What's the power usage operating at this speed?
The chip is running at 0.84V and 2.5A, with a clock of 230MHz.
Topic: [ANN] Bitfury is looking for alpha-testers of first chips! FREE MONEY HERE! - page 39. (Read 176728 times)
July 02, 2013, 08:21:27 AM
The chip is running at 0.84V and 2.5A, with a clock of 230MHz.
July 02, 2013, 08:02:54 AM
Cool! But how can it estimate hash rate without any accepted shares?
Few minutes later:
intron_2 2,149.59 MH/s 174 (100.00%) 0 / 1 / 0
July 02, 2013, 06:52:14 AM
Hi All,
I just received some BitFury ASIC chips today. Big thanks to BitFury for giving myself and the BitCentury team this opportunity. We are currently waiting for our initial prototype PCBs to be manufactured. I will keep you up to date with progress of the testing, but in the meantime check the following link for the pics: http://bitcentury.io/blog/bitfury-asics-have-arrived.
Regards,
Nigel / BitCentury
I just received some BitFury ASIC chips today. Big thanks to BitFury for giving myself and the BitCentury team this opportunity. We are currently waiting for our initial prototype PCBs to be manufactured. I will keep you up to date with progress of the testing, but in the meantime check the following link for the pics: http://bitcentury.io/blog/bitfury-asics-have-arrived.
Regards,
Nigel / BitCentury
July 02, 2013, 03:14:25 AM
Feeding external clock with its own dedicated power supply will improove these parameters a lot and will help for more stable work at higher clock speeds if chip allows them. I have prepared one variable for selecting the best frequency and then will use fixed for maximised performance.
And do you feed the clock to multiple chips then?
With clock buffers or fanning tree or matched lines?
intron
I am talking about checking performance of one chip with external clock vs integrated, if it is cost effective depending on the hash rate increase buffers for each chip will be nice.
July 02, 2013, 03:03:42 AM
The disadvantage with an external clock is that there's no practical way of tuning the clock frequency to individual ASIC performance. Given variations in process and circumstances, it may be better to optimize each ASIC's internal clock, rather than using a single external clock for all of them.
I don't know if a respin is planned, but maybe a separate VDD for the oscillator circuit would help improve stability.
I don't know if a respin is planned, but maybe a separate VDD for the oscillator circuit would help improve stability.
July 01, 2013, 08:37:19 AM
Feeding external clock with its own dedicated power supply will improove these parameters a lot and will help for more stable work at higher clock speeds if chip allows them. I have prepared one variable for selecting the best frequency and then will use fixed for maximised performance.
And do you feed the clock to multiple chips then?
With clock buffers or fanning tree or matched lines?
intron
July 01, 2013, 07:55:51 AM
Feeding external clock with its own dedicated power supply will improove these parameters a lot and will help for more stable work at higher clock speeds if chip allows them. I have prepared one variable for selecting the best frequency and then will use fixed for maximised performance.
July 01, 2013, 02:12:47 AM
We haven't tried external oscillator. I don't have a generator that can do 200MHz+ outputs. The internal oscillator is fast enough, but it does look jittery on the scope, so internal setup/hold timing may be improved with stable external oscillator, and a higher clock may be used.
July 01, 2013, 02:07:00 AM
intron have you tried to increase hash rate with external oscillator or higher voltage or you do not have proper cooling for now.
June 30, 2013, 01:50:11 PM
Just send him some chips as well pls as he made boards (ultrix, sorry, as I said I don't have at hands most). Maybe Silverpike could share some with you or Dave (in USA) ?
Its not worry, as it actually works out a bit better given my current work schedule. I should have more free time to tinker in a week or two.
June 30, 2013, 01:35:44 PM
Just send him some chips as well pls as he made boards (ultrix, sorry, as I said I don't have at hands most). Maybe Silverpike could share some with you or Dave (in USA) ?
I have only enough chips to make a single prototype H-board to verify the function and build quality of the US-side board manufacturer. I'll post here if I end up with a few spares, but it would be less than 5 at most.
June 30, 2013, 09:01:53 AM
Intron you were using the diodes to protect INSCK INMOSI inputs in the first board.
But bitfury said that there are diodes inside the chip. Are you still using them on new boards?
But bitfury said that there are diodes inside the chip. Are you still using them on new boards?
No. I was alarmed a bit by some of bitfury's comments about
the chip being highly sensitive and could be 'fried' in a whimp.
Later, when the boards were ready I read the ASIC had
protection inside the chip. So on the new boards the
zeners are omitted.
intron
June 30, 2013, 08:48:36 AM
Intron you were using the diodes to protect INSCK INMOSI inputs in the first board.
But bitfury said that there are diodes inside the chip. Are you still using them on new boards?
But bitfury said that there are diodes inside the chip. Are you still using them on new boards?
June 30, 2013, 06:50:37 AM
Just send him some chips as well pls as he made boards (ultrix, sorry, as I said I don't have at hands most). Maybe Silverpike could share some with you or Dave (in USA) ?
