, yes, this is a coinbase field problem what including extranonce. In the EthereumStratum v1 (v2 is not actually used) there are many more possibilities to manipulate this parameter. Many pools do not provide extranonce at all, using Id & job ID only.
Rest data can be easily swapped on the fly - this device contains a high-speed FPGA specialized in telecommunications.
During pplns retransmission, device can lazily palpate the both pools and choose the best hole (requests to change various parameters in the simplest case). This is not a stupid relay, it has advanced logic which is noticeable from the element base. For typical MITM it is overkill.
If it is impossible to work with this pool(s), it will signal. But on the verge of POS, no one is in a hurry to close vulnerabilities...

I am testing known BTCStratum vulnerabilities for now and I am not claiming btc at this stage.

P.S. Why should I infect my own device? :)

Indeed this is impossible. We have software that detects if something like this is running and we ban the miner.

I'm a little lost. A week ago this was impossible due to a stratum connection. Now you talking about software to block the impossible.
Why block the impossible?  ???

WoolyPooly was attacked in a similar way recently.
I believe half of the young pools are subject to similar attacks. This is difficult to detect such at low hashrate. I first learned about this more than two years ago and continue to notice. This is the eternal struggle between good and evil  :(

It's a fight of skills. Yes, many vulnerabilities have been known over the years. Some were received recently. I have not invented anything new, I have only combined it all into a userfrendly device.
Because it is based an FPGA it can be updated via the internet for new vulnerabilities & algo.

If I have mining power to find block in reasonable timeframe, I don't need pool.
If I need pool, I don't need FPGASwitch because I don't have mining power to find block in reasonable timeframe.

What you say means subsidizing the weak miners at the expense of the strong. But the pool is not created for charity.
In reality everyone has a chance to find a block. Even using one processor in Solo. You can do two iterations and find a golden nonce, or you can do two million iterations and find nothing. Everything depends on the starting nonce.

You choose a low salary and no bonus. Well, this is your choice.

But you are cheating.  If it works as you describe you are on pool when you only submit shares but not a found block.  When you do hit a block you do not share with pool.

So it clearly only benefits you and harms the pool.  Either solo mine or pool mine.  What you propose is to cheat the system.  That is not an honest approach.

The truth is, it has been in use for years  :) Recently, a simple proxy server could route eth shares (on some pools this is still possible). In getwork days, it was used everywhere. I just made a simple one button device.
It's like VPN or tor. Complex technology, but there are solutions with one button.

Hey! Some news.
1. We will try to mount all the videos in a small movie soon.
2. Now we are negotiating the production of a batch. We agreed on two options - in a plastic case and naked (in transparent heat-shrink tube). Naked will be cheaper.
The price will depend on the quantity too. We have several hundred pre-orders now and are targeting one thousand devices. This is not the biggest batch for China... If we order five thousand, the price will drop by half. We hope that interest in the project will rise...
3. The end of testing is scheduled for mid-June.

So... The silicone crisis has affected all areas of electronics manufacturing. We cannot buy a couple of chips for our batch. We are negotiating the purchase from several warehouses, while I am not ready to predict the delivery time  :(

Your idea is intriguing (I believe it could work, just not sure its ethical, not debating, just giving my opinion) and it seems you possess the know how of what I was asking for a week or so ago.

https://bitcointalk.org/index.php?topic=5345317.0 (https://bitcointalk.org/index.php?topic=5345317.0)

Maybe you have some ideas while you are waiting chips for your project/equipment.

My device can do this, but it's much easier to use the https://miningrigrentals.com/ IMHO
It is not necessary to rent on. Although sometimes it gives x2 to the market.
But this has a user-friendly interface for switching between coins and pools and any more controls the traffic.

Yes, I have used them in the past and could still...I would just like to be in control of it all and not worry about anyone "borrowing" hash rate for free.

I'm using the MRR for several years and they never stole hashrate...
You can of course use my device. But it will not be cost-effective. This uses a expensive FPGA to work.
Any open source router is your choice IMHO. OpenWrt etc.

Do you have a vision when it will go on sale?..

