8th Alt coin thread. Or what to do now that asics are all over the place.

[grendel25]

I'm in Alaska and I have been very interested in adding some solar capability.  I see Philipma posted something like 1500 watt setups for around $1K? 

Can anyone point me to a link or something where I could read more about this kind of setup, scalability, setup scenarios, etc?  I would be very grateful.

Feel free to PM me too. 

Thank you

[1714537879]

1714537879

[1714537879]

1714537879

[1714537879]

1714537879

[1714537879]

1714537879

[1714537879]

1714537879

[martyroz]

I'm in Alaska and I have been very interested in adding some solar capability.  I see Philipma posted something like 1500 watt setups for around $1K? 

Can anyone point me to a link or something where I could read more about this kind of setup, scalability, setup scenarios, etc?  I would be very grateful.

Feel free to PM me too. 

Thank you

I don't use batteries, but in my Country, selling excess generation (to offset night time usage) is a viable option, in fact many times cheaper than any battery scenario.

I made a thread about it and I'm happy to answer any questions. Basically all you need to know is;
* ball-park installation cost per kw
* Your usage charge per Kwh
* Your feed-in / generation tarrif / credit per Kwh

For me, I built a 21.6Kw system at a cost of $14,000 USD which allows perpetual use of 1600W (plus normal household use).
Right now I'm pulling 6000W though so I still have bills.

https://bitcointalk.org/index.php?topic=2486273.0

[philipma1957]

I'm in Alaska and I have been very interested in adding some solar capability.  I see Philipma posted something like 1500 watt setups for around $1K?  

Can anyone point me to a link or something where I could read more about this kind of setup, scalability, setup scenarios, etc?  I would be very grateful.

Feel free to PM me too.  

Thank you

we are in development only

the large array which is grid tied and no batteries

https://bitcointalk.org/index.php?topic=1369207.new;boardseen#new


the smaller system which mines in the day only

https://bitcointalk.org/index.php?topic=4415357.0

not started testing

we are off to a bad start with the specialty inverters we purchased.

we may need to use a different supplier.

[ZedZedNova]

I am considering having an ASIC or two again;  but it needs to fall into the realm of the S3;  ~400W power consumption, and very inexpensive ($200 max purchase price) and easily re-sellable.     A smaller miner like me can easily justify it and find paths to ROI with my power and budget constraints.   When I see things like the D3 hit that range, where I can buy, use sell and buy again like I did with S3's a bit ago....  Winning strategy.

Unless something crazy happens, the next ASIC miner I buy will probably be from sidehack. Although with the price of BTC down, he is sitting on a sizable number of 2Pac miners that is slowing him down a bit. I'm hoping that he can get the 16nm usb stick miners done before too long, and then get to the 16nm pod miner he's been wanting to do for a while.

When I look at the prices of new ASIC miners and the value of any of the coins they mine, ROI is usually one or more years away with the current network hashrates and difficulty, and we all know how that works out.

[jmigdlc99]

Unless something crazy happens, the next ASIC miner I buy will probably be from sidehack. Although with the price of BTC down, he is sitting on a sizable number of 2Pac miners that is slowing him down a bit. I'm hoping that he can get the 16nm usb stick miners done before too long, and then get to the 16nm pod miner he's been wanting to do for a while.

When I look at the prices of new ASIC miners and the value of any of the coins they mine, ROI is usually one or more years away with the current network hashrates and difficulty, and we all know how that works out.

Are you serious about buying USB based stick miners at this point in time? Haven't those been long obsolete even with free electricity? As far as i know they earn a few cents per month which really isn't worth it so you will have to be playing solo-mining lottery?

I'm interested to learn more about your "plan" for USB stick miners.

[KaydenC]

Moderately hopeful rendertoken or other forms of gpu marketplace gains traction this year.

Where are rendertoken currently at? Beta-testing still? Not currently giving out live rendering jobs yet are they?

...