June 30, 2013, 06:18:54 AM
The c-scape firmware has its own TCP/IP stack, JSON parser and mining software,
so it's lightweight and compact, and can run on single chip $5 ARMs in less than
32 KB flash memory. The only code from external sources is the SHA-256 implementation
(about 150 lines), needed to prepare the midstate, and to verify the bitfury results,
and of course bitfury's example SPI code to initialize the chip and get the results.
Currently it only supports the getwork protocol, so it depends on an external
stratum proxy for best results.
c-scape
Which ARM chip? I was thinking about doing something similar with an LPC1768. Has USB OTG, Ethernet, 512kb flash memory with in app programming, and a bunch of other useful features. You can get them in prototyping quantities for ~$6-8 and in lots of 100 for $3-4 (depending on how many chinese distributors you want having your email and phone number).
I've got a small stratum implementation tested against stratum proxy, eloipool, etc. I'll upload somewhere when I get home from traveling.
We use the NXP LPC1758, quite similar.
And looking forward to see your code:)
intron
June 30, 2013, 05:52:41 AM
The c-scape firmware has its own TCP/IP stack, JSON parser and mining software,
so it's lightweight and compact, and can run on single chip $5 ARMs in less than
32 KB flash memory. The only code from external sources is the SHA-256 implementation
(about 150 lines), needed to prepare the midstate, and to verify the bitfury results,
and of course bitfury's example SPI code to initialize the chip and get the results.
Currently it only supports the getwork protocol, so it depends on an external
stratum proxy for best results.
c-scape
Which ARM chip? I was thinking about doing something similar with an LPC1768. Has USB OTG, Ethernet, 512kb flash memory with in app programming, and a bunch of other useful features. You can get them in prototyping quantities for ~$6-8 and in lots of 100 for $3-4 (depending on how many chinese distributors you want having your email and phone number).
I've got a small stratum implementation tested against stratum proxy, eloipool, etc. I'll upload somewhere when I get home from traveling.
June 30, 2013, 05:41:31 AM
Thank you very much for your reply. Felipeo and I were using the term "standalone" in the old software engineering sense: not using the OS services like network stack, dynamic linking and memory management.
A lot of Bitcoin software is hopelessly entwined with humounguos OS-dependent components like Python interpreter or OpenSSL library.
I was just trying to confirm my guess that your miner software isn't dragging that baggage.
Thanks again.
A lot of Bitcoin software is hopelessly entwined with humounguos OS-dependent components like Python interpreter or OpenSSL library.
I was just trying to confirm my guess that your miner software isn't dragging that baggage.
Thanks again.
The c-scape firmware has its own TCP/IP stack, JSON parser and mining software,
so it's lightweight and compact, and can run on single chip $5 ARMs in less than
32 KB flash memory. The only code from external sources is the SHA-256 implementation
(about 150 lines), needed to prepare the midstate, and to verify the bitfury results,
and of course bitfury's example SPI code to initialize the chip and get the results.
Currently it only supports the getwork protocol, so it depends on an external
stratum proxy for best results.
c-scape
June 29, 2013, 05:34:51 PM
Not totally, it still needs a PC running a stratum proxy.
When there is time and the poor little ARM can handle
the workload it would be nice to have truly it stand-alone.
Thank you very much for your reply. Felipeo and I were using the term "standalone" in the old software engineering sense: not using the OS services like network stack, dynamic linking and memory management.When there is time and the poor little ARM can handle
the workload it would be nice to have truly it stand-alone.
A lot of Bitcoin software is hopelessly entwined with humounguos OS-dependent components like Python interpreter or OpenSSL library.
I was just trying to confirm my guess that your miner software isn't dragging that baggage.
Thanks again.
Yhym standalone in that sens
Thank You for clarifying this for me 
June 29, 2013, 05:30:47 PM
Nice... your fun project seems to be the pre-eminent Bitfury DIY project on the net currently.
Hopefully chips will be available for sale soon and you can enjoy the fun of profit as well.
Hopefully chips will be available for sale soon and you can enjoy the fun of profit as well.
Sorry to disappoint but except for a handful of testers only Metabank and 100TH will have access to these chips as far as we know.
June 29, 2013, 05:28:20 PM
Not totally, it still needs a PC running a stratum proxy.
When there is time and the poor little ARM can handle
the workload it would be nice to have truly it stand-alone.
Thank you very much for your reply. Felipeo and I were using the term "standalone" in the old software engineering sense: not using the OS services like network stack, dynamic linking and memory management.When there is time and the poor little ARM can handle
the workload it would be nice to have truly it stand-alone.
A lot of Bitcoin software is hopelessly entwined with humounguos OS-dependent components like Python interpreter or OpenSSL library.
I was just trying to confirm my guess that your miner software isn't dragging that baggage.
Thanks again.
Jump to: