For the record, I believe that neither DeepBit nor Inaba are shills for BFL
And I can't imagine what deluded frame of mind is causing other people to think otherwise. It's insane.
Topic: 1GH/s, 20w, $700 (was $500) — Butterflylabs, is it for real? (Part 2) - page 75. (Read 146909 times)
December 08, 2011, 12:44:06 PM
December 08, 2011, 12:40:45 PM
For the record, I believe that neither DeepBit nor Inaba are shills for BFL
And I can't imagine what deluded frame of mind is causing other people to think otherwise. It's insane. December 08, 2011, 12:38:05 PM
For the record, I believe that neither DeepBit nor Inaba are shills for BFL
December 08, 2011, 12:24:46 PM
Long polling don't affect that much, especially in 10 minutes run. Propably chips are overheating and make errors.
Usually mining software detects such errors and filters them out. I didn't saw any share submission errors there.Long Polling reduces stales by 0.5%-1.5%, but there may be also some another reason.
I think that the take away is the 946 MH/s avg for 10 minutes, that is unless deepbit is in on the scam with Inaba.
I can acknowledge that this value is real (enhanced with local luck).Stales percentage is considerably reduced now.
Of course I'm not, why would I help my potential competitors ? :)
December 08, 2011, 12:17:41 PM
Since they generated 2388 shares, the stales are probably just caused from the debugging/testing process BFL is going through, since we really don't know how long it took to generate those shares.
I think that the take away is the 946 MH/s avg for 10 minutes, that is, unless DeepBit is in on the scam with Inaba.
I wouldn't mind seeing the performance of multiple units working together though. That would be cool.
I think that the take away is the 946 MH/s avg for 10 minutes, that is, unless DeepBit is in on the scam with Inaba.
I wouldn't mind seeing the performance of multiple units working together though. That would be cool.
December 08, 2011, 10:37:54 AM
Long polling don't affect that much, especially in 10 minutes run. Propably chips are overheating and make errors.
December 08, 2011, 07:39:42 AM
That's hot, thanks for the update.
Btw, are us dumb Americans the only ones that still use the month/day/year format? It still throws me off for a second to see the the day/month/year even though it is much more common every where else it seems.
Cheers
Btw, are us dumb Americans the only ones that still use the month/day/year format? It still throws me off for a second to see the the day/month/year even though it is much more common every where else it seems.
Cheers
Americans also do not use the metric system. Everyone else does but Burma and Ivory Coast I think:)
Computer scientists should use ISO 8601. Forget your local custom.
strftime("%Y-%m-%d %H:%M:%S")oh, crap, I replied to another one of these threads...
December 08, 2011, 07:37:03 AM
@BFL Thanks for posting an update and being open to sharing what's going on.
I look forward to receiving my Bitforce Single (hopefully for Christmas)!
I look forward to receiving my Bitforce Single (hopefully for Christmas)!
December 08, 2011, 07:15:32 AM
Nice results but still far off from the promised 1050 MHash/s performance. Are we going to be seeing those figures at the final launch ? 100 MHash/s does make a difference.
December 08, 2011, 07:00:01 AM
By the way, the 3.98% stales looks very high... I'm getting 0.11% to 0.17% and I'm not even on a very fast internet link... where is the bottle neck? The linkup between the FPGA and the intertubes?
May be they don't support Long Polling (one of possible reasons). December 08, 2011, 06:51:03 AM
May be they just don't have spare unit to leave mining because they are working on it's tuning.
As I see, it was up for more than 10 minutes, but not during the entire round, so the right column would show lower speed.
10 minutes is too short to get actual data because luck variation will be usually more than ±30%
(Of course I can't know for sure if those shares are coming from FPGA or GPU because useragent string can be easily forged)
As I see, it was up for more than 10 minutes, but not during the entire round, so the right column would show lower speed.
10 minutes is too short to get actual data because luck variation will be usually more than ±30%
(Of course I can't know for sure if those shares are coming from FPGA or GPU because useragent string can be easily forged)
Not very confidence building if they only have 2 boards (one handed out to Inaba and one in the lab for development)... surely they must have multiple units laying around for endurance testing / burn in / ...
Of course we can't know if it's from the actual device, but let's give them the benefit of the doubt ;-)
By the way, the 3.98% stales looks very high... I'm getting 0.11% to 0.17% and I'm not even on a very fast internet link... where is the bottle neck? The linkup between the FPGA and the intertubes?
December 08, 2011, 06:18:30 AM
Hi guys, I thought you might enjoy a screen shot update of the BitForce unit at work.
why don't you let a unit work for 1 or 2 days on deepbit and post here the full screen shot of https://deepbit.net/account with the right column of stats and the https://deepbit.net/stats page, that way we can see it's been up and running and how it's performing.Right now one can only guess your unit has been up for 10 minutes at 946 MHz (with some earlier activity to generate the 0.085 BTC) as conveniently you don't include the 24h stats column (which is fuel for speculation of course ;-)
As I see, it was up for more than 10 minutes, but not during the entire round, so the right column would show lower speed.