EDIT: not considering ASICs at all. Not out of snobism, but because they really make no sense currently unless you're in need of a big door-stopper.

Their roadmap: https://rendertoken.com/#roadmap
With any luck we will be rendering by end of the year.

I'm not a huge fan of ASICs too, but I will still grab some ASICs in promising algorithms to diversify, mainly those with good revenue to electricity cost ratio.

[adaseb]

Unless something crazy happens, the next ASIC miner I buy will probably be from sidehack. Although with the price of BTC down, he is sitting on a sizable number of 2Pac miners that is slowing him down a bit. I'm hoping that he can get the 16nm usb stick miners done before too long, and then get to the 16nm pod miner he's been wanting to do for a while.

When I look at the prices of new ASIC miners and the value of any of the coins they mine, ROI is usually one or more years away with the current network hashrates and difficulty, and we all know how that works out.

Are you serious about buying USB based stick miners at this point in time? Haven't those been long obsolete even with free electricity? As far as i know they earn a few cents per month which really isn't worth it so you will have to be playing solo-mining lottery?

I'm interested to learn more about your "plan" for USB stick miners.

You are correct, he is most likely going to buy them for fun and not for profit.

They were even obsolete years ago and maybe for a certain period during a crazy bitcoin price boom did they produce a decent profit and people had like 12 of them plugged into a USB hub while filling up their entire bedroom with those USB miners.

I remember seeing a photo on here with a guy who had like 10 hubs and about 10-12 USB miners plugged into each hub.

They are a novelty item and never for actual profit.

[fr4nkthetank]

I still have a usb block erupter, 335mhs.  Its on solo ckpool.  I bought those from China in mid 2013 i think, I had some usb hubs too  What people dont realize is that I saved on christmas decorations that year because the green lights were so nice.  I sold them in 2015 (i think), for more than I paid them.  They became a novelty item, yes.

I think you cant really lose by buying ASIC's at a low price point if you have low electric cost.  You can chase profits by buying the first batch, or you can buy the last batches at cheap prices.  I like the latter.  I dont have any ASics or server grade equipment because I run my things with low decibels.  Granted I pay ~4 cents usd/kwh (total kwh / total bill) all in, but I suggest running your nvidia cards on lyra2z until fpga's destroy the algo.  Power usage is wayyyyyy down.  My amd cards are still ETH/forks but to reduce the heat I could switch them to CN-heavy or other.

And in the winter, its pure profit because the house is heated 99% with mining.  Even Bitmain can't compete with me.  Its just a different scale lol.

[philipma1957]

Unless something crazy happens, the next ASIC miner I buy will probably be from sidehack. Although with the price of BTC down, he is sitting on a sizable number of 2Pac miners that is slowing him down a bit. I'm hoping that he can get the 16nm usb stick miners done before too long, and then get to the 16nm pod miner he's been wanting to do for a while.

When I look at the prices of new ASIC miners and the value of any of the coins they mine, ROI is usually one or more years away with the current network hashrates and difficulty, and we all know how that works out.

Are you serious about buying USB based stick miners at this point in time? Haven't those been long obsolete even with free electricity? As far as i know they earn a few cents per month which really isn't worth it so you will have to be playing solo-mining lottery?

I'm interested to learn more about your "plan" for USB stick miners.

  the 16nm pod miner is what he really wants.

[grendel25]

Thank you martyroz and philipma1957 for the responses.  Still sorting things out... could be next year before I'm ready to move forward with any solar project.  Hope it's okay if I may PM you sometime in the future for more details.  For now, the information you provided gives me something to look into further.  


Oh hey, Martyroz  ... I just realized we pay about the same for power.  Very interesting.  But the difference here is that I don't get a credit for feeding back excess to the grid

[philipma1957]

Hey I like that mounting system that can tilt depending on the season.  Is that something buysolar makes or does he fab those himself?

sent a pm to you and to buysolar.

