Community Miner Design Discussion

[navigatrix]

Speaking of connectors: I was trying to find the list of bands & performers endorsing Neutrik and I couldn't find it. Maybe they discontinued the endorsement program because nowadays if any well-known national or worldwide act is performing on stage they have their equipment plugged in through the Neutrik connectors.

Take a look at the Neutrik's version of Ethernet, USB and power plugs:

http://www.neutrik.us/en-us/ethercon/
http://www.neutrik.us/en-us/multimedia/
http://www.neutrik.us/en-us/powercon/

and compare them to the stuff available at Amazon or your local electronics retailer.

I'm not going to try to post the prices, because some readers could get a heart attack and some others could choke laughing. But think of Neutrik next time you'll be warming your glue gun to fix loose connectors.

Maybe a bit off the topic here, but having spent a lot of time in a recording studio, I am a big fan of Neutrik. Audio demands robust and accurate connections at every junction.

[chiguireitor]

So, after skimming the thread a little... i only have one question to y'all:

Why the fuck aren't we organizing to get a community chip done?

I know, i know, scary asic costs and such, but does anyone have idea of the process? is anyone informed at all?

PlanetCrypto once suggested on another thread that they could deal with the asic side, if sidehack was willing to build the miners, did this die off? is someone around up to snuff for FPGA to ASIC design?

[NotFuzzyWarm]

So, after skimming the thread a little... i only have one question to y'all:

Why the fuck aren't we organizing to get a community chip done?

I know, i know, scary asic costs and such, but does anyone have idea of the process? is anyone informed at all?

PlanetCrypto once suggested on another thread that they could deal with the asic side, if sidehack was willing to build the miners, did this die off? is someone around up to snuff for FPGA to ASIC design?

For a start. perhaps someone here can do an ASIC die layout. If so,   Then the real expensive part comes: I don't know first-hand about TSMC but - we work with the #1 chip packaging company in Taiwan - folks that mount the silicon die into the final chip package  - and they will not even talk to a customer without a 100k$ up-front minimum to even look at the design & mfg of it. That is just to get past the initial sales meeting and have them evaluate the feasibility of your product with no guarantees of it being viable.

If going for 16/14nm node I'd bet TSMC and any other foundry is going to ask for no less to look at producing the dies. Also based solely and explicitly on what you tell them to do they will do it, even if the design itself is wrong or not properly spec'd to the last detail. If you want design suggestions from them regarding your design specs and expectations - that costs extra and you MUST ask for their input.

[kilo17]

So, after skimming the thread a little... i only have one question to y'all:

Why the fuck aren't we organizing to get a community chip done?

I know, i know, scary asic costs and such, but does anyone have idea of the process? is anyone informed at all?

PlanetCrypto once suggested on another thread that they could deal with the asic side, if sidehack was willing to build the miners, did this die off? is someone around up to snuff for FPGA to ASIC design?

If you read the thread and not skim it you will see that we do not have interest in producing large amounts -- all that does is help the big boys -- we want the hobbyists and home miners to have a miner. 
So to be blunt, offsetting the costs of manufacturing an ASIC would require either a REALLY expensive miner or a whole bunch of miners -- NEITHER PERTAIN TO THIS THREAD 

[sidehack]

Although... it would be pretty cool.

[chiguireitor]

So, after skimming the thread a little... i only have one question to y'all:

Why the fuck aren't we organizing to get a community chip done?

I know, i know, scary asic costs and such, but does anyone have idea of the process? is anyone informed at all?

PlanetCrypto once suggested on another thread that they could deal with the asic side, if sidehack was willing to build the miners, did this die off? is someone around up to snuff for FPGA to ASIC design?

If you read the thread and not skim it you will see that we do not have interest in producing large amounts -- all that does is help the big boys -- we want the hobbyists and home miners to have a miner. 
So to be blunt, offsetting the costs of manufacturing an ASIC would require either a REALLY expensive miner or a whole bunch of miners -- NEITHER PERTAIN TO THIS THREAD 

As i've stated before on other threads, if there was a custom asic from the community, it could be done with either TSSOP or QFP, which would really be a hobbyist's boon. That, and the fact that everyone says it is expensive, but no-one has real numbers (apart from NotFuzzyWarm).

I just registered on OnSemi and they seem rather open about their process and even have FPGA to ASIC tutorials and such... It has been like, what, 10 years since the last time i touched an FPGA, and still seems like Spartan 3 is the hottest thing in the block. It shouldn't be THAT difficult to get a real quote based on a bitstream.... just sayin'

[kilo17]

Maybe I need a sandwich also 

[chiguireitor]

Maybe I need a sandwich also 

I just had two, imagine my happiness

[2112]

As i've stated before on other threads, if there was a custom asic from the community, it could be done with either TSSOP or QFP, which would really be a hobbyist's boon. That, and the fact that everyone says it is expensive, but no-one has real numbers (apart from NotFuzzyWarm).

