Bitcoin Forum
May 06, 2024, 02:22:00 AM *
News: Latest Bitcoin Core release: 27.0 [Torrent]
 
   Home   Help Search Login Register More  
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 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 ... 221 »
  Print  
Author Topic: Avalon ASIC users thread  (Read 438337 times)
Beepbop
Full Member
***
Offline Offline

Activity: 126
Merit: 100



View Profile
February 02, 2013, 02:23:35 PM
 #81

Have you tried measuring the power without the UPS square wave being invovled? I.o.w. plug it straight into the wall (or perhaps with a transient protector only)?
1714962120
Hero Member
*
Offline Offline

Posts: 1714962120

View Profile Personal Message (Offline)

Ignore
1714962120
Reply with quote  #2

1714962120
Report to moderator
1714962120
Hero Member
*
Offline Offline

Posts: 1714962120

View Profile Personal Message (Offline)

Ignore
1714962120
Reply with quote  #2

1714962120
Report to moderator
If you see garbage posts (off-topic, trolling, spam, no point, etc.), use the "report to moderator" links. All reports are investigated, though you will rarely be contacted about your reports.
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction.
1714962120
Hero Member
*
Offline Offline

Posts: 1714962120

View Profile Personal Message (Offline)

Ignore
1714962120
Reply with quote  #2

1714962120
Report to moderator
1714962120
Hero Member
*
Offline Offline

Posts: 1714962120

View Profile Personal Message (Offline)

Ignore
1714962120
Reply with quote  #2

1714962120
Report to moderator
1714962120
Hero Member
*
Offline Offline

Posts: 1714962120

View Profile Personal Message (Offline)

Ignore
1714962120
Reply with quote  #2

1714962120
Report to moderator
jgarzik (OP)
Legendary
*
Offline Offline

Activity: 1596
Merit: 1091


View Profile
February 02, 2013, 02:25:17 PM
 #82

What are the odds that you could get an Avalon to run with a program like 50miner or guiminer?

Supposed the source will be opened, probably after Yifu gets some sleep and batch #2 sales </guess>

In that case, other miners may be modified and used.  It appears to be a modified cgminer, modified kernel (Avalon device driver), and otherwise stock OpenWRT install.


Jeff Garzik, Bloq CEO, former bitcoin core dev team; opinions are my own.
Visit bloq.com / metronome.io
Donations / tip jar: 1BrufViLKnSWtuWGkryPsKsxonV2NQ7Tcj
jgarzik (OP)
Legendary
*
Offline Offline

Activity: 1596
Merit: 1091


View Profile
February 02, 2013, 02:25:52 PM
 #83

Have you tried measuring the power without the UPS square wave being invovled? I.o.w. plug it straight into the wall (or perhaps with a transient protector only)?

No, though several have requested that.

It would be good to eliminate the UPS as a factor, agreed.


Jeff Garzik, Bloq CEO, former bitcoin core dev team; opinions are my own.
Visit bloq.com / metronome.io
Donations / tip jar: 1BrufViLKnSWtuWGkryPsKsxonV2NQ7Tcj
wndrbr3d
Hero Member
*****
Offline Offline

Activity: 914
Merit: 500


View Profile
February 02, 2013, 03:20:38 PM
 #84

Have you tried measuring the power without the UPS square wave being invovled? I.o.w. plug it straight into the wall (or perhaps with a transient protector only)?

Most consumer UPS systems don't run inline AC->DC->AC (this is why your UPS doesn't act as a space heater while the power is still on), but basically act as a surge protector until power cuts then a relay throws starting the DC->AC which usually converts a square DC to a staggered AC.

At worst the UPS is just trickle charging off the AC line. Usually only data center UPS's run 100% live through battery.
organofcorti
Donator
Legendary
*
Offline Offline

Activity: 2058
Merit: 1007


Poor impulse control.


View Profile WWW
February 02, 2013, 03:36:48 PM
 #85

I have a post in progress at http://organofcorti.blogspot.com/2013/02/98-asic-choices-avalon-update.html

All the charts are up and are self explanatory - it's an update based on Jeff's 620W power consumption figure, and the charts show, amongst other things, what you might earn if you have a batch1 or batch2 Avalon and you're paying the high electricity prices.