[Elder III]

Moderately hopeful rendertoken or other forms of gpu marketplace gains traction this year.

Where are rendertoken currently at? Beta-testing still? Not currently giving out live rendering jobs yet are they?

...

EDIT: not considering ASICs at all. Not out of snobism, but because they really make no sense currently unless you're in need of a big door-stopper.

Their roadmap: https://rendertoken.com/#roadmap
With any luck we will be rendering by end of the year.

I'm not a huge fan of ASICs too, but I will still grab some ASICs in promising algorithms to diversify, mainly those with good revenue to electricity cost ratio.

I'm keeping an eye on them too. I really hope that blockchain rendering takes off since I think it's an excellent use of "traditional" GPUs and may prove to be a safeguard against Asics (I think).

[vapourminer]

Moderately hopeful rendertoken or other forms of gpu marketplace gains traction this year.

Where are rendertoken currently at? Beta-testing still? Not currently giving out live rendering jobs yet are they?

...

EDIT: not considering ASICs at all. Not out of snobism, but because they really make no sense currently unless you're in need of a big door-stopper.

Their roadmap: https://rendertoken.com/#roadmap
With any luck we will be rendering by end of the year.

I'm not a huge fan of ASICs too, but I will still grab some ASICs in promising algorithms to diversify, mainly those with good revenue to electricity cost ratio.

I'm keeping an eye on them too. I really hope that blockchain rendering takes off since I think it's an excellent use of "traditional" GPUs and may prove to be a safeguard against Asics (I think).

probably have to rebuild most miner rigs to render. with a lot of rigs having 1 pcie lane per card, 4 gigs ram, undernourished cpus and little storage (120 gig ssd or 32 gig flash drives..  well i cant see how its going to be able to get all the textures and such to the cards in enough time to be useful.

maybe rigs with 4 cards at pcie x4 or 3 cards at pcie x8 along with a good quad+ core cpu with much more ram, 16+ gigs maybe. and storage for jobs so 512 gigs or more. <-- all specs guesswork on my part

wonder how much bandwidth it needs?

[KaydenC]

Moderately hopeful rendertoken or other forms of gpu marketplace gains traction this year.

Where are rendertoken currently at? Beta-testing still? Not currently giving out live rendering jobs yet are they?

...

EDIT: not considering ASICs at all. Not out of snobism, but because they really make no sense currently unless you're in need of a big door-stopper.

Their roadmap: https://rendertoken.com/#roadmap
With any luck we will be rendering by end of the year.

I'm not a huge fan of ASICs too, but I will still grab some ASICs in promising algorithms to diversify, mainly those with good revenue to electricity cost ratio.

I'm keeping an eye on them too. I really hope that blockchain rendering takes off since I think it's an excellent use of "traditional" GPUs and may prove to be a safeguard against Asics (I think).

probably have to rebuild most miner rigs to render. with a lot of rigs having 1 pcie lane per card, 4 gigs ram, undernourished cpus and little storage (120 gig ssd or 32 gig flash drives..  well i cant see how its going to be able to get all the textures and such to the cards in enough time to be useful.

maybe rigs with 4 cards at pcie x4 or 3 cards at pcie x8 along with a good quad+ core cpu with much more ram, 16+ gigs maybe. and storage for jobs so 512 gigs or more. <-- all specs guesswork on my part

wonder how much bandwidth it needs?

Here's an extremely detailed 3hr long discussion on setting up an octane render rig. At 1hr 37 mins they talked about pcie lanes: https://youtu.be/XrOBvz454fU?t=1h37m40s

TLDR: A 8x pcie 3.0 connection per gpu is highly recommended

Long version: Pcie 1x lane works too, and the rendering speed is unrelated to lane bandwidth. However during data transfer the rig will be idle, which happens when you are receiving new job or swapping out the textures in gpu vram. The penalty will be especially heavy for out of core rendering (Scene is 64gb, gpu only has 11gb, there will be loads of transfer from system ram to vram).


