Bitcoin Forum
October 31, 2020, 01:22:12 PM *
News: Latest Bitcoin Core release: 0.20.0 [Torrent]
 
  Home Help Search Login Register More  
  Show Posts
Pages: « 1 2 3 4 5 6 7 8 9 10 11 [12] 13 14 15 16 17 18 19 20 21 22 23 24 »
221  Bitcoin / Hardware / Re: Bitfury: "16nm... sales to public start shortly" on: March 05, 2016, 01:38:44 AM
Being able to get an appreciable hashrate at 0.06 and not catch on fire at 0.15 will be difficult. I mean that's going to be drawing a lot of current, and making a lot of heat, per chip. The range I'm building for is about 0.07-0.12 like Phil said. Also hey, PlanetCrypto. Long time no see.

I missed ya'll. Good to communicate, again, with individuals with more than 4 active neurons.
That came out all wrong, but I think you get my meaning.

"Opinions are like butts, everybody's got one and they all stink."
So here's the logic behind my stench:

It's obvious, any implementation is going to require logic level shifting. Cuz' comms to/from the chip are likely to be at 1.8V (or higher 2.5V, 3.3V, or dare I put forth 5V). So that complexity needs to be included in any design.

In vid part 1 (https://www.youtube.com/watch?v=3zPpj1JYw38) he's demo'ing @ Vcore of .378V pulling ~6.0A which equals ~38GH/s per chip with efficiency @ .062W/GH/s.
COLD. A reasonable expectation is current draw will increase as the chip heats up, How much, no clue.
While not an easy 10A single phase buck design, it's doable and been done before. Heatsink optional.
My .06 thought was mirroring his demo. For an exclusively air cooled design I'd consider dropping that to .045 - .050. (buck efficiencies may drop horribly??).
That'd put it back to compac hash rates but with significantly better overall efficiencies.

vid 2, heatsink required.

Vid 3 where he "cranks it up" the heatsink is replaced with immersion cooling. Although the chip, conspicuously, is NOT boiling the Novec fluid. So it ain't getting THAT hot. And is probably still heatsink'able. For comparison Intel CPU's surface temperature in Novec run ~160F in 7100. And they boil like water on the stove. 3M (Phil Tuma) has vids on youtube showing this.

Novec 7000 boils @ 34C (93.2F), Novec 7100 boils @ 61C (141.8F), Novec 7200 boils @ 76C (168.8F), Novec 7300 boils @ 98C (208.4F)

The fact that the fluid is not boiling probably indicates poor heat transfer into the fluid (cuz' it's not hot enough to cause the fluid to change state) i.e. it can't contribute enough energy to rise above the latent heat of vaporization curve. Contributing to my belief that even at that Vcore/clock rate/current draw it's still heatsink'able and doesn't require immersion cooling to survive. Is it gonna' get hot, HELL YES. Does it require esoteric cooling, don't think so. Non boiling fluid provides little convection fluid flow, thereby leaving hot fluid on/near the surface of the chip. And even then, the fluid didn't boil.

The buck to drive it is another potential bone of contention. Cuz' it's drawing just under 40A (38.8A) @ a Vcore of ~.580V PER CHIP. Imagine the current draw of a string of 3 or 4 chips. Similar to the current draw that'll start a car (albeit at a higher V). Unless somebody knows of some magic I'm not aware of, that's probably a multiphase buck. Probably 4, 5, or 6 phases at least for dependability and longevity. Each supplying 10A+. These are a bitch to design, increase component cost, and copper trace content/width/cost/board real estate. All in all, would us immersion cooling guys like to see it, YUP. Matter of fact why not drive it to 60A  Grin. Is it the market segment GS is trying to service, HELL NO  Huh

Hence my upper limit cap of .15W/GH/s for this project.

Nuff said, time to lurk.
222  Bitcoin / Hardware / Re: Community Miner Design Discussion on: March 04, 2016, 06:35:02 PM
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,  Cool 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$ minimum to even try the design & mfg of it. If going for 16/14nm node I'd bet TSMC and any other foundry is going to ask for no less to produce the dies. Also based solely and explicitly on what you tell then 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.