I'll update it with words when I have time. Right now I'm going to bed, sad that the Avalon store did not work for me.

Bitcoin network and pool analysis 12QxPHEuxDrs7mCyGSx1iVSozTwtquDB3r
follow @oocBlog for new post notifications
PuertoLibre
Legendary
*
Offline Offline

Activity: 1834
Merit: 1003


View Profile
February 02, 2013, 04:19:53 PM
 #86

thanks for clearing that up ngzhang.

Makes sense as it's a stand alone unit.  When quoting BFL's numbers, no one includes the PC it's hooked up to either.
Which should be referred to as The "Tethered Wattage".

No BFL customer is going to bring that bit up without figuring out a cheap lower power alternative. Even a laptop adds 50% to that power figure. A desktop, much more.
PuertoLibre
Legendary
*
Offline Offline

Activity: 1834
Merit: 1003


View Profile
February 02, 2013, 04:22:26 PM
 #87

Chip's power consumption is <6.6W/Ghs

ATX power loses(82%), DC/DC power loses(87%), Fan(5~20w), etc…

So…

Dear Dr. NGZHANG,

That's not the reason for you to claim the power consumption of the unit is 450w.

The power consumption of the unit is what all it matters. You picked on BFL for the same thing and now you talked like its all fine and dandy.

LOL at ... saying fans power and all that. Should I add the power I need to cool my room?

Actually, yes.

When summer comes around you will need it regardless of which vendor you go with. If you "need it" to run a Bitcoin mining operation then it should be included in the final calculations.
organofcorti
Donator
Legendary
*
Offline Offline

Activity: 2058
Merit: 1007


Poor impulse control.


View Profile WWW
February 02, 2013, 04:43:21 PM
 #88

I have a post in progress at http://organofcorti.blogspot.com/2013/02/98-asic-choices-avalon-update.html

All the charts are up and are self explanatory - it's an update based on Jeff's 620W power consumption figure, and the charts show, amongst other things, what you might earn if you have a batch1 or batch2 Avalon and you're paying the high electricity prices.

I'll update it with words when I have time. Right now I'm going to bed, sad that the Avalon store did not work for me.

I was too sad to go to bed so I finished the post. Beware! It's nearly 4am and there will be typos.

http://organofcorti.blogspot.com/2013/02/98-asic-choices-avalon-update.html

Bitcoin network and pool analysis 12QxPHEuxDrs7mCyGSx1iVSozTwtquDB3r
follow @oocBlog for new post notifications
PuertoLibre
Legendary
*
Offline Offline

Activity: 1834
Merit: 1003


View Profile
February 02, 2013, 05:53:21 PM
Last edit: February 02, 2013, 06:03:39 PM by PuertoLibre
 #89

Chip's power consumption is <6.6W/Ghs

ATX power loses(82%), DC/DC power loses(87%), Fan(5~20w), etc…

So…

650W * 0.82 * 0.87 - 20W = 443W, About right.

so an equivalent bfl number would be something like:

60w / .87 / .82 + 2W (fan) + 5W (host) = 91 W    (host can be discounted for many since it can run several units)

or "bfl's worst case" 60W * 1.2 = 72W ===> 72w / .87 / .82 + 2W (fan) + 5W (host) = ~108W


Unless you want Inaba to eat you alive, never quote that reality, ever again. Wink

By the way, if BFL overclocks, it will raise even that amount by a bit more. Adding the host just raises it further.
DeathAndTaxes
Donator
Legendary
*
Offline Offline

Activity: 1218
Merit: 1079


Gerald Davis


View Profile
February 02, 2013, 06:19:18 PM
 #90

DC Power supplies are marked by the DC wattage provided, not by the AC wattage consumed.  There is nothing at all strange about a device marked "450W" pulling 600+ watts from the wall.  

This is a 91% efficiency power supply. So even if that was "normal" 620W AC is ~565W DC.  565W DC > 400W DC.