I hope no more than two weeks.

id like to preorder

This is a scam.

People tried doing this back in 2011 or so to cheat pools and this is why most pools send a special hash which if a block is actually found will credit the pool AND NOT the individual miner.

You get a hash sent from the pool where you generate your nonces and if you find one with a difficulty high enough, which results in a block, you CANNOT send that hash yourself to the network because it contains the info of the pool which will get credit for finding the block.

This is simply not possible and it seems people are already about to get scammed.

EDIT, what a poster above said is true. This is called a Man in the middle attack. Before anyone get scammed just do some googling first and you will see what he is describing is just not possible.

1. Do you understand that bitcoin's stratum2 is based on similar principle (https://assets.website-files.com/5e5fcd39a7ed2643c8f70a6a/5f8337d0eb8a19587ffcb1aa_5ef35164b69ea7f1d27c6a2a_BRAIINS%2520OS%2520DEVELOPMENT%2520UPDATE.jpeg)?  ::) 
2. https://bitcointalk.org/index.php?topic=5331736.msg56872924#msg56872924
3. In getwork protocol, in 2011, in BTC, this worked great! And in chia for the past two months  ;D For this reason, the BTC stratum1 was launched. But I will repeat again: BTC stratum != eth stratum. Stratum1 != Stratum2 etc. If you undertook to criticize the project, specify the coin, the stratum version, the pool version, etc. Otherwise, it is philosophy.

I am not too technical with the details, but from what I understand it works like this. Pools implemented a type of safety to prevent this man in the middle attack from happening. So when a miner is sent a hash to "hash" and generate a nonce which is less than the difficulty, one of the parameters includes the pools info.

So miners end up hashing this code and if they find a block, the hash originally contains the pool info in the paramter. You cannot change this parameter after you find the nonce which solves the block. If you change the pool info, your entire nonce won't work since the hash will be different.

I don't know if you already know this and are trying to scam people or if you are innocent and you just don't know that its impossible. Either way, what you are proposing here is just not possible. It won't work. If you don't believe me choose some low hashrate ETH clone type coin OR fork your own and see if you can actually succeed in doing this.

Mine on your local pool and if you find a block, try and send that block to the network with your solo client, and see what happens. Most likely it will get rejected by the network or if it gets accepted the pool will get the block credit and not the one which sent it.

Yes it's a scam. Even if there is a bug in stratum used for eth it would be so obvious and easy to
exploit it would be common knowledge. The only reason to use an FPGA is to steal the
block from the miner who tried to sleal it in a true MIM attack, ie attack both ends.

That's *if* such a bug exists, and if it did it would have destroyed Eth mining by now.

scam.  taking the time to write this incase anyone has hope that it isn't, so I can save you some money.

here's the tell...  It's all marketing blah blah blah, with one slight nod to technical.  He copied some ideas/text from how real FPGAs are done.  But look how it is artfully danced around:

To do this, the device creates a clone of the miner and both pools inside itself. This must be done because network traffic sniffing/decryption in real time is too difficult and we have to create endpoints.

So, I have some questions:

• "clone of the miner and both pools".  What?  Why?  You're saying you are creating a hardware protocol in programmable logic?  OK, this is because you need microsecond response times?  Why?
• "real time is too difficult".  No it's not.  The task is done in real time.  A malicious task could also be done in real time.  Also, what exactly do you mean by "real time", because I don't think that means what you think that means.  Software-driven-implementation delays on the order of miliseconds don't matter for mining.  Your task is single purpose; the reason smart firewall traffic sniffing & decryption is done with FPGA assist is to greatly expand the capabilities of the device.  This is not necessary or useful in your case.  You are not selling a general purpose device that must have broad and expandable capabilities to compete in a marketplace of similar devices.  You are selling a toaster, it does one thing only, and it's unique with zero competition.  It's far easier to implement everything you describe in software, but if you use only software you can't make the argument for asking for pre-orders because you have to fund physical device development (which is of course very expensive, made more expensive due to component shortages), can you?
• "we have to create endpoints".  Right, I get it, you have to end the sentence with something that sounds like it matches.  Try harder next time, this makes no sense.  Your premise is not very complex.  By "endpoint" you mean "miner".  But "endpoint" sounds more technical, because that's hardware lingo (PCIe CEM spec among others).  All you're claiming to do is mine on two pools at once.  It's not rocket science.  It's not even particularly hard.  You open two sockets.
• why do you need microsecond speed on a task that can see  milliseconds of delay without issue?  if you say "a millisecond is enough to lose a block", I respond:  that's ok if a rare event causes you to lose the block, because you're already implementing a ratio mechanism to prevent it from being obvious that you're cheating.  So this imprecision becomes a slight variance to your ratio.  

