DIY FPGA Mining rig for any algorithm with fast ROI

[s1gs3gv]

There's only so many people in this business. Sales and more specifically, the authorized sales reps, like to brag.

How do you think bitmain knows what their competitors are doing? How do you think I know what amazon paid Xilinx, or about recent purchases Bitmain made?

I give up. Tell me

I just did. The first thing any sales rep is going to do is tell you how wonderful of a company they are and how many huge businesses in your sector they're doing business with. All you need to do is ask some leading questions the right way. They'll happily provide you information that you can use a little bit of supposition with or in some instances, just directly tell you information they shouldn't.

So you have some secret info that Intel is going to flood the market with cheap UltraScale+ class products ?

If that is the case, why should we order your $4000 boards ?

[senseless]

There's only so many people in this business. Sales and more specifically, the authorized sales reps, like to brag.

How do you think bitmain knows what their competitors are doing? How do you think I know what amazon paid Xilinx, or about recent purchases Bitmain made?

I give up. Tell me

I just did. The first thing any sales rep is going to do is tell you how wonderful of a company they are and how many huge businesses in your sector they're doing business with. All you need to do is ask some leading questions the right way. They'll happily provide you information that you can use a little bit of supposition with or in some instances, just directly tell you information they shouldn't.

So you have some secret info that Intel is going to flood the market with cheap UltraScale+ class products ?

Secret info? No, All you need to do is look at the information they're publishing, what they're developing, etc to determine what their goal is. As I've said before, intel could lose market cap (stock valuation) equivalent to 100% of Xilinx's stock and it would just be a "bad day". Xilinx and Intel are not in the same class. Intel owns their own fabs. Xilinx pays for fab. Intel can run as many tape outs as quickly as they want with no delay. Xilinx has to wait for TSMC scheduling. Intel can produce chips and sell them at a profit for below what Xilinx's cost to produce is. If they did get into a pricing war, Xilinx would be destroyed. Not only will they lose on cost, but they're also going to lose on performance because.... Well, vivado isn't great. Even ISE was better than vivado.

The Stratix 10 will also have HBM2 and they also have a dual port memory technology equivalent of the ultraram.

[GPUHoarder]

There's only so many people in this business. Sales and more specifically, the authorized sales reps, like to brag.

How do you think bitmain knows what their competitors are doing? How do you think I know what amazon paid Xilinx, or about recent purchases Bitmain made?

I give up. Tell me

I know you’re poking to get reactions, but to be serious without sharing things I cannot - some of us have been in business discussions about this market and plans in it with many large players for quite some time. Anyone who thinks every company named is not reading this thread is quite mistaken.

Regarding companies private plans, none of which I have presented or revealed here, nor will I. You would do well enough just actually researching what is public in various corners of the internet, developer forums, answers to questuons, publically announces deals and products, etc.

[senseless]

I know you’re poking to get reactions, but to be serious without sharing things I cannot - some of us have been in business discussions about this market and plans in it with many large players for quite some time. Anyone who thinks every company named is not reading this thread is quite mistaken.

Regarding companies private plans, none of which I have presented or revealed here, nor will I. You would do well enough just actually researching what is public in various corners of the internet, developer forums, answers to questuons, publically announces deals and products, etc.

Ya, Once I realized that my VCU1525 thread was #2 on google for vcu1525... I figured they'd be aware of it... I'm hoping that whoever is blocking the sales at Xilinx will wisen up and open the flood gates.. If they don't want my money, I'm more than happy to give it to intel. $50,000 for a proto chip 

[s1gs3gv]

I know you’re poking to get reactions

Not really. I'm just amused by some of the stories i'm hearing. For example, why would anyone want to pay $4000 for an FPGA board containing an export controlled FPGA that might, allegedly, implement some algo written by somebody which might ROI sometime, in the perfect world, when Intel is going to flood the market with cheap ultrascale+ class FPGAs real soon now ?

Hope and greed spring eternal.

Let me guess - to save mining from centralization ! lol

[senseless]

I know you’re poking to get reactions

Not really. I'm just amused by some of the stories i'm hearing. For example, why would anyone want to pay $4000 for an FPGA board containing an export controlled FPGA that might, allegedly, implement some algo written by somebody which might ROI sometime, in the perfect world, when Intel is going to flood the market with cheap ultrascale+ class FPGAs real soon now ?

Hope and greed spring eternal.