When beta testnet launch, we will get a clearer picture on the ideal rig setup. Risers MIGHT still be the best price performance because a Threadripper mobo + 1900x will run $700+ USD and only fits 4 gpu. A gigabyte B250 fintech with 12 gpu and 64gb ram might be better bang for buck.

On a side note the Gigabyte b250 fintech is the only mining mobo with 4x ram slots, and 64gb imo is a must for rendertoken.

I hope someone comes up with a Onda B250 form factor threadripper board, with 7x pcie 8x slots and 2x m2 pcie slots. To run 7x 1180/2080 + 2x acorn fpga accelerator.

[philipma1957]

Moderately hopeful rendertoken or other forms of gpu marketplace gains traction this year.

Where are rendertoken currently at? Beta-testing still? Not currently giving out live rendering jobs yet are they?

...

EDIT: not considering ASICs at all. Not out of snobism, but because they really make no sense currently unless you're in need of a big door-stopper.

Their roadmap: https://rendertoken.com/#roadmap
With any luck we will be rendering by end of the year.

I'm not a huge fan of ASICs too, but I will still grab some ASICs in promising algorithms to diversify, mainly those with good revenue to electricity cost ratio.

I'm keeping an eye on them too. I really hope that blockchain rendering takes off since I think it's an excellent use of "traditional" GPUs and may prove to be a safeguard against Asics (I think).

probably have to rebuild most miner rigs to render. with a lot of rigs having 1 pcie lane per card, 4 gigs ram, undernourished cpus and little storage (120 gig ssd or 32 gig flash drives..  well i cant see how its going to be able to get all the textures and such to the cards in enough time to be useful.

maybe rigs with 4 cards at pcie x4 or 3 cards at pcie x8 along with a good quad+ core cpu with much more ram, 16+ gigs maybe. and storage for jobs so 512 gigs or more. <-- all specs guesswork on my part

wonder how much bandwidth it needs?

Here's an extremely detailed 3hr long discussion on setting up an octane render rig. At 1hr 37 mins they talked about pcie lanes: https://youtu.be/XrOBvz454fU?t=1h37m40s

TLDR: A 8x pcie 3.0 connection per gpu is highly recommended

Long version: Pcie 1x lane works too, and the rendering speed is unrelated to lane bandwidth. However during data transfer the rig will be idle, which happens when you are receiving new job or swapping out the textures in gpu vram. The penalty will be especially heavy for out of core rendering (Scene is 64gb, gpu only has 11gb, there will be loads of transfer from system ram to vram).


When beta testnet launch, we will get a clearer picture on the ideal rig setup. Risers MIGHT still be the best price performance because a Threadripper mobo + 1900x will run $700+ USD and only fits 4 gpu. A gigabyte B250 fintech with 12 gpu and 64gb ram might be better bang for buck.

On a side note the Gigabyte b250 fintech is the only mining mobo with 4x ram slots, and 64gb imo is a must for rendertoken.

I hope someone comes up with a Onda B250 form factor threadripper board, with 7x pcie 8x slots and 2x m2 pcie slots. To run 7x 1180/2080 + 2x acorn fpga accelerator.

so I have a ryzen 1800x  on a board that can do three cards and I have 64gb ram and I have a 2tb ssd

so I guess it would work well to render.

I have this board  https://www.newegg.com/Product/Product.aspx?Item=N82E16813157757

I have this cpu  https://www.newegg.com/Product/Product.aspx?Item=19-113-499

I have this ssd  https://www.newegg.com/Product/Product.aspx?Item=9SIA8TK78R3886

I have this ram https://www.newegg.com/Product/Product.aspx?Item=N82E16820156086

I have 3 of these gpu's  https://www.newegg.com/Product/Product.aspx?Item=N82E16814487346

So this should be good to render.

it is fully paid off.  So roi is not an issue.