Bottom line:  usec speed isn't a requirement to do the task.  The use of an FPGA provides no value and is a complicated, difficult solution to something that can be easily done in software.





Can we just get this loser banned already, or does he have to successfully scam people first? 

This has been known for a very long time. But you must have the Russian hacker level for this  ;D
https://miningclub.info/threads/woolypooly-ili-kak-kinut-pplns-majnerov-iz-za-sobstvennoj-zhadnosti.70514/

adaseb
I am not too technical with the details
But that didn't stop you from calling the project a scam ;)

So miners end up hashing this code and if they find a block, the hash originally contains the pool info in the paramter.
ExtraNonce, yes, this is a problem in bitcoin's stratum1. But in ETH, the situation is somewhat different.
EthereumStratum1 (https://www.programmersought.com/article/4562796333/) does not have a mandatory ExtraNonce (except the mining.set_extranonce notice that can be rejected), but it have etherbase and task ID.
The etherbase hashed, but task ID and message id - not. Obviously, the solo pool will not accept a task with an incorrect ID and we have to repack the data asap.
It is expensive to have different nodes (for PPLNS and SOLO) and sometimes pools are tricky here. The daemon is the same, but the task changes a little. From the outside, it looks like different nodes, but a little traffic analysis allows you to determine this trick and use this. The task changes every 3 seconds (or less) and so I use Cyclone IV FPGA :)
This is the tip of iceberg known to everyone. There are also lesser-known vulnerabilities that I exploit (https://youtu.be/1dF2K_D1cis?t=39). In any case, the ShareSwitch (https://shareswitch.online) scans the pool first and tries to configure the most profitable exchange. And only then analyzes and changes the traffic. If the pool is well protected, this will inform the user.
Each pool's admin is lazy in his own way. And each pool has its own vulnerabilities. EthereumStratum2 has a range of ExtraNonces for each miner, for example. This opens up new possibilities and closes some outdated ones. Some of these vulnerabilities could have been closed years ago, but the developers are too lazy... Most likely, it will never be closed again, because there is a POS ahead.

If you don't believe me choose some low hashrate ETH clone type coin OR fork your own and see if you can actually succeed in doing this.
Wrong way. It does not depend on the hashrate. However, ETC has even more holes, eah...

Mine on your local pool and if you find a block, try and send that block to the network with your solo client, and see what happens.
I've explained to you how it works. It's a little more complicated than you imagine :)
But for example in Bitcoin Stratum V2 such a scenario is possible after some manipulations. I have a video where ASIC works (https://youtu.be/SB6VW-VImzQ?t=168) in this mode.

JayDDee
used for eth it would be so obvious and easy to exploit it would be common knowledge
It is not easy to use and it has been known for a long time. It's just that every teenager could do it before, now you need an FPGA for this.

The only reason to use an FPGA is to steal the block from the miner who tried to sleal it in a true MIM attack, ie attack both ends.
So, can I steal the block or not? ;)
This device is placed between the miner and the router. There are thousands ways to uncover unnecessary connections and traffic. And of course it can be blocked at the router level. What is the point of such a scenario?

That's *if* such a bug exists, and if it did it would have destroyed Eth mining by now.
I'll just remind you that CryptoNight ASICs (Baikal Giant-N etc.) existed long before the official release and everyone claimed that this impossible.
The first FPGA miners for bitcoin did not appear in the RadioShack.
This device has been around for a long time too. This only manifests itself if you have found a block. If not, no one will notice the changes. Don't worry, POS will kill mining much earlier ;)

malefactor
You don't like FPGA, I see :)

He copied some ideas/text from how real FPGAs are done.
The real FPGAs are done on the factory. Your question about bitstream maybe?