I just registered on OnSemi and they seem rather open about their process and even have FPGA to ASIC tutorials and such... It has been like, what, 10 years since the last time i touched an FPGA, and still seems like Spartan 3 is the hottest thing in the block. It shouldn't be THAT difficult to get a real quote based on a bitstream.... just sayin'

I'm sorry to say this, but someone has to do it.

You must have spent too much time soldering with non-lead-free solder and are now suffering some latent effects of https://en.wikipedia.org/wiki/Lead_poisoning .

The numbers (of dollars and of days) just don't work out for that business model. Many people have had run that calculation independently and they all came to the same conclusion. They all could be wrong or you could be wrong.

I mean seriously: Spartan 3? Xilinx is already depreciating Spartan 6 and sampling Spartan 7. If not lead then maybe you need less sandwiches.

[chiguireitor]

...

I'm sorry to say this, but someone has to do it.

You must have spent too much time soldering with non-lead-free solder and are now suffering some latent effects of https://en.wikipedia.org/wiki/Lead_poisoning .

The numbers (of dollars and of days) just don't work out for that business model. Many people have had run that calculation independently and they all came to the same conclusion. They all could be wrong or you could be wrong.

I mean seriously: Spartan 3? Xilinx is already depreciating Spartan 6 and sampling Spartan 7. If not lead then maybe you need less sandwiches.

It strikes me as weird this ad-hominem response of you, you have been generally toughtful on the forums when talking about hardware... maybe more sandwiches?

As always, when there's a market, there's opportunity, and there's little to no presence on the low-volume, low-tech requirement part of the market.

Maybe the numbers have been ran by peeps just looking for a quick buck. Wouldn't surprise me, most peeps here get scared by things as 12 month ROIs and such, which is like the minimum when dealing with traditional manufacture.

Also, as i said, 10 years without touching FPGAs, i know i'm out of touch there, but the market seems ripe with Spartan 3E and 6 modules, which i wouldn't have expected (i tought that would be long depreciated on the market already, but it seems i was wrong).

Finally, someone's bound to fill this niche, whoever does it first will have a good start in the field, as it stands today, we're probably watching a whole revolution start in front of our eyes, and big chip developers will eye this technology eventually.

[NotFuzzyWarm]

Unless in the future they get the per-chip power needs below ~1 or 2watts at most you can't use TSSOP or QFP packages. I've never seen any used to remove that much heat from a die without slathering their bottoms with compound to let the board pull more heat away. Even then, iffy at best for power chips. QFN package are not only smaller but have the large contact pads to give direct bonding for power feed and heat removal.

[sidehack]

I've seen TSSOP and QFP with belly pads.

[NotFuzzyWarm]

I've seen TSSOP and QFP with belly pads.

Good to learn something new. Makes sense since there is all that otherwise unused real estate under it.

[sidehack]

Yeah. I like the 10W/chip neighborhood but that pretty much requires a belly pad or a really big chip. I also like pins (makes verifying connections so very easy, and 95% of the problems I have with Compac assembly are ASIC connections), so something like a QFP with a pad would be nice.

[2112]

As always, when there's a market, there's opportunity, and there's little to no presence on the low-volume, low-tech requirement part of the market.

The "hobby miner" is about the same market opportunity as the impoverished compulsive gamblers converging on the racetrack on the race days. When I was a kid I had to transfer between buses on a stop right in front of the racetrack. I remember the type very well: their grim determination, the fantabulous stories they were telling each other: how they are going to buy or breed a horse who will win the race for them; how in some distant city there's a racetrack that offers the odds that are less of a rip-off; how they instead of buying tip-sheet will start publishing a tip-sheet; etc.

This place is like that racetrack, except that the racetrack wouldn't let in those who are underage.

The niche markets in obsolete electronics exist when buyers have money. Here's an example for the those with a fetish for obsolete electronics: hand-wound moving-coil phonograph record pickups from Koetsu: http://www.elusivedisc.com/Koetsu/products/248/ ranging in price from $19,949.99 for Blue Lace Platinum

to the pedestrian $2,495.00 for Black Goldline

You could then play your treasured vintage vinyl records through a vacuum tube amplifier from
McIntosh Labs http://www.mcintoshlabs.com/us/Products/pages/ProductDetails.aspx?ProductId=MC275B

.