Getting back late to the discussion, so apologies if I state what has already been said.
I'll take the debit for the community ASIC discussion death.
I got sick and basically fell off the planet several months ago. Sucks to get old, BTW. Don't do it, if possible.

At a minimum, the cost to bring an ASIC to tape out is $1.5m
NotFuzzyWarm is spot on in his post.
From a business standpoint, for an entity to layout $1.5-$2.0 mil requires some surety of sales/profit.
If for no reason other than being able to continue operations (i.e. net revenue = $0).
On paper / in posts, it seems that BF has "stepped up to the plate" regarding chip availability.
Time will tell if/when chips actually make it into the community's hands.
And while I'm hopeful, I'm not "holding my breath" waiting for it to happen.

IMHO the kilo17/GS alliance is the communities best bet for affordable miners to empower the continued de-centralization of BTC.
For what little it's worth, we support this endeavor and look forward to it's success.

Also from a business standpoint, kilo17's financial participation is a gutsy move in an unsure industry.

223  Bitcoin / Hardware / Re: Bitfury: "16nm... sales to public start shortly" on: March 04, 2016, 04:23:52 AM
Well right now Kilo is pretty much the designated agent for getting chips. My conversation with Punin got as far as giving him some info to make out and send me an NDA around the end of January (which I never received) and then about two weeks later he dropped a line mostly to let me know he'd posted the 28nm info stuff. That's pretty much it.

I have 2 other people that are actually communicating with me and I will have the chips when they are released.  Wink
AFIK the chip is a fact. In some posts Punin wrote about  eng samples and the video part 1,2,3 in youtube are showing this.
But... when production run will be available and if they are going to sell small quantities or samples is another question .
I think that it is not only up to Punin and big USD will talk Wink

the videos on youtube show a fine chip.

If sidehack builds us a piece of gear that volts down to .07 and up to .12  it will be very nice

Vcore of .3V to .8V would be better, IMHO, spawning efficiencies of ~.06W/GH/s and ~.15W/GH/s respectively.
Guestimates per the vid's
224  Alternate cryptocurrencies / Announcements (Altcoins) / Re: [ANN][DCR] Decred - Hybrid PoW/PoS | btcsuite Devs | Tons of New Features | Go on: January 04, 2016, 12:07:07 PM
Patiently awaiting code release
Signed up for AirDrop.
225  Bitcoin / Hardware / Re: GekkoScience BM1384 Project Development Discussion on: September 24, 2015, 02:08:47 PM
*NIX is more common than you think.

 Pretty much every router "appliance" runs some form of LINUX (Cisco is the only major exception, they have their own propriatary OS stuff).

 Most of the Internet runs on some sort of *NIX - the exceptions are mostly (again!) Cisco boxes in the bigger routers.

 Many older smartphones run on a *NIX of some sort, though propriatary seems to be making a comeback lately with Android getting popular.

 Do keep in mind that the Mac OS is *NIX under the hood.

 To get technical, Windows "borrowed" a LOT of *NIX design concepts in the NT series and it's later derivations, and somewhat to a lesser degree MS-DOS and consumer Windows versions did as well. MS-DOS also borrowed heavily from the older DEC RT-11 OS though (both RT-11 and UNIX borrowed from older OSs as well).

Early Cisco IOS's were heavily based on Linux.
When NT was released NT had dynamic memory allocation/deallocation, Linux did not.
To me DOS stole more from CP/M & MP/M than the PDP 11 OS.
Much of the NT design team was "lured" away from DEC, so it makes sense that NT has "under the hood" similarities to RT-11.

Core Dumped Blues

Well, my terminal's locked up, and I ain't got any mail,
And I can't recall the last time that my program didn't fail;
I've got stacks in my structs, I've got views in my queues,
I've got them : Segmentation violation -- Core dumped blues.

If you think that it's nice that you get what you C,
Then go : illogical statement with your whole family,
'Cause the Supreme Court ain't the only place with : Bus error views.
I've got them : Segmentation violation -- Core dumped blues.