Also remember the 400W number comes from the product listing for the entire unit.  It is "strange" for a company to have all the specs related to the entire unit except the power which just happens to be just the chip (not even the total DC wattage).  Now for the module only listing putting a wattage of ~180W DC would be realistic. 

Still even if you take 620W AC remove the PSU inefficiency getting 565W DC then drop out 25W for the host and 3 5W fans = 540W not 400W.  A single module by itself has to be pulling ~180W ea (3x180 = 540) not ~133W ea (3x133W=400W).   So the 400W number is dishonest, there is no way or reason to spin that.  The sad thing is first out the door @ 180W per module and 620W per unit is still impressive.  It is dishonesty that isn't even needed.
Aseras
Hero Member
*****
Offline Offline

Activity: 658
Merit: 500


View Profile
February 02, 2013, 06:31:45 PM
 #91

So, it does not work with p2pool. right?

Just mined all night on p2pool.

It looks like fixing the air flow/physical configuration solved the issues.

No miner or machine restarts while I was sleeping, and while it was mining p2pool.

(no p2pool blocks either, alas)



Are you using getwork or stratum on p2pool? Did you try both? Any difference between the two?
firefop
Sr. Member
****
Offline Offline

Activity: 420
Merit: 250


View Profile
February 02, 2013, 06:49:55 PM
 #92

Did I tell you to shut up?

BFL can plug into a OpenWRT device LIKE one in AVALON. Even an Android phone would work.

Your question is does Raspberry Pi gonna have enough bandwidth as a host? Fck yes, it also shows you know crap about this.


Avalon engineered the OpenWRT solution to work effectively.

I don't know what they had to do to make that happen, but I know that you can get away with using a low computer power single purpose device if well engineered to the application.

Everyone in this thread seems to be suggesting the end consumer should slap some 5W toy device called a "Raspberry Pi" into their BFL miner. My experience tells me this is not a real solution.

What operating system and mining software should they install on it? What optimizations should they perform on the OS build to ensure the bitcoin miner is not interrupted by other processes?

If BFL isn't going to provide a mining controller, then most users will end up using a real computer, not a "Raspberry Pi." You should be more realistic.


In addition to that - serious miners are already running a beefy system as a host to control mining equipment... the simple fact is... unless you want to be physically close to your mining location, you need a machine to remote into and control your power switches and other gear. If you're running a big enough farm - it's also going to be your bitcoind node.

Slapping openwrt on a device doesn't make it easier to administer remotely - why expose your avalon (or any mining machine) to potential intrusion via forwarded ports? It makes no sense, you'll have a secure vpn link to a machine on that local subnet.




jgarzik (OP)
Legendary
*
Offline Offline

Activity: 1596
Merit: 1091


View Profile
February 02, 2013, 07:50:18 PM
 #93

Are you using getwork or stratum on p2pool? Did you try both? Any difference between the two?

Stratum, with "+8" (difficulty 8.0).  Stratum seems to work better in general, with Avalon's modified cgminer 2.10.4.


Jeff Garzik, Bloq CEO, former bitcoin core dev team; opinions are my own.
Visit bloq.com / metronome.io
Donations / tip jar: 1BrufViLKnSWtuWGkryPsKsxonV2NQ7Tcj
loshia
Legendary
*
Offline Offline

Activity: 1610
Merit: 1000


View Profile
February 02, 2013, 08:42:13 PM
 #94

Jeff,

You can take a look here:

https://github.com/ckolivas/cgminer/commit/7e2b1e80aafc431f8572ad57bb012f2dfe7f7c97


I am testing this on my openwrt and i am having zero memory leaks for almost 10 hours. I guess when you mine with stratum, Avalon memory will be exhausted very soon, which can cause system reboot.

Are you having cgminer patches (source) related to avalon?
Someone from Avalon team has to send you recompiled/updated cgminer package probably.


Please help the Led Boy aka Bicknellski to make us a nice Christmas led tree and pay WASP membership fee here:
https://bitcointalk.org/index.php?topic=643999.msg7191563#msg7191563
And remember Bicknellski is not collecting money from community;D
greyhawk
Hero Member
*****
Offline Offline

Activity: 938
Merit: 1009


View Profile
February 02, 2013, 08:55:59 PM
 #95

Paging Jgarzik.  Party of one.  Jgarzik, party of 1!  Your seat at the ASIC table is ready.  JGarzik, party of 1!

