Avalon ASIC users thread

[jgarzik]

Uptime: 1d 13h 49m 29s

solo mining w/ eloipool + high diff

Good luck. Are you going for a block no matter how long it takes?

Thanks for the luck, it apparently helped!    The following is, as far as I know, the first block found by a publicly released ASIC miner:
http://blockexplorer.com/block/00000000000001528a3fa72b86032459e1fb6ab38720e19a26e3a1f4a64e461a

So far all known ASIC vendors have been highly ethical and not used their mining power on mainnet, which could mean that block #220386 is the first ASIC block found.

Will it mine for as long as it takes?  I imagine it will make more sense to mine on a pool once other ASIC miners start hitting mainnet.  The average block production time, for the moment, should be around 2 days, 4 hours.

[1714129943]

1714129943

[1714129943]

1714129943

[1714129943]

1714129943

[jgarzik]

Uptime: 2d 0h 40m 23s

[jgarzik]

I'm sure Jeff can confirm these numbers on his device as well.

My review included raw numbers -- search for "cgminer status"

Here is an updated cgminer status snapshot:  http://pastebin.com/8mKjmTSn

Readers may search for "temp" to find what I presume to be temperature sensor readings.  Fan speed is also observed via "fan" data.

[420]

Why's there 12 pages and there's only one damn ASIC owner besides the federation

If I'm right; give me reasons

[coinnewb]

Uptime: 1d 13h 49m 29s

solo mining w/ eloipool + high diff

Good luck. Are you going for a block no matter how long it takes?

Thanks for the luck, it apparently helped!    The following is, as far as I know, the first block found by a publicly released ASIC miner:
http://blockexplorer.com/block/00000000000001528a3fa72b86032459e1fb6ab38720e19a26e3a1f4a64e461a

So far all known ASIC vendors have been highly ethical and not used their mining power on mainnet, which could mean that block #220386 is the first ASIC block found.

Will it mine for as long as it takes?  I imagine it will make more sense to mine on a pool once other ASIC miners start hitting mainnet.  The average block production time, for the moment, should be around 2 days, 4 hours.

Very nice. Congrats!
Do you mind sharing how you set up eloipool for ASIC mining?  I want to learn how to do so in case I want/need to solo mine.  Thanks!

[kano]

I'm sure Jeff can confirm these numbers on his device as well.

My review included raw numbers -- search for "cgminer status"

Here is an updated cgminer status snapshot:  http://pastebin.com/8mKjmTSn

Readers may search for "temp" to find what I presume to be temperature sensor readings.  Fan speed is also observed via "fan" data.

Heh funny I didn't think of it before ...
They've named it a PGA (PGA0) I chose PGA as the 3 letter shortening of FPGA
I guess using PGA avoided having to create a new ASIC type in the API
(which I'll decide on shortly)

[jgarzik]

They've named it a PGA (PGA0) I chose PGA as the 3 letter shortening of FPGA
I guess using PGA avoided having to create a new ASIC type in the API

Or just google for "pin grid array"

[kano]

They've named it a PGA (PGA0) I chose PGA as the 3 letter shortening of FPGA
I guess using PGA avoided having to create a new ASIC type in the API

Or just google for "pin grid array"

I came up with the name, idiot.
It says in the README ... ...

[Icoin]

The air leaves the module with around 50°C. On the same side is the air intake for the PSU, means the PSU gets 50°C air for cooling. Even if it should be just 30°C its not realy suitable. Compare it with the dark blue area outside the device, thats the right temperature.

I guess to turn the PSU 180°  in a way the air intake would point to the outside of the case would lower the energy consumption.
I allso dont see a heat barrier (spacing) between the elements and the casing, the heat is distributed trough the thick groundplate on the various components.


The heat distribution trough the massive ground plate aswell that the PSU beeing at a higher temperature on its cables and on the air intake is clearly visible. Even the grill on the PSU shows a temperature around 40°C

One of the temperature monitors consistently reads close to 50, and "temp_max" is often 100-125, so it is possible or even likely that temperature is playing a factor in these restarts.