On a PDP-11, life should be a breeze,
But with VAXen in the house even magnetic tapes would freeze.
Now you might think that unlike VAXen I'd know who I abuse,
I've got them : Segmentation violation -- Core dumped blues.
226  Bitcoin / Hardware / Re: GekkoScience BM1384 Project Development Discussion on: September 21, 2015, 12:18:04 PM
got them running great on a debian 7 build .  19 sticks .

@Jake36 thanks for help on the cgminer install



 I will post a shot later. -------- shot below

This was the first linux build I ever did.  I burned a boot cd  took a few hours to set up I

still have some quirks , but those are my quirks.  Since I never did this os before.

 I have good numbers for the sticks

19 on 2 hubs  freq 125  


Mighty impressive.
AND
Congrats on your first Linux build.

Most build Linux 4-6 times the first time, there are just SOOOOOO many options cuz' Linux is so configurable.
What distro did ya' start with?
ALL our servers run Centos 6, which may not be optimum for what your doing.
227  Bitcoin / Hardware / Re: GekkoScience BM1384 Project Development Discussion on: September 19, 2015, 09:02:24 PM
Sticks came in yesterday, Muy Bueno.
Took me a couple of hours to find the support thread and Novak's brilliant work.
Now hashing @ Eligius @ 200 MHZ, ~10Gh/s each.
Them LEDs are BRIGHT, Tooooooo F'ing cool (I'm such a redneck).
228  Bitcoin / Hardware / Re: GekkoScience BM1384 Project Development Discussion on: September 15, 2015, 11:36:49 AM

"... but how else do you stop companies from doing this if you require their service for your own business?"

Make it yourself, and if you can't do it where you are, move to where you can.

I moved from Dover, DE in 97-98 partially because power was becoming SOOOOO expensive (I read the writing on the wall) and the city wouldn't let me do wind/solar/bio-fuels at my house on Division Street.
First ~5-7 years here in MN, I made ALL my own power (and did not live like Grizzly Adams).
In '01 I got involved in county government and wrote the code and zoning for wind/solar (it's easy to win, if you write the rules).
In '05 I had'em (my local corrupt co-op) run in power, cuz' it was cheap.
It's tripled in cost since then.
With the cost of solar being so cheap, I'm re-visiting self generation.
And since I wrote the code, I know what I can do and what I can't.

If I can't get efficient chips, I'll fight this battle from the supply side versus the consumption side.
Besides, I remember how good it felt to NOT have a re-occurring monthly electrical expense.

During the school year, every Friday I mentor the local 7-12 grade science classes with my "radical" ideas.
Presenting strong economic reasons for doing what I do/believe (the proof is in the pudding, so to speak).
Cuz' it's hard to argue with numbers in black'n'white (Reality What a Concept -- Robin Williams).

What I'm trying to say is: it's not enough to just do it, one has to bank it forward to the next generations.
I endeavor to teach them they have options, to critically think about what's possible and look for better solutions.

Even the USB stick miners are profitable if there is no electrical expense.
Think about what it would cost to buy an RPi, a 13 port USB hub, 10 stick miners, a big battery (L16 HC's), a solar charge controller, a 12Vdc to 5Vdc Buck, and 1-3 solar panels.
There is always more than one way to skin a rat.

I'll get off my soap box . . . . . . .
http://www.ted.com/talks/david_rothkopf_how_fear_drives_american_politics?utm_source=newsletter_daily&utm_campaign=daily&utm_medium=email&utm_content=image__2015-09-14
229  Bitcoin / Hardware / Re: GekkoScience BM1384 Project Development Discussion on: September 12, 2015, 02:34:59 AM
Bit Innosilicon isn't willing to sell sample chips. Right now they're looking for large batch investments, which sure as heck isn't gonna happen without playing with a few first to make sure they work right. If PC hasn't heard fresh news from Innosilicon, I have to assume that's still their stance.

We left our conversations with Innosilicon at "when you get productions chips we'll buy them in quantity".
To be honest, I'm not the slightest bit interested in buying experimental prototypes which will likely differ considerably from production units.
And having been badly burned in pre-sales schemes before, the statement "not only NO, but F#$% NO" comes to mind.

Acquiring experimental prototype chips (assuming they'd sell them) and forwarding them to GS for board design seems like a huge waste of their time (and a waste of our money).
Given the high probability that the specs on the production chip will be radically different as a result of prototyping.