The export control is no worse than the export control for a lot of other multi-purpose hardware. There's an ITAR code for your CPU, I'm sure even GPUs have one. A license per country you wish to ship to (Canada excluded). It's really not a big deal at all, and, doing production outside the USA with a non-usa corporation you can ship to wherever you want (according to whatever laws are applicable to the entity in the territory that it's registered, operating, and shipping from).

As far as I've seen, the only person who's really publicly announced or mentioned performance has been whitefire990. I was more than happy keeping the whole operation a secret and continue mining in secret. If you don't want to buy one. DONT.

[LTCMAXMYR]

Don't waste time. Hurry up and make Bytom miner. The mining profit is 20BTC everyday.It's a big business,better than selling boards here.

[vrdelta]

What is your proposal for traction?  Between what those of us with our own RTL have posted here, what I know of in private, and what is already in public repos I’m unaware of any algorithms for which there is already code written to run on FPGAs. Actions are being taken to make sure the community has access to various hardware options, at better pricing that’s has previously existed.

Consolidate and release all the code in a repository so we have a starting point.  I don't see much point in keeping the algorithm codes a secret. The masses aren't going to be making bitstreams for their FPGAs anytime soon and you still need a miner for it to work.

I am making a couple of bitsreams targeting the xcvu9p so when my 1525's arrive on Tuesday I can see what the numbers look like.

[s1gs3gv]

Don't waste time. Hurry up and make Bytom miner. The mining profit is 20BTC everyday.It's a big business,better than selling boards here.

It won't be 20BTC/day if 10000 people are mining it with the same FPGA you are.

[LTCMAXMYR]

Don't waste time. Hurry up and make Bytom miner. The mining profit is 20BTC everyday.It's a big business,better than selling boards here.

It won't be 20BTC/day if 10000 people are mining it with the same FPGA you are.

shitman's ASIC miner is 1KH /400W
GPU miner 6x 1060, 1KH/300W
bytom global hashrate is about 20MH,that is 20000 miners
let's see you beat the GPU

[senseless]

Don't waste time. Hurry up and make Bytom miner. The mining profit is 20BTC everyday.It's a big business,better than selling boards here.

It won't be 20BTC/day if 10000 people are mining it with the same FPGA you are.

shitman's ASIC miner is 1KH /400W
GPU miner 6x 1060, 1KH/300W
bytom global hashrate is about 20MH,that is 20000 miners
let's see you beat the GPU

It depends what the algo looks like. There are already tensor bitstreams available for lease on amazon from third party companies.

[GPUHoarder]

Don't waste time. Hurry up and make Bytom miner. The mining profit is 20BTC everyday.It's a big business,better than selling boards here.

It won't be 20BTC/day if 10000 people are mining it with the same FPGA you are.

It would be impressive if that’s a VU9P, seeing as how that volume of the card doesn’t even exist.

Edit: Read the Tensority algorithm, as I was excited for the idea of proof of useful work rather than endless hashing to no purpose. Sad to see that is all the ‘marketing’ piece, and the actual algorithm is more meaningless calculations.

[ilia_2s]

senseless, can you send me your e-mail to PM. I cant write to your PM due to small amount of postings.

[gameboy366]

Don't waste time. Hurry up and make Bytom miner. The mining profit is 20BTC everyday.It's a big business,better than selling boards here.

It won't be 20BTC/day if 10000 people are mining it with the same FPGA you are.

shitman's ASIC miner is 1KH /400W
GPU miner 6x 1060, 1KH/300W
bytom global hashrate is about 20MH,that is 20000 miners
let's see you beat the GPU

I am mining it at 700 h/s with few 1050ti. Are you personally mining it too ? I can't good performance out of it. 10*1050ti gives same hash as 5*1050ti. Can't open another instance with more then 5 gpus.

[saint78]

I'm pretty sure intel's plan is to flood the market. Why else would they be developing hybrid cpu/fpgas (like APUs) or why would they bother to create the CCIX interconnect? They're definitely going to flood the market. The question is when. Xilinx will almost surely wait and only react hoping that Intel will also try to preserve high margins. The way I see it, Xilinx has one last chance to gain market share before Intel opens the flood gates. Get your options placed on XLNX while you still can -- Looking forward to making money on their downfall.

Plus, as you and I both know, Quartus is capable of placement / routing that is orders of magnitude better than anything Vivado could do automatically without floorplanning.

I'm looking forward to my call with Intel next week! Hopefully I'll be able to get something rolling quickly and at a lower price point so we (crypto community as a whole) can avoid Xilinx all together.