I think this is one dinner you'll be happy to eat alone. Smiley

At least he gets dinner. Unlike at "Chez Papillon" where they're seating guests while the kitchen is still waiting for ingredients.
Bogart
Legendary
*
Offline Offline

Activity: 966
Merit: 1000


View Profile
February 02, 2013, 09:43:22 PM
 #96

DC Power supplies are marked by the DC wattage provided, not by the AC wattage consumed.  There is nothing at all strange about a device marked "450W" pulling 600+ watts from the wall.  

This is a 91% efficiency power supply. So even if that was "normal" 620W AC is ~565W DC.  565W DC > 400W DC.

Also remember the 400W number comes from the product listing for the entire unit.  It is "strange" for a company to have all the specs related to the entire unit except the power which just happens to be just the chip (not even the total DC wattage).  Now for the module only listing putting a wattage of ~180W DC would be realistic. 

Still even if you take 620W AC remove the PSU inefficiency getting 565W DC then drop out 25W for the host and 3 5W fans = 540W not 400W.  A single module by itself has to be pulling ~180W ea (3x180 = 540) not ~133W ea (3x133W=400W).   So the 400W number is dishonest, there is no way or reason to spin that.  The sad thing is first out the door @ 180W per module and 620W per unit is still impressive.  It is dishonesty that isn't even needed.

There is a second stage of DC-DC conversion where the 12V output from the ATX power supply is further reduced, adding another layer of loss.

"All safe deposit boxes in banks or financial institutions have been sealed... and may only be opened in the presence of an agent of the I.R.S." - President F.D. Roosevelt, 1933
Askit2
Hero Member
*****
Offline Offline

Activity: 981
Merit: 500


DIV - Your "Virtual Life" Secured and Decentralize


View Profile
February 02, 2013, 10:31:36 PM
 #97

I guess the point people menioned about the power can be pointed to as this. When BFL missed estimates they didn't say they used a laptop PS that was 80% efficient and their box had fans so those take power, plus they needed additional fans. They said it uses 80 Watts (at the wall) and people checked at the wall to verify power claims.

Why didn't Avalon at least list their system by their actual draw? because 400 sounds wayyyyy better then 600. It is easy enough to plug in and verify. This is the cost in power the customer pays for is the one that would matter not that the modules apparently take 2/3rds of the power.

Now do I care if the unit is stand alone. No. Do I care about power draw. Yes. Why because I can run linux on a Rpi at ~5W and some BFL gear for less watts (if the numbers are off by 800%) and have less cooling required through the hot part of the year.

To me the fact that the main face of Icarus and Lancelot trolled BFL over being wrong about power use makes me wonder why such a great and honest person wouldn't publish an accurate power consumption figure for a unit that actually plugs into the wall. It isn't difficult. It is expected that the power use listed for a device with 1 power plug will include the power burned up by the Power supply, the hubs, the Open WRT, the fans and anything else in the case that would be powered to use the device.

Maybe I missed the part where other companies forgot to take into concideration their Power supplies or fans for a boxed unit. Maybe the Avalon team was inexperienced with this as the previous offerings where not as complete of a unit. I don't know.

          ▄▄
        ▄█▀▀█▄
      ▄█▀ ▄▄ ▀█▄
      ▀ ▄████▄ ▀
   ▄▀ ▄ ▀████▀ ▄ ▀▄
 ▄▀ ▄███▄ ▀▀ ▄███▄ ▀▄
█  ███████  ███████  █
 ▀▄ ▀███▀ ▄▄ ▀███▀ ▄▀