Put bluntly, if we were going to buy a large quantity of chips (250K-500K) we'd fund the design effort to make one of our own.
This concept is not "off the table".
230  Bitcoin / Hardware / Re: GekkoScience BM1384 Project Development Discussion on: September 12, 2015, 02:19:24 AM
If PC gets SFARDS chips PC won't get GekkoScience boards. I've said it a dozen or two times already, but big BGA is stupid and hard to work with.

We have little to no interest in the SFARDS chip, for a plethora of reasons already stated in this thread.
231  Bitcoin / Hardware / Re: GekkoScience BM1384 Project Development Discussion on: September 11, 2015, 05:28:22 PM
Heard from anyone else? Avalon, Innosilicon?

Negative.
232  Bitcoin / Hardware / Re: GekkoScience BM1384 Project Development Discussion on: September 11, 2015, 03:48:08 AM
BTW, still no response from BM about chips.

I guess the squeaky wheel gets ignored.

Just call me the Minnesota Mushroom: Kept in the cold dark and not feed shit.
233  Bitcoin / Hardware / Re: GekkoScience BM1384 Project Development Discussion on: September 09, 2015, 04:53:48 PM
One of the things that has always pissed me off about buying electricity is one never knows "how deep" one is in till the bill comes in.
Doesn't give me any current usage feedback and consequently the ability to trim usage to stay within a budget.

First 5 years I lived in my house I made all my electricity with no grid tie backup.
And did NOT live like "Grizzly Adams" (Big screen TV, Microwave, Surround sound stereo, forced air heat, etc. . .)
Was very apparent when consumption rose above norms.
Electricity cost here has tripled since Dec 2005 (the date I ran in grid power) and am re-visiting generating all my own power again given the current trends.

Pretty sad statement when the end consumer can generate electricity cheaper than multi-billion dollar companies.
234  Bitcoin / Hardware / Re: GekkoScience BM1384 Project Development Discussion on: September 06, 2015, 03:59:57 PM
MN isn't THAT much colder than where I'm at in IA, and has similar wind levels.
6-8 months perhaps if he's in the northern part, and some of that only part of the day.

I beg to differ. I'm in North Central MN (near Leech Lake), > 300 miles north of the IA/MN border.
Our winters are significantly colder and longer than say Des Moines.
About 3 months (June, July, and August) outta' the year ambient temps rise above ~75F

Now I'm trying to figure out how his electric company generates the power cheap enough to SELL it at that rate.
Only thing I can think of in THAT area would be nuclear....

Yup, lotta nukes (Excel Energy).
GRE (Great River Energy) is coal, coal, and coal.
Minnesota Power is a mix of nukes, coal, wind, and up here, wood gasification (this is logging country).

We also have some solar and wind on site.
The $0.0465/KwH is off peak (12 hours/day) which will feed/recharge our UPS's/battery bank at night.
During the day solar augments the $0.12/KwH service.
We are expanding our little solar farm as money permits.
With the goal to be exclusively solar ASAP.
The concept is to buy the power once (a depreciating capex expense) versus suffer an ever increasing re-occurring expense (CODB, Cost Of Doing Business).
Once the off peak is online our aggregate power cost should be around $0.05 - $0.06 / KwH and decreasing as we bring more solar online.

MN was one of the few states that implemented (back in the 70's) fair net excess generation statutes.
Put simply, we can buy power @ $0.12/Kwh and can sell power @ $0.105 /Kwh.
Obviously, if our power generation is in a deficit we want to use everything we generate (grid tieing our generation facilitates this).
IF/when BTC mining becomes unprofitable (which it is becoming) we just turn off the miners and the excess generation becomes the profit center.

Another option is to power the (inefficient) miners exclusively with off peak power and only mine 12 hours/day.

Like so many things in this world, there is no single solution to a problem.
It takes a mix of technologies, intelligently arranged, to solve a problem.
In this case the electrical cost to mine BTC.
I like having options.
235  Bitcoin / Hardware / Re: GekkoScience BM1384 Project Development Discussion on: September 05, 2015, 04:35:38 PM
To date have not heard back from BM about chips.
That's either real good or real bad.
236  Bitcoin / Hardware / Re: GekkoScience BM1384 Project Development Discussion on: September 05, 2015, 04:32:22 PM
I hear tell he's working on negotiating lower power rates.