Edit:

Oh ya, I find it hilarious that devs are changing their algos based on what we say in this thread. Which reminds me, I just developed this new code for cryptonight which makes use of a little hack that's able to bypass some steps to obtain a result quicker. I'm now mining monero at 100Kh/s per fpga.

Can you give anymore details on what exactly you are going to try to get going with Intel? Are you talking Stratix10 HBM? Or some form of prototype?

[senseless]

Can you give anymore details on what exactly you are going to try to get going with Intel? Are you talking Stratix10 HBM? Or some form of prototype?

Yes, the Stratix 10 parts are equivalent to the Ultrascale+ parts. I believe we'll be able to get the same or better performance for a much lower price point. If FPGA are really going to be viable in the long term to displace GPUs the cost has to be reduced in a significant way. $3995 is ok, right now, while margins are high... But if we really wanted to displace GPUs completely both the cost / perf and power / perf will need to be better than GPUs. I'm hoping that we'll eventually (6-12 months) be able to sell a $1000-2000 board with a stratix 10 or ultrascale+ part that has HBM.

[d57heinz]

Can you give anymore details on what exactly you are going to try to get going with Intel? Are you talking Stratix10 HBM? Or some form of prototype?

Yes, the Stratix 10 parts are equivalent to the Ultrascale+ parts. I believe we'll be able to get the same or better performance for a much lower price point. If FPGA are really going to be viable in the long term to displace GPUs the cost has to be reduced in a significant way. $3995 is ok, right now, while margins are high... But if we really wanted to displace GPUs completely both the cost / perf and power / perf will need to be better than GPUs. I'm hoping that we'll eventually (6-12 months) be able to sell a $1000-2000 board with a stratix 10 or ultrascale+ part that has HBM.

Now that’s a price point I can sort of agree with. Of course again the barrier for entry is awfully high.  In order to maintain any sense of decentralization profits have to be low low low and cost of card has to be such that anyone can afford to play. Sadly this system will be taken over by the rich.  Again the rich just continue to get richer.  I had high hopes for crypto. But then bitmain came along and has everyone scrambling.  I’ve been warning of bitmain for years now and everyone basically ignored me.  Minority you know.  Anyway good luck chasing the dragon with these.  It’s a rich mans trick crypto is nowadays. Good luck all

BR

[senseless]

Can you give anymore details on what exactly you are going to try to get going with Intel? Are you talking Stratix10 HBM? Or some form of prototype?

Yes, the Stratix 10 parts are equivalent to the Ultrascale+ parts. I believe we'll be able to get the same or better performance for a much lower price point. If FPGA are really going to be viable in the long term to displace GPUs the cost has to be reduced in a significant way. $3995 is ok, right now, while margins are high... But if we really wanted to displace GPUs completely both the cost / perf and power / perf will need to be better than GPUs. I'm hoping that we'll eventually (6-12 months) be able to sell a $1000-2000 board with a stratix 10 or ultrascale+ part that has HBM.

Now that’s a price point I can sort of agree with. Of course again the barrier for entry is awfully high.  In order to maintain any sense of decentralization profits have to be low low low and cost of card has to be such that anyone can afford to play. Sadly this system will be taken over by the rich.  Again the rich just continue to get richer.  I had high hopes for crypto. But then bitmain came along and has everyone scrambling.  I’ve been warning of bitmain for years now and everyone basically ignored me.  Minority you know.  Anyway good luck chasing the dragon with these.  It’s a rich mans trick crypto is nowadays. Good luck all

BR

Unfortunately that's human nature. The same can be said for any system of proofing or economics (even socialism). There is a new netflix show called "Explained". The third episode on Monogamy had some interesting things to say about tribal nature and raising children. It really got me thinking about society as a whole, why we do the things we do, and how ridiculous it all is. We're all in this together. If the military budgets globally were diverted into science / research... Most of the ails of today would be no more... Unfortunately, empowering people also means losing power for those who currently hold it... It reminds me of that line from cloudatlas... "There is a natural order to this world, and those who try to upend it do not fare well.". (Oh reminds me of this too..)

Even if we dropped the price to $10 per HBM2 Stratix 10 part.. It wouldn't stop people with more money from buying more than the rest. Even if we limited sales it would just create a secondary resale market that those with more money would take advantage of.

I've been warning the same -- I've been very vocal against a lot of these ASIC companies and would refuse to do business with most all of them.

[rezin385]

I’ve yet to have someone explain to me why they think FPGAs are so bad but GPUs are so good.

Thats a red herring. Who is making that argument ?

The developers who have forked explicitly over their algorithms being listed in this thread.

The economies of scale argument also applies to GPU based mining. For example, with modest capital I could easily build a host system supporting up to 128  GPUs  (since no one uses the PCIe for anything other than making them boot and comm that is lower than serial speed), supporting individually resetting and reinitializing them and all the benefits of smaller hosts, but at 5% of system cost instead of as much as 30%, and also provide power savings. The difference is Bitmain doesn’t represent enough demand to independently control the FPGA market, and the companies in that market are not going to sacrifice decades of high margin business for a short term cash play.

There is no down side for TSMC or commodity memory manufacturers to take big money for cheap part orders from Bitmain or any large player.. There is a downside for Xilinx/Intel flooding the market with cheap FPGAs. Similarly you don’t see NVIDIA and AMD letting Bitmain build custom mining GPUs with their chips at cheap prices. Companies with massive R&D into their chip products want to very carefully maintain control over markets to keep the balance between volume and margin exactly where it is most profitable.

I'm pretty sure intel's plan is to flood the market. Why else would they be developing hybrid cpu/fpgas (like APUs) or why would they bother to create the CCIX interconnect? They're definitely going to flood the market. The question is when. Xilinx will almost surely wait and only react hoping that Intel will also try to preserve high margins. The way I see it, Xilinx has one last chance to gain market share before Intel opens the flood gates. Get your options placed on XLNX while you still can -- Looking forward to making money on their downfall.

Plus, as you and I both know, Quartus is capable of placement / routing that is orders of magnitude better than anything Vivado could do automatically without floorplanning.

I'm looking forward to my call with Intel next week! Hopefully I'll be able to get something rolling quickly and at a lower price point so we (crypto community as a whole) can avoid Xilinx all together.

Edit:

Oh ya, I find it hilarious that devs are changing their algos based on what we say in this thread. Which reminds me, I just developed this new code for cryptonight which makes use of a little hack that's able to bypass some steps to obtain a result quicker. I'm now mining monero at 100Kh/s per fpga.

Whoa. 100kh/s for monero is awesome. If its cnv7 then that's a game changer!

[GPUHoarder]

I’ve yet to have someone explain to me why they think FPGAs are so bad but GPUs are so good.

Thats a red herring. Who is making that argument ?

The developers who have forked explicitly over their algorithms being listed in this thread.

The economies of scale argument also applies to GPU based mining. For example, with modest capital I could easily build a host system supporting up to 128  GPUs  (since no one uses the PCIe for anything other than making them boot and comm that is lower than serial speed), supporting individually resetting and reinitializing them and all the benefits of smaller hosts, but at 5% of system cost instead of as much as 30%, and also provide power savings. The difference is Bitmain doesn’t represent enough demand to independently control the FPGA market, and the companies in that market are not going to sacrifice decades of high margin business for a short term cash play.

There is no down side for TSMC or commodity memory manufacturers to take big money for cheap part orders from Bitmain or any large player.. There is a downside for Xilinx/Intel flooding the market with cheap FPGAs. Similarly you don’t see NVIDIA and AMD letting Bitmain build custom mining GPUs with their chips at cheap prices. Companies with massive R&D into their chip products want to very carefully maintain control over markets to keep the balance between volume and margin exactly where it is most profitable.

I'm pretty sure intel's plan is to flood the market. Why else would they be developing hybrid cpu/fpgas (like APUs) or why would they bother to create the CCIX interconnect? They're definitely going to flood the market. The question is when. Xilinx will almost surely wait and only react hoping that Intel will also try to preserve high margins. The way I see it, Xilinx has one last chance to gain market share before Intel opens the flood gates. Get your options placed on XLNX while you still can -- Looking forward to making money on their downfall.

Plus, as you and I both know, Quartus is capable of placement / routing that is orders of magnitude better than anything Vivado could do automatically without floorplanning.

I'm looking forward to my call with Intel next week! Hopefully I'll be able to get something rolling quickly and at a lower price point so we (crypto community as a whole) can avoid Xilinx all together.

Edit:

Oh ya, I find it hilarious that devs are changing their algos based on what we say in this thread. Which reminds me, I just developed this new code for cryptonight which makes use of a little hack that's able to bypass some steps to obtain a result quicker. I'm now mining monero at 100Kh/s per fpga.

Whoa. 100kh/s for monero is awesome. If its cnv7 then that's a game changer!

Whoosh