Jeff, this simple modification will save you power costs

[organofcorti]

The air leaves the module with around 50°C. On the same side is the air intake for the PSU, means the PSU gets 50°C air for cooling. Even if it should be just 30°C its not realy suitable. Compare it with the dark blue area outside the device, thats the right temperature.
......

So yours arrived?

[MrTeal]

Inaba: Hey, lets do some scientific and procedural testing so we can determine an outcome given a set of constants (hardware) and variables (temperature).
Dalkore: That's stupid. It's either going to be situation A or situation B. I don't know which, and I have no way of finding out unless we follow Inaba's suggestion, but it's still stupid.

Really? 

Maybe not stupid, but probably not the best idea. It's more along that lines of "Hey guy who paid for some mining hardware, we want you to stop earning a couple hundred bucks a day from your Avalon so you can spend time doing testing for us."

Right now if intake air temps aren't 85F, why take the unit offline to test that? Worry about it in a couple months when it's getting warmer and there's a couple hundred TH/s on the network, but for now milk that thing for all it's worth. If anyone should be doing this testing, it's the Avalon team.

The air leaves the module with around 50°C. On the same side is the air intake for the PSU, means the PSU gets 50°C air for cooling. Even if it should be just 30°C its not realy suitable. Compare it with the dark blue area outside the device, thats the right temperature.

Jeff, this simple modification will save you power costs

No, the heatsinks are 50C assuming that the camera was properly calibrated to view flat metal surfaces. The air temps are going to be less than the heatsink temperature, looking at the picture it seems around 30C. As you say, 20C would be better but 30C is pretty standard for what a PSU can see drawing air from the inside of a case so it's not like it won't handle it.


Jeff, if you're wondering if it's heating related, grab a couple high flow 120mm fans and put them in the missing holes running full out. You could also make an air guide in the empty slot to force the air from the fan with only module into the heatsink. Both things could be done without having to take the unit offline to tinker with, and they should drop the temps substantially. See if that helps with your restarting issue.

[jgarzik]

Jeff, if you're wondering if it's heating related,

I'm not.

Seems pretty clear heat is not a problem here.

[jgarzik]

Come on Bitcoin Foundation, you gotta give up more details than that

[Yankee (BitInstant)]

Come on Bitcoin Foundation, you gotta give up more details than that

I honestly don't know the answer...Yifu will be here today, I dont have the logins.

[kaerf]

Come on Bitcoin Foundation, you gotta give up more details than that

I honestly don't know the answer...Yifu will be here today, I dont have the logins.

Can you kindly do us a favor and ask him what the shipping status of batch #1 is and what's happening with batch #2 ?

[Dalkore]

Come on Bitcoin Foundation, you gotta give up more details than that

I honestly don't know the answer...Yifu will be here today, I dont have the logins.

Can you kindly do us a favor and ask him what the shipping status of batch #1 is and what's happening with batch #2 ?

This would be great, it seems like he was updating us a lot before pre-sale batch #2 and now its a blackhole in some regards.   Good customer service would be at least an update once a week. 

[AlexWaters]

The ASIC definitely sheds heat well. I tried working in the office where it's kept the other day (which has a couch.) And almost fell asleep from the warmth and pleasant hum of Bitcoin minting. 

Also, this morning I somehow spilled OJ on myself and my laptop keyboard. I pulled the keyboard out, washed it in the sink, and used the ASIC's exhaust to dry it. 10 minutes later my laptop is as good as new.

Cheers for thermal efficiency!

[SolarSilver]

The ASIC definitely sheds heat well. I tried working in the office where it's kept the other day (which has a couch.) And almost fell asleep from the warmth and pleasant hum of Bitcoin minting. 

Can you quantify how much noise it makes? I guess because these are big fans, the noise is not as irritating as little GPU fans at full blast?

I know for sure I would not want to share an office with 3 GPU cards at 100% fan speed

[420]

Whats the total count of ASIC's in customer hands?? 3 or 4?

[niko]

Whats the total count of ASIC's in customer hands?? 3 or 4?

Two units are confirmed in this thread (jgarzik and the BF), and network hashrate indicates no jncrease beyond the usual variance.