No secrets here, am working through getting $0.0465/Kwh power online.
Which requires a BUTTLOAD of BS paperwork and all new equipment (Load centers, outlets, conduit, meter bases, etc, ad nausem), which ofc has to be to NEC.
237  Bitcoin / Hardware / Re: GekkoScience BM1384 Project Development Discussion on: September 05, 2015, 04:26:55 PM
Don't forget to consider you'll probably want to switch coins relatively often as fads change and bubbles burst.

Yet another reason I've never been an Altcoin fan.
238  Bitcoin / Hardware / Re: GekkoScience BM1384 Project Development Discussion on: September 05, 2015, 04:25:57 PM

The Rocks Compute Cluster is used primarily to generate 4K/8K/16K DH keys and large Mersenne primes.


 GIMPS?
 Or something else?


Something else, an in-house written DH Key generation routine which uses as it's seed a large Mersenne Prime.
These large primes are no where near as large as the primes GIMPS is working on, but still fall into the definition of large.


 I'm guessing it's the DH key that's soaking all the computation power, as all Mersenne primes up to the 44'th one have been calculated and verified, along with 4 more Mersenne primes past that but not all of the "inbetween" candidates after the 44'th prime have been verified.

 I can only think of one person/group that is likely to be putting $4k/month into the GIMPS effort - Dr. Cooper of one of the Missouri universities - and that's WITH the backing of the university administration to run GIMPS on most of the computers at that university.

 At the kind of power level you use, why aren't you in a "cheap power" area? Seems like dropping your power costs from 12/kwh to under 5/kwh would save you a TON of money.


You're spot on, the DH key generation is the core hog.
Using large primes as seeds, to some extent, speeds up large DH key generation but also makes the keys more deterministic due to the finite number of seeds (large primes).

To those ends, we are developing, in H/W, a non-deterministic large TRNG (True Random Number Generator) ASIC.
Where large is defined as 16K+ bits, implemented as a USB dongle, and can generate 5-10 TRN's per second.

Most, if not all of the H/W TRNG's I've evaluated are deterministic. As most use the noise off of an electronic circuit as the seed. Given a large enough sample it becomes repetitive (shades of chaos theory), i.e. crackable with a brute force attack. And a  large number of those also have "backdoors", i.e. one can force the TRNG output to a known value through simple code implemented in the corresponding driver software.

Wonder why that is?Huh Ahhhh, the illusion of security/privacy.

So if someone tells me they have a TRNG, IMHO, they're lying through their teeth.
We aim to change that.
239  Bitcoin / Hardware / Re: GekkoScience BM1384 Project Development Discussion on: September 05, 2015, 03:57:25 PM
Huh, have you considered using your cluster to mine altcoins that are cpu-friendly? CryptoNight comes to mind as an algorithm with little to no gpu advantage (and no FPGA or ASIC implementations).

Since you've already got the hardware, it's free money--provided the revenue is greater than your power bills.

Have never been a Altcoin kinda' guy so never looked into it, think that's about to change.
Thanks for the suggestion.
Most of the nodes are nVidia card enhanced (C20's and Quadro 5600's).
My preliminary guess is the power cost will be larger than the revenue stream.
As each node "wastes" a lot of energy on "house keeping" chores.
i.e. power not used in computation.

I'll run some calcs and see if it's worth the hate, hassle, and discontent to run a couple of day test.
And report back with findings.
240  Bitcoin / Group buys / Re: [SIDEHACK STICK] Official sales thread for everywhere not already covered on: September 04, 2015, 05:48:04 PM
Sent the 0.242536 BTC ($56) for 2 Compacs to 1BURGERAXHH6Yi6LRybRJK7ybEm5m5HwTr.
Shipping address in message.

Thanks!
Pages: « 1 2 3 4 5 6 7 8 9 10 11 [12] 13 14 15 16 17 18 19 20 21 22 23 24 »
Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!