rpi B for cgminer

[kano]

Finally switched mine off today to replace the SD card (attempt 3 - this time Class 10, 8GB)
A bit over 23 days non-stop mining on an MMQ at 820MH/s with cmginer2.11.3a
(was actually mining for 26 days but I restarted cgminer when I updated it to 2.11.3a)

[1715289826]

1715289826

[1715289826]

1715289826

[1715289826]

1715289826

[1715289826]

1715289826

[BitMinerN8]

Finally switched mine off today to replace the SD card (attempt 3 - this time Class 10, 8GB)
A bit over 23 days non-stop mining on an MMQ at 820MH/s with cmginer2.11.3a
(was actually mining for 26 days but I restarted cgminer when I updated it to 2.11.3a)

That is good to hear, I really hope they can handle a Single SC.

[kano]

Jala rpi screen shot ... mining 256diff - using approx 3% CPU
Click to see:

[BitMinerN8]

Jala rpi screen shot ... mining 256diff - using approx 3% CPU
Click to see:

Yeah!

[astutiumRob]

I have actually even had bitcoind running running on a PI but I gave up long before the blockchain synced.  I suspect if I let it go I would have ended up with a toasted SD card.

I tried a while back, it to ~275 days to-go worth of data, then seems to take longer than a day to add the blockchain for that day - so was neverexpected to finish - left it a while and my rPi died

Another good reason for having "balance as at" version of the blockchain

[crazyates]

Jala rpi screen shot ... mining 256diff - using approx 3% CPU
Click to see:

How much do you think the rPi could handle? If a SC Single is ~10x as much hashpower, are we gonna see ~10x as much CPU usage?

[kano]

Jala rpi screen shot ... mining 256diff - using approx 3% CPU
Click to see:

How much do you think the rPi could handle? If a SC Single is ~10x as much hashpower, are we gonna see ~10x as much CPU usage?

No the CPU usage isn't directly linear.

However, USB seems to be the larger factor in CPU usage on the rpi.
My driver also doesn't linearly increase the polling for work results - with the Jala, it polls more often than needed since latency is still an issue when it takes almost 1s to complete a nonce range.

The 50or60GH/s device will need polling for results about twice as often as I do with the Jala
Sending work to it of course is 10x as much and processing replies is also 10x as much.
Anyway, we'll see once I get one ... not sure when yet.