10 minutes is too short to get actual data because luck variation will be usually more than ±30%
(Of course I can't know for sure if those shares are coming from FPGA or GPU because useragent string can be easily forged)
December 08, 2011, 05:55:29 AM
Hi guys, I thought you might enjoy a screen shot update of the BitForce unit at work.
why don't you let a unit work for 1 or 2 days on deepbit and post here the full screen shot of https://deepbit.net/account with the right column of stats and the https://deepbit.net/stats page, that way we can see it's been up and running and how it's performing.
Right now one can only guess your unit has been up for 10 minutes at 946 MHz (with some earlier activity to generate the 0.085 BTC) as conveniently you don't include the 24h stats column (which is fuel for speculation of course ;-)
December 08, 2011, 02:37:01 AM
That's hot, thanks for the update.
Btw, are us dumb Americans the only ones that still use the month/day/year format? It still throws me off for a second to see the the day/month/year even though it is much more common every where else it seems.
Cheers
Btw, are us dumb Americans the only ones that still use the month/day/year format? It still throws me off for a second to see the the day/month/year even though it is much more common every where else it seems.
Cheers
Americans also do not use the metric system. Everyone else does but Burma and Ivory Coast I think:)
Aye, for every day stuff. Its fuggin annoying. As soon as we do any kind of surveying, molecular engeering work, etc etc, there is no reason not to use metric. But yet we insist on having fuggin 'standard' rulers handed out to all our school children. Dirves me nuts. Which also come in 'American' and Metric here. >.<
December 08, 2011, 02:07:32 AM
That's hot, thanks for the update.
Btw, are us dumb Americans the only ones that still use the month/day/year format? It still throws me off for a second to see the the day/month/year even though it is much more common every where else it seems.
Cheers
Btw, are us dumb Americans the only ones that still use the month/day/year format? It still throws me off for a second to see the the day/month/year even though it is much more common every where else it seems.
Cheers
December 08, 2011, 12:38:04 AM
Hi guys, I thought you might enjoy a screen shot update of the BitForce unit at work. At this current stage of tweaking, we're getting pool reported speeds from 850 to 970 mh/s. We'll be making additional adjustments prior to demo this weekend with Inaba so it's reasonable to expect these figures will change.
You can change the averaging window of pool's luckmeter at "Advanced" settings page. Setting it to 40 or even 60 minutes can give much better result than 10 minutes. December 08, 2011, 12:03:48 AM
For the few of us that aren't quite as concerned with power usage, will there be a way that we can flash alternate firmwares in order to increase or reduce speed?
Econo-mode and "Warp-6" for when bitcoin is at $30.Also, these are standard 80-mm fans correct? That is, if they fail, they can be replaced with a standard 80mm case fan?
Nice screen shot BTW!
December 07, 2011, 11:52:04 PM
Hi guys, I thought you might enjoy a screen shot update of the BitForce unit at work. At this current stage of tweaking, we're getting pool reported speeds from 850 to 970 mh/s. We'll be making additional adjustments prior to demo this weekend with Inaba so it's reasonable to expect these figures will change.
Our development has been extended due to increased power demands. The speed of the unit had to be reduced to compensate. Our current focus is finding the most economical balance between the two prior to finalization & delivery production.
Regards,
BFL
Our development has been extended due to increased power demands. The speed of the unit had to be reduced to compensate. Our current focus is finding the most economical balance between the two prior to finalization & delivery production.
Regards,
BFL
For the few of us that aren't quite as concerned with power usage, will there be a way that we can flash alternate firmwares in order to increase or reduce speed?
December 07, 2011, 11:45:24 PM
Hi guys, I thought you might enjoy a screen shot update of the BitForce unit at work. At this current stage of tweaking, we're getting pool reported speeds from 850 to 970 mh/s. We'll be making additional adjustments prior to demo this weekend with Inaba so it's reasonable to expect these figures will change.
Our development has been extended due to increased power demands. The speed of the unit had to be reduced to compensate. Our current focus is finding the most economical balance between the two prior to finalization & delivery production.
Regards,
BFL
Our development has been extended due to increased power demands. The speed of the unit had to be reduced to compensate. Our current focus is finding the most economical balance between the two prior to finalization & delivery production.
Regards,
BFL
December 07, 2011, 09:13:47 PM
Just a couple of days to go till the weekend.
BTW, why do all of your demos happen then and not during the working week ? Is this a part time venture or ... ?
BTW, why do all of your demos happen then and not during the working week ? Is this a part time venture or ... ?
Jump to: