FPGA miner for altcoin

[arm_race]

Hi everyone, I would like to know your opinions on the FPGA miners for altcoins. My partner and I have been working on a FPGA miner project for an altcoin(sorry i can't reveal now, but not a very popular one), after being fully optimized we believe the fpga miner should be able to achieve the same level of hash power as of a mid-range graphic card with only 1/10 of the power consumption. It would essentially look like one of those early day bitcoin FPGA miners, with 2 or more FPGA "chips" on a PCB board (depending on the demand), like the one in this link
http://fpgamining.com/. The other beauty of FPGA miners is that a cheap raspberry pi can be used as the host to drive all FPGA miners connected to it via USB cables, there is no GPU -like rigs needed.
Due to the development, custom board design and manufacture the final price might end up in the $500 range. Do you think someone would buy such a product for mining nowadays? Just want to hear your general opinions on it, as both of us are relatively new to the crytocurrency and didn't experience the fpag miner era of bitcoin.

Any suggestion is appreciated!

[1714605537]

1714605537

[1714605537]

1714605537

[1714605537]

1714605537

[1714605537]

1714605537

[1714605537]

1714605537

[1714605537]

1714605537

[ylpkm2]

So 2.50 a day profit or so, almost no power draw for $500 bucks.. at a ROI of more than 6 months, I would pass. At least I can reuse and sell gpus easily. A fpga that can lose its value quick due to competition would be a poor choice. Someone will use a cheaper board and make a faster one. I know a few programmers here that have made fpga miners. They could potentially make a better version pretty easily. Maybe if price was lower I would go for it..

[Bimmber]

Way too expensive for zero resale value later on.

[stash2coin]

Depending on that what coins can be mined with your fpga, but i think there will be buyers for this, if this 1/10 of power is true it will be tempting for countries with high electricity prices. People were skeptical about bitcoin fpgas but ended buying them at loads until the ASICs came.
Will be interesting to post more info about this development!

What about selling it at lower price like 350 and getting a percentage from mining like the miners devs do? Of course there will be a problem with people trying to avoid your devfee

[Ruhtilg]

Way too expensive for zero resale value later on.

You can reprogram it to mine other coins. But it depends on the developer.

[Bimmber]

With new nvidia cards coming out, this FPGA will be making less and less money. And at some point it won`t be worthwhile any more. At least gpu`s can be sold to gamers.

[arm_race]

So 2.50 a day profit or so, almost no power draw for $500 bucks.. at a ROI of more than 6 months, I would pass. At least I can reuse and sell gpus easily. A fpga that can lose its value quick due to competition would be a poor choice. Someone will use a cheaper board and make a faster one. I know a few programmers here that have made fpga miners. They could potentially make a better version pretty easily. Maybe if price was lower I would go for it..

thank you for your valuable inputs. I am a GPU miner myself and it would also be my no.1 concern of no resale value when buying those dedicated fpga/asic miners.
so what would make you buy a fpga miner? you mentioned the price, currently there are two designs in my mind but i don't know which one customers would prefer:

1. a "big" fpga board with 2 or 4 cores on it, the hash power can beat a graphic card, with only a quarter of its power consumption. priced around $400-$800 (depending on the number of cores), most efficient hash/power and better hash/cost perhaps.
2. a "tiny" fpga mining usb stick. It would look a usb flash drive with heatsink on it (as showed in the link below). it has only one fpga core on it, maybe only a quarter of the GPU miner hash power, but power consumption is tiny as well, as it is powered solely from the USB. priced around $100 (i haven't run through all the numbers yet, just rough one). potentially you can connect multiple of those sticks on a usb hub with good power source. you can also take it with you and plug in your laptop and mine anywhere you want! this would be less efficient in hash/power and hash/cost if you compare to the big one though.
https://www.amazon.com/GekkoScience-2-Pac-Compac-Bitcoin-BM1384x2/dp/B06XBWK2W5/

I agree with you there are so many talented fpga/asic programmers here, but the design, manufacture and validation cycles may cost time and money. during which time we can constantly improve our design to make it more competitive.

[arm_race]

Depending on that what coins can be mined with your fpga, but i think there will be buyers for this, if this 1/10 of power is true it will be tempting for countries with high electricity prices. People were skeptical about bitcoin fpgas but ended buying them at loads until the ASICs came.
Will be interesting to post more info about this development!

What about selling it at lower price like 350 and getting a percentage from mining like the miners devs do? Of course there will be a problem with people trying to avoid your devfee

HI thank you for your interest, I will definitely keep you updated when rubber hits the road! right now we are still working on prototyping our miner design on development boards from different fpga vendors. the fpga devices nowadays are made very power efficient, if a typical gpu takes 150W when mining, i think it is very likely to achieve the same level of hashrate with only 15W power intake.
the price is just a rough number as we haven't run through all the numbers with our suppliers.. technically it would be little tricky to implement a dev fee, and morally i am against it  (no offense) I wouldn't sleep well to take a percentage of miners' hard mined coins...

Yes please take a look at the big and tiny designs i proposed in another reply i would like to know what miners want the most, time has changed since bitcoin fpga era, we should consider changing the design if the is demand from miner community. thanks.

[arm_race]

With new nvidia cards coming out, this FPGA will be making less and less money. And at some point it won`t be worthwhile any more. At least gpu`s can be sold to gamers.

you are right, but technically, from the hash/power efficiency perspective, the gpu can never beat fpga, just like fpga can never be faster than ASIC (which also requires millions $ investment). but like you said gpu is generic and has resale potential. but fpga is versatile, we can custom build the product the way miners want. what would make you to buy a fpga miner? does any of the designs i proposed appeal to you? thanks.

[Bimmber]

For me personally price is key factor. I would not pay whopping 500$ for it.

[arm_race]

For me personally price is key factor. I would not pay whopping 500$ for it.

good point, nobody wants to invest in that much $$$ for a miner which has low resale value. the reason i said $500 is that price range i can manage to compete with the GPU miner at 1/10 of its power consumption. but maybe that's not what miners want? you would prefer $100 fpga miner with only 1/4 of GPU hash rate?

Thanks.

[Rochester]

Hi man, i am trying to msg you but you set no msg from new_bie, which i am here.  

do you mind send me a private message? i am here with an oppty for you.

email also works. mine can be find in my profile.
thank you!
Rochester

[DigitalCruncher]

Consider making of non-standard (incompatible to NOMP etc) pool for these miners to not disclose the algorithm's name. Just disclosing the name of algo will ruin your business immediately. FPGA developers will build their miners for that altcoin, and altcoin developers may wish to change POW algorithm.  Please think about it.

Also your pool solves the problem of devfees.

[whoismoses]

"Mid-range"?

If I assume that is an RX 570... That card can get 600-800 H/s on Cryptonight and 28-30 MH/s on ETH...

On both of those with electricity you would make $0.80 to $1.50 per day.

That is almost a 1 year break even...

with the reduced power it is almost the same as buying a mid-range cpu. I'd probably buy the GPU... Now $250, different story...

Cheaper, just as fast or faster, and less power sells.

[Rochester]

For me personally price is key factor. I would not pay whopping 500$ for it.

good point, nobody wants to invest in that much $$$ for a miner which has low resale value. the reason i said $500 is that price range i can manage to compete with the GPU miner at 1/10 of its power consumption. but maybe that's not what miners want? you would prefer $100 fpga miner with only 1/4 of GPU hash rate?

Thanks.

it is not about the resale value from my perspective. all about the efficiency and profitbility of mining and flexibility of switch on altcoin to another, no?

[Miz]

As far as I understand, the payback of such equipment is at the level of 6+ months. Now it seems to me, given the unstable market it is very risky.

[arm_race]

Consider making of non-standard (incompatible to NOMP etc) pool for these miners to not disclose the algorithm's name. Just disclosing the name of algo will ruin your business immediately. FPGA developers will build their miners for that altcoin, and altcoin developers may wish to change POW algorithm.  Please think about it.

Also your pool solves the problem of devfees.

it is a valuable advice, thanks! just see what happened to monero when antminer x3 is introduced to market...

[arm_race]

"Mid-range"?

If I assume that is an RX 570... That card can get 600-800 H/s on Cryptonight and 28-30 MH/s on ETH...

On both of those with electricity you would make $0.80 to $1.50 per day.

That is almost a 1 year break even...

with the reduced power it is almost the same as buying a mid-range cpu. I'd probably buy the GPU... Now $250, different story...

Cheaper, just as fast or faster, and less power sells.

I hear you . the price of those dedicated miners are outrageously high nowadays. we will be working hard to control the price to below $200, and performance as good as RX 570 type of GPU.

[arm_race]

As far as I understand, the payback of such equipment is at the level of 6+ months. Now it seems to me, given the unstable market it is very risky.

what is the reasonably ideal payback period in your mind?

[arm_race]

Hi man, i am trying to msg you but you set no msg from new_bie, which i am here.  

do you mind send me a private message? i am here with an oppty for you.

email also works. mine can be find in my profile.
thank you!
Rochester

sorry man i haven't logged in for a while and missed your post, would you mind sending me a pm with your email address? because i don't see that from your profile, probably because i am also a newbie?
thanks.