[Marvell2]

10k for 5kw that sounds too good to be true

[Marvell2]

Moderately hopeful rendertoken or other forms of gpu marketplace gains traction this year.

Where are rendertoken currently at? Beta-testing still? Not currently giving out live rendering jobs yet are they?

...

EDIT: not considering ASICs at all. Not out of snobism, but because they really make no sense currently unless you're in need of a big door-stopper.

Their roadmap: https://rendertoken.com/#roadmap
With any luck we will be rendering by end of the year.

I'm not a huge fan of ASICs too, but I will still grab some ASICs in promising algorithms to diversify, mainly those with good revenue to electricity cost ratio.

I'm keeping an eye on them too. I really hope that blockchain rendering takes off since I think it's an excellent use of "traditional" GPUs and may prove to be a safeguard against Asics (I think).

probably have to rebuild most miner rigs to render. with a lot of rigs having 1 pcie lane per card, 4 gigs ram, undernourished cpus and little storage (120 gig ssd or 32 gig flash drives..  well i cant see how its going to be able to get all the textures and such to the cards in enough time to be useful.

maybe rigs with 4 cards at pcie x4 or 3 cards at pcie x8 along with a good quad+ core cpu with much more ram, 16+ gigs maybe. and storage for jobs so 512 gigs or more. <-- all specs guesswork on my part

wonder how much bandwidth it needs?

Here's an extremely detailed 3hr long discussion on setting up an octane render rig. At 1hr 37 mins they talked about pcie lanes: https://youtu.be/XrOBvz454fU?t=1h37m40s

TLDR: A 8x pcie 3.0 connection per gpu is highly recommended

Long version: Pcie 1x lane works too, and the rendering speed is unrelated to lane bandwidth. However during data transfer the rig will be idle, which happens when you are receiving new job or swapping out the textures in gpu vram. The penalty will be especially heavy for out of core rendering (Scene is 64gb, gpu only has 11gb, there will be loads of transfer from system ram to vram).


When beta testnet launch, we will get a clearer picture on the ideal rig setup. Risers MIGHT still be the best price performance because a Threadripper mobo + 1900x will run $700+ USD and only fits 4 gpu. A gigabyte B250 fintech with 12 gpu and 64gb ram might be better bang for buck.

On a side note the Gigabyte b250 fintech is the only mining mobo with 4x ram slots, and 64gb imo is a must for rendertoken.

I hope someone comes up with a Onda B250 form factor threadripper board, with 7x pcie 8x slots and 2x m2 pcie slots. To run 7x 1180/2080 + 2x acorn fpga accelerator.

so I have a ryzen 1800x  on a board that can do three cards and I have 64gb ram and I have a 2tb ssd

so I guess it would work well to render.

I have this board  https://www.newegg.com/Product/Product.aspx?Item=N82E16813157757

I have this cpu  https://www.newegg.com/Product/Product.aspx?Item=19-113-499

I have this ssd  https://www.newegg.com/Product/Product.aspx?Item=9SIA8TK78R3886

I have this ram https://www.newegg.com/Product/Product.aspx?Item=N82E16820156086

I have 3 of these gpu's  https://www.newegg.com/Product/Product.aspx?Item=N82E16814487346

So this should be good to render.

it is fully paid off.  So roi is not an issue.

yeah a custom board with 4 ram slots and 4 gpu slots
would be good , i hate risers

[ZedZedNova]

Unless something crazy happens, the next ASIC miner I buy will probably be from sidehack. Although with the price of BTC down, he is sitting on a sizable number of 2Pac miners that is slowing him down a bit. I'm hoping that he can get the 16nm usb stick miners done before too long, and then get to the 16nm pod miner he's been wanting to do for a while.

When I look at the prices of new ASIC miners and the value of any of the coins they mine, ROI is usually one or more years away with the current network hashrates and difficulty, and we all know how that works out.