But here nobody has significant money, thus it is a niche, but not a market.

That picture is from https://en.wikipedia.org/wiki/Bootleg_mining, but much better article is under https://en.wikipedia.org/wiki/Artisanal_mining .

[NotFuzzyWarm]

You left out the turntables. https://www.google.com/search?q=audiophile+turntable&tbm=isch&imgil=yR8ZGCJ59_gEhM%253A%253BQmX2Kwxn5zxeFM%253Bhttp%25253A%25252F%25252Fwww.dagogo.com%25252Famg-viella-v12-turntable-review&source=iu&pf=m&fir=yR8ZGCJ59_gEhM%253A%252CQmX2Kwxn5zxeFM%252C_&usg=__AtcSIlEJ08EseIYih8JOxRSgKWM%3D&biw=1399&bih=881&ved=0ahUKEwiEn-fV_ZDLAhVDn4MKHX-KDz0QyjcIWw&ei=9e_NVsTQFsO-jgT_lL7oAw#imgrc=yR8ZGCJ59_gEhM%3A
Some of those bad boys can cost over 100k$. Love the ones with 4 tonearms. Prolly each for playing the owners specific type of record or music I guess. Toss in 4 of those 20k$ pickups ya linked to, woof...

Must be nice to find and supply a market like that....

[Meech]

A little late to the discussion but very pleased to hear Notfuzzywarm and jstefanop are in talks with sidehack.  Great pool of minds and collective..   

[sidehack]

I like serving niche markets and also specifically not overcharging.

[toptek]

You left out the turntables. https://www.google.com/search?q=audiophile+turntable&tbm=isch&imgil=yR8ZGCJ59_gEhM%253A%253BQmX2Kwxn5zxeFM%253Bhttp%25253A%25252F%25252Fwww.dagogo.com%25252Famg-viella-v12-turntable-review&source=iu&pf=m&fir=yR8ZGCJ59_gEhM%253A%252CQmX2Kwxn5zxeFM%252C_&usg=__AtcSIlEJ08EseIYih8JOxRSgKWM%3D&biw=1399&bih=881&ved=0ahUKEwiEn-fV_ZDLAhVDn4MKHX-KDz0QyjcIWw&ei=9e_NVsTQFsO-jgT_lL7oAw#imgrc=yR8ZGCJ59_gEhM%3A
Some of those bad boys can cost over 100k$. Love the ones with 4 tonearms. Prolly each for playing the owners specific type of record or music I guess. Toss in 4 of those 20k$ pickups ya linked to, woof...

Must be nice to find and supply a market like that....

I had no idea turntables were still there and cost so much  are they used mostly for live rape?  sense all the media in stores is either in CD Or DVD form these days unless you go to a used record store and yea there are still some of those around .
I know in a recording studio they kind of use custom turn tables.  

Thanks sidehack

for being there that can't be said enough.


cya.

[2112]

You left out the turntables.

Some of those bad boys can cost over 100k$. Love the ones with 4 tonearms. Prolly each for playing the owners specific type of record or music I guess. Toss in 4 of those 20k$ pickups ya linked to, woof...

Must be nice to find and supply a market like that....

For some this is probably just a status symbol. For professionals multiple tonearms are just a time-saving convenience. Precisely adjusting a tone-arm & cartridge combo takes time. Vinyl records didn't precisely follow standards and on top of it needles&cartridges react differently to different kinds of wear of the groove and warping of the whole record.

Here's an example of a high-end transcription turntable with optical contactless pickup:

https://www.bostonglobe.com/arts/2014/04/05/pushing-back-silence-new-technology-and-battle-save-old-recordings/8ccQ3EPHdc7TI6GnxK8QtM/story.html

I couldn't find pictures of the US Library of Congress vinyl record transcription setups, that one will have to do.

There's a big difference between those niche markets for electronics and Bitcoin mining.

If somebody is capable of producing a top-performing money printing machine he is also capable of running it at the top performance in his own garage. He doesn't need to sell it to the owners of other garages. He's also capable of buying another garage where price of electricity is cheaper than his original garage.

On the other hands vendors of gold pans and shovels couldn't profitably employ them in their own garage. They had to sell them to those willing to carry them up the mountains in Klondike.

So almost by definition the ASIC vendor is forced to either: (a) self-mine or (b) find a buyer who will willingly overpay. The other profitable strategy is to sabotage competitor's mines. One way to make buyer overpay is deprive the buyer from the full and correct documentation of the sold ASIC.

This is the essence of the difference between a niche and a niche market.