OK, this is because you need microsecond response times?
Because it contains a lot of operations that the AR9331 will not be able to handle. AEAD, for example.

Software-driven-implementation delays on the order of miliseconds don't matter for mining.
Yep. But without FPGA, it will take seconds. The task changes every 3 seconds or less in ETH, do you know?

You are selling a toaster, it does one thing only, and it's unique with zero competition.
The iPhone is also very simple. Plastic and glass only. If you analyze it without any knowledge, of course ;)

It's far easier to implement everything you describe in software
Again an easy implementation of an impossible technology?.. Please decide: either it's impossible, or it's easy to implement?

you can't make the argument for asking for pre-orders because you have to fund physical device development (which is of course very expensive, made more expensive due to component shortages), can you?
The development is completed in April, check the publication dates. It could have been ready in May, but I couldn't buy the components until July.
No pre-order has been paid. No one has paid before. Keep your fantasies to yourself.

By "endpoint" you mean "miner".  But "endpoint" sounds more technical, because that's hardware lingo
By "endpoint" I mean endpoint (https://valelab4.ucsf.edu/svn/3rdpartypublic/boost/doc/html/boost_asio/reference/ip__tcp/endpoint.html). Learn the terminology (https://en.wikipedia.org/wiki/Network_socket) first.

PCIe CEM spec among others
USB have endpoints (https://microchipdeveloper.com/usb:endpoints) and many other interfaces (https://deploy-docs.aptible.com/docs/endpoints#endpoint-types). It is a good practice to use Google before posting. No need to thank me, sir  ::)

All you're claiming to do is mine on two pools at once.  It's not rocket science.  It's not even particularly hard.
For all the mining platforms/OS in one box? You can implement it at any time and compete with me!

why do you need microsecond speed on a task that can see  milliseconds of delay without issue?
Because I know a little more than you do.

Bottom line:  usec speed isn't a requirement to do the task.  The use of an FPGA provides no value and is a complicated, difficult solution to something that can be easily done in software.
I will look at your developments with undisguised pleasure. Especially in FPGAs.

sergioheadache
id like to preorder
The devices are in production now. The sale is open on our website (https://shareswitch.online), welcome.

Do I have to look into the web interface for control every time? If something changes on the pool during operation?
You said earlier it will be possible to adjust how many non gold shares to send to solo pool, but I don't see this in the interface.

Front panel LEDs indicate that operation cannot be continued. During this time, the device will operate as a straight bridge. Perhaps in the future I will add an e-mail notice.

This function has been removed, for most pools the required number of shares can be understood from the node version and other public parameters.

Will the EIP1559 be an obstacle to work?

Of course not. This will work until the POS.

So, the batch was made ahead of time and the first 50 devices left today! All buyers will receive automatic track notifications as they are sent.
The documentation will be available in your personal account, look for the link in the email.
It was difficult for us, but thank for your support to all buyers!  :D

The whole batch was shipped!

Dear customers! Please check spam folder for update link!
Thanks.

An important update for Chinese miners and other cypherpunks.

Now you can activate the bridge between any other device (for Chinese miners, devices outside of China are available for output only). 它有双重功能--它可以保护你的连接也没有人会发现你在挖任何加密货币可以让你获得全部奖励，而不是一部分.

These bridges mix your traffic randomly and encrypt with an additional key + salt between the bridges. This obfuscate your pool connections and makes traffic analysis difficult. At the local level this makes it impossible to understand what you are mining.
Unlike public VPN services, it does not use shared widely known servers and does not keep logs (because the device does not have free memory for this, ha-ha). This does not mix your mining traffic with other traffic that can identify you. The mining traffic usualy is very small and it is lost against the background of other legal data.

The work logic resembles a torrent: you use everyone and everyone uses you.

Please note that the feature will be enabled by default in the March update. 获取更多信息请联系我们。