Are you serious about buying USB based stick miners at this point in time? Haven't those been long obsolete even with free electricity? As far as i know they earn a few cents per month which really isn't worth it so you will have to be playing solo-mining lottery?

I'm interested to learn more about your "plan" for USB stick miners.

Profit is not my only motive here. My plan for stick miners is to learn and introduce others to cryptocurrencies. I do not, and have not had enough money to be a player in the crypto world, although I did get lucky when I mined LTC in late 2011/early 2012 and then just hodl it. The LTC profit I "discovered" in 2016 has funded my forays into other areas, like GPU mining, ASIC mining, and buy/hodl strategies with other altcoins. I have profited from the limited mining I do at home, but not earned a ton.

Besides, sidehacks USB stick miners have cool blinking lights. 

[m.vina]

Besides, sidehacks USB stick miners have cool blinking lights. 

This alone makes the entire purchase worth it.

Kidding aside, congratulations on your LTC profits. Hopefully you've made constructive use of it and actually re-invested / enjoyed the money. If you ever get to buy some of sidehack's USB would you consider putting in an order for me as well? Long live USB mining.

[citronick]

https://www.ccn.com/cryptocurrency-mining-affects-amd-stock-while-nvidia-overestimates-gpu-demand/

Cryptocurrency Mining Affects AMD Stock while Nvidia Overestimates GPU Demand

Demand from cryptocurrency miners has indirectly influenced the fortunes of GPU manufacturers Advanced Micro Devices (AMD) and Nvidia, with the former’s share prices increasing by 30 percent in May 2018 alone.

AMD Stocks Rise on Mining Demand
Since 2017, opportunistic cryptocurrency miners have started to bulk purchase GPU cards to build powerful mining rigs. While the move has augmented both profits and product demand for AMD, analysts are concerned about the immediate drop in stock prices if GPU mining proves to be unprofitable, or ceases to exist, in the near future.

On June 20, Bernstein analyst Stacy Rasgon called out AMD’s widely popular Vega GPU, which was disappointing for investors until cryptocurrency enthusiasts brought relevance to the product.  

“Given AMD’s GPUs have been preferred for mining, and GPU supply in general to gamers was constrained industry-wide, we believe it is plausible that much of AMD’s GPU ramp has benefitted miners, rather than gamers, over this period,” stated Rasgon.

As reported by CCN, AMD earlier confirmed cryptocurrency miners create 10 percent of overall revenue for the company, adding that a “modest decline” is forecasted for Q3 2018.

Joseph Moore, an analyst with Morgan Stanley, echoes Rasgon’s thoughts and remains concerned with the company’s exposure to cryptocurrencies. Moore believes the cryptocurrency sector has “offset the slow and steady progress” of the microprocessor market but has ultimately “raised the bar” for those companies to perform in case the momentum from cryptocurrencies fades.

According to Marketwatch, traders continue to watch AMD stock. 13 company analysts have buy ratings, 12 have hold ratings, and six have sell ratings.

Nvidia Misappropriated GPU Demand
As per a GadgetNow report on June 20, Nvidia’s CEO Jen-Hsun Huang announced the company “would not launch any new GPUs for a long time” after the company overestimated the demand from cryptocurrency miners.

As it stands, Nvidia is stuck with an excess inventory of its flagship 10 series cards, with an Asian OEM partner going to the extent of returning 300,000 GPUs to the American chip maker.

A Taiwanese OEM partner noted:

“This is quite notable news as Nvidia usually exerts massive influence over its partners and can be quite ruthless with allocations of new GPUs if partners step out of line. The fact that this partner returned these GPUs says a lot about the current state of supply in the channel.”

As reported by CCN on May 11, AMD posted revenue of $1.65 Billion for the first quarter of 2018 and claimed 10% of that was from GPU sales to miners. Nvidia, on the other hand, reported revenue of $3.21 billion, and $289 million (9% of their total revenue) was from sales to miners.