   ▀▄ ▀ ▄████▄ ▀ ▄▀
      ▄ ▀████▀ ▄
      ▀█▄ ▀▀ ▄█▀
        ▀█▄▄█▀
          ▀▀
███████████████████████████████████████████████████████████████████
██████▀▀▀▀▀▀▀▀▀▀▀██████████▀▀▀▀▀████▀▀▀▀▀█████▀▀▀▀█████▀▀▀▀▀███████
██████            ▀████████     ████     █████    █████     ███████
██████     ▄▄▄▄▄    ▀██████     █████    ████      ████    ████████
██████     ██████▄    █████     █████    ▀██▀  ▄▄  ▀██▀    ████████
██████     ███████    █████     ██████    ██   ██   ██    █████████
██████     ███████    █████     ██████    ██   ██   ██    █████████
██████     ███████    █████     ██████     █   ██   █     █████████
██████     █████▀    ██████     ███████       ████       ██████████
██████     ▀▀▀▀▀    ▄██████     ████████     ██████     ███████████
██████            ▄████████     ████████     ██████     ███████████
██████▄▄▄▄▄▄▄▄▄▄▄██████████▄▄▄▄▄█████████▄▄▄▄██████▄▄▄▄████████████
███████████████████████████████████████████████████████████████████
.DIWtoken.com.
▄██████████████████▄
███       ▀███████
███       █████████
███       █████████
███       █████████
███              ██
███   ▄▄▄▄▄▄▄▄   ███
███   ▄▄▄▄▄▄▄▄   ███
███              ███
███▄▄▄▄▄▄▄▄▄▄▄▄▄▄███
██████████████████▀

▄██████████████████▄
███████████▀ ███████
█████████▀   ███████
███████▀     ██▀ ███
███ ▀▀       █▄▄████
███          █▀▀▀▀██
███ ▄▄       ███████
██████▄     █▄ ▀███
█████████▄   ███▄███
███████████▄ ███████
▀██████████████████▀

▄██████████████████▄
████████████████████
███████████████▀▀ ██
█████████▀▀     ███
████▀▀     ▄█▀   ███
███▄    ▄██      ███
█████████▀      ▄██
█████████▄     ████
█████████████▄ ▄████
████████████████████
▀██████████████████▀
......SECURITY DECENTRALIZED...
kano
Legendary
*
Offline Offline

Activity: 4494
Merit: 1808


Linux since 1997 RedHat 4


View Profile
February 02, 2013, 11:22:54 PM
 #98

Jeff,

You can take a look here:

https://github.com/ckolivas/cgminer/commit/7e2b1e80aafc431f8572ad57bb012f2dfe7f7c97


I am testing this on my openwrt and i am having zero memory leaks for almost 10 hours. I guess when you mine with stratum, Avalon memory will be exhausted very soon, which can cause system reboot.

Are you having cgminer patches (source) related to avalon?
Someone from Avalon team has to send you recompiled/updated cgminer package probably.

No one (but Avalon) has any source code yet ...

Pool: https://kano.is - low 0.5% fee PPLNS 3 Days - Most reliable Solo with ONLY 0.5% fee   Bitcointalk thread: Forum
Discord support invite at https://kano.is/ Majority developer of the ckpool code - k for kano
The ONLY active original developer of cgminer. Original master git: https://github.com/kanoi/cgminer
pazor
Legendary
*
Offline Offline

Activity: 966
Merit: 1000



View Profile
February 03, 2013, 12:27:55 AM
 #99

jeff, you should verify your measurement (with load and without load) without ups asap.

treuhand-Dienst gewünscht? - frag per PM an
BTC 174X17nR7vEQBQo4GXKRGMGaTmB49Gf1yT
vapourminer
Legendary
*
Offline Offline

Activity: 4326
Merit: 3519


what is this "brake pedal" you speak of?


View Profile
February 03, 2013, 12:38:33 AM
 #100

Have you tried measuring the power without the UPS square wave being invovled? I.o.w. plug it straight into the wall (or perhaps with a transient protector only)?

No, though several have requested that.

It would be good to eliminate the UPS as a factor, agreed.

I believe you mentioned you had it plugged into an APC backUPS 550. if its like my APC backUPS ES-550, it only trickle charges the battery when power is on. normally your avalon will be running on wall power with only surge suppression.

also if your avalon draws 600+ watts it wont power your avalon when the powers out. the ES-550 only delivers 330 watts when in battery backup mode.

this is if its plugged into the battery backup outlets of course. there are surge suppression only outlets on the ES-550 too.
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 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 ... 221 »
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!