ZTEX USB-FPGA Modules 1.15x and 1.15y: 215 and 860 MH/s FPGA Boards - page 40.

ztex

donator

Activity: 367

Merit: 250

ZTEX FPGA Boards

Quote from: CA Coins on January 27, 2012, 04:57:22 PM

I got that problem too. I tested the new firmware on two boards and it will do that somewhat randomly. I tried it out for at least 30 minutes and it wouldn't stabilize. Got only about combined 180MH/s on two boards at deepbit. Went back to the d1 firmware and it ran fine at about 400MH/s.

What exactly happened to the frequency? Does the the d2 firmware clocks down, but the d1 firmware runs stable?

If yes, it is a clock stability problem (too much jitter). I saw that earlier and thought it would have been fixed because it does not appear anymore on my 1.15d FPGA board test cluster. I will change a few parameters and post a test release next week (because I cannot reproduce this problem here).

If that problem appears, just use the d1 firmware (from the new release). It is as fast.

CA Coins

donator

Activity: 305

Merit: 250

Quote

What I noticed is, the software kept switching frequencies like crazy, in 6 MHz increments, whereas the old software switched in 8 MHz increments and only very rarely.

I got that problem too. I tested the new firmware on two boards and it will do that somewhat randomly. I tried it out for at least 30 minutes and it wouldn't stabilize. Got only about combined 180MH/s on two boards at deepbit. Went back to the d1 firmware and it ran fine at about 400MH/s.

Inspector 2211

sr. member

Activity: 448

Merit: 250

Quote from: rupy on January 27, 2012, 03:20:07 PM

Don't forget the new firmware: -f ztex_ufm1_15d2.ihx

The Java software detects which firmware it is targeted at, and will complain if you try to run it against the wrong (== too old or too new) firmware.

rupy

hero member

Activity: 725

Merit: 503

Don't forget the new firmware: -f ztex_ufm1_15d2.ihx

ztex

donator

Activity: 367

Merit: 250

ZTEX FPGA Boards

Quote from: Inspector 2211 on January 27, 2012, 01:53:56 PM

I just looked at the screen output, the number on the far right side. And yes, I did wait a minute or two to see whether I get a non-zero hash rate.

This is the "submitted hash rate". Never, ever, ever use this value for performance evaluation if you run the board only for a few minutes. Use the frequency or the "hash rate" in the center of the line.

(Maybe I should put the actual "hash rate" to the end of the line and the "submitted hash rate" into the center. This will save al lot of support time ;-) )

Quote

What I noticed is, the software kept switching frequencies like crazy, in 6 MHz increments, whereas the old software switched in 8 MHz increments and only very rarely.

It switches more often at start-up but will stabilize after a while.

Inspector 2211

sr. member

Activity: 448

Merit: 250

Quote from: ztex on January 27, 2012, 01:50:02 PM

Did you evaluate the "submitted has rate" or "has rate"?

If you find no shares within the fist few minutes the "submitted hash rate" is zero because you submitted nothing.

The actual hash rate is the frequency minus errors, because the FPGA calculates on hash per clock. In the new version this value is called "hash rate" and can be found ind the logs/output between "maxErroRate" and "submitted new nonces".

I just looked at the screen output, the number on the far right side. And yes, I did wait a minute or two to see whether I get a non-zero hash rate.

What I noticed is, the software kept switching frequencies like crazy, in 6 MHz increments, whereas the old software switched in 8 MHz increments and only very rarely.

ztex

donator

Activity: 367

Merit: 250

ZTEX FPGA Boards

Quote from: Inspector 2211 on January 27, 2012, 01:48:09 PM

Will do, but right now I'm at work, so by the time I can send them, it'll be [early] Saturday morning in Germany.
I'm mining in a small office I rent in Silicon Valley, mostly with HD 5830 cards. (Yes, it gets "toasty" in there...)

Did you evaluate the "submitted hash rate" or "hash rate"?

If you find no shares within the fist few minutes the "submitted hash rate" is zero because you submitted nothing.

The actual hash rate is the frequency minus errors, because the FPGA calculates on hash per clock. In the new version this value is called "hash rate" and can be found ind the logs/output between "maxErroRate" and "submitted new nonces".

Inspector 2211

sr. member

Activity: 448

Merit: 250

Quote from: ztex on January 27, 2012, 01:19:12 PM

Quote from: Inspector 2211 on January 27, 2012, 01:13:06 PM

I tried it this morning, pointing it at deepbit.net (just like I was doing with the December 2011 version).

Result: 0.0 MH/s

After a few minutes of getting 0.0 MH/s, I reflashed the "old" (December 2011) firmware and was back to my usual 199 MH/s again.

Any ideas?

Send me the logs / output.

Will do, but right now I'm at work, so by the time I can send them, it'll be [early] Saturday morning in Germany.
I'm mining in a small office I rent in Silicon Valley, mostly with HD 5830 cards. (Yes, it gets "toasty" in there...)

ztex

donator

Activity: 367

Merit: 250

ZTEX FPGA Boards

Quote from: Inspector 2211 on January 27, 2012, 01:13:06 PM

I tried it this morning, pointing it at deepbit.net (just like I was doing with the December 2011 version).

Result: 0.0 MH/s

After a few minutes of getting 0.0 MH/s, I reflashed the "old" (December 2011) firmware and was back to my usual 199 MH/s again.

Any ideas?

Send me the logs / output.

Quote

Is the long-polling command line parameter now mandatory? I have never used it.

Long polling URL is determined automatically (you see it in the logs/output) or can be defined by "-lp"

Inspector 2211

sr. member

Activity: 448

Merit: 250

Quote from: rupy on January 26, 2012, 08:36:38 PM

New firmware out with long-polling. http://www.ztex.de/btcminer/

I tried it this morning, pointing it at deepbit.net (just like I was doing with the December 2011 version).

Result: 0.0 MH/s

After a few minutes of getting 0.0 MH/s, I reflashed the "old" (December 2011) firmware and was back to my usual 199 MH/s again.

Any ideas?

Is the long-polling command line parameter now mandatory? I have never used it.

rjk

sr. member

Activity: 448

Merit: 250

1ngldh

Quote from: DeathAndTaxes on January 27, 2012, 12:20:48 PM

Quote from: ztex on January 27, 2012, 11:35:52 AM

100 unit price is going to be reduced to about $250 for the next batch. So license production makes only sense for 250 units or more.

Hmm now that is interesting. 20 GH for $25K huh @ 1KW. Well a guy can dream. I got a 5.76KW circuit. Could put 100 GH in my garage. Of course then I would need a full time security guard.

Well that pretty much blows my dream of a 20ghash rig using video cards, drawing 7 kw.

And this isn't vaporware either..... hmmm.....

DeathAndTaxes

donator

Activity: 1218

Merit: 1079

Gerald Davis

Quote from: ztex on January 27, 2012, 11:35:52 AM

100 unit price is going to be reduced to about $250 for the next batch. So license production makes only sense for 250 units or more.

Hmm now that is interesting. 20 GH for $25K & 1KW.

Well a guy can dream. I got a 5.76KW circuit. Could put 100 GH/s in my garage. Of course then I would need a full time security guard so thats cuts into the ROI.

ztex

donator

Activity: 367

Merit: 250

ZTEX FPGA Boards

Quote from: iambaboon on January 27, 2012, 10:50:24 AM

I don't have experience with hardware. Does this mean that one could get all the instructions to build one 1.15x module at home ? In that case, a friend of mine with some experience in electronics/electronic part mounting would know enough ?

100 unit price is going to be reduced to about $250 for the next batch. So license production makes only sense for 250 units or more.

You should have some experience with electronic. Hand soldering is not worth the time and money (lower yield rate)

Quote

2. By using the adaptations you mentioned, would it be possible to use one of the ATX power supplies supporting GPUS at the moment to power, let's say, a 30-module cluster ?

A 350 W ATX supply would be sufficient to power a 30 FPGA boards.

DeathAndTaxes

donator

Activity: 1218

Merit: 1079

Gerald Davis

Quote from: iambaboon on January 27, 2012, 10:50:24 AM

I don't have experience with hardware. Does this mean that one could get all the instructions to build one 1.15x module at home ? In that case, a friend of mine with some experience in electronics/electronic part mounting would know enough ?

You could but assembly is usually pretty cheap on a simple & small board design. You wouldn't be saving much at all. Trying to do it without very expensive specialized gear and experience means anything theoretical savings ($20 per board maybe) would be offset by losses due to defects.

Say you could manage to hand build boards w/ only a 5% defect rate. Net-net, you would knock a whole $10 off each board. Not a good trade for the hundreds of hours it would take to build hundreds of units.

iambaboon

member

Activity: 116

Merit: 10

bitcoin afficionado

2 questions:
1.

Quote from: ztex on October 20, 2011, 02:19:47 PM

License production programs can be offered too. The customer would purchase the empty PCB including a license fee and gets assembly data (stencil data, bill of material, pick and place data, ...). Estimated costs (including PCB, license fee, parts, assembly) based on the prices here in Germany (from official Xilinx Distributor) would be:
100 units: 170 EUR (about 230 USD)
250 units: 140 EUR (about 190 USD)

I don't have experience with hardware. Does this mean that one could get all the instructions to build one 1.15x module at home ? In that case, a friend of mine with some experience in electronics/electronic part mounting would know enough ?

2. By using the adaptations you mentioned, would it be possible to use one of the ATX power supplies supporting GPUS at the moment to power, let's say, a 30-module cluster ?

CA Coins

donator

Activity: 305

Merit: 250

Quote from: rupy on January 26, 2012, 08:36:38 PM

New firmware out with long-polling. http://www.ztex.de/btcminer/

Nice! Will test LP out. Thanks.

rupy

hero member

Activity: 725

Merit: 503

New firmware out with long-polling. http://www.ztex.de/btcminer/

DeathAndTaxes

donator

Activity: 1218

Merit: 1079

Gerald Davis

Quote from: rupy on January 14, 2012, 11:14:48 AM

D&T are you using p2pool? What would you say of the payout compared to say deepbit?

I have only one rig on it. With 10GH I need to be careful moving them around. Everything looks solid. There is significant variance though so it isn't suitable (at this point) for miners who obsess over how much they make per day. For me my power bill comes once a month so what matters is how much variance I have month to month not day to day.

As p2pool grows variance should be less of an issue. At ~150GH/s it should find blocks on average every 12 hours. That means even an unlucky period should be 10+ blocks per week.

TL/DR version:
p2pool works BUT it is "cutting edge" expect it to be rough around the edges.

rupy

hero member

Activity: 725

Merit: 503

D&T are you using p2pool? What would you say of the payout compared to say deepbit?

DeathAndTaxes

donator

Activity: 1218

Merit: 1079

Gerald Davis

Quote from: rupy on January 13, 2012, 06:48:02 PM

But p2pool might be an important step for bitcoin so that makes it a bit more interesting!

Yeah and on p2pool you will get about 2 to 3 LP per minute. Roughly 20x as often as Bitcoin main network. So LP support is mandatory (well without massive stale rates).

Topic: ZTEX USB-FPGA Modules 1.15x and 1.15y: 215 and 860 MH/s FPGA Boards - page 40. (Read 182450 times)