ROCKMINER ASIC miner official thread

[Ntrain2k]

^^^ 6 New R-Box's, Official Main Branch CGMiner identifies these as LIR. ^^^

Damn, I'm gonna have to give this a try. Looks like she's up to speed now!

[hurricandave]

^^^ 6 New R-Box's, Official Main Branch CGMiner identifies these as LIR. ^^^

Damn, I'm gonna have to give this a try. Looks like she's up to speed now!

The rejects at the poolside are MUCH BETTER. In comparison Kano.is WAS showing an average 5.23% Invalids w/Rockminer Fork but after 12 hrs on this (4.8.0) I'm at 1.68% Invalids, I bet Kano will appreciate it!

[Ntrain2k]

^^^ 6 New R-Box's, Official Main Branch CGMiner identifies these as LIR. ^^^

Damn, I'm gonna have to give this a try. Looks like she's up to speed now!

The rejects at the poolside are MUCH BETTER. In comparison Kano.is WAS showing an average 5.23% Invalids w/Rockminer Fork but after 12 hrs on this (4.8.0) I'm at 1.68% Invalids, I bet Kano will appreciate it!

What's your bat look like?

My speed is still slow with this one.



cgminer.exe --icarus-options 115200:1:1 --rock-freq 310 -o stratum+tcp://stratum.westhash.com:3334 -u XXXX -p XXXX --no-submit-stale

[bitsolutions]

^^^ 6 New R-Box's, Official Main Branch CGMiner identifies these as LIR. ^^^

Damn, I'm gonna have to give this a try. Looks like she's up to speed now!

The rejects at the poolside are MUCH BETTER. In comparison Kano.is WAS showing an average 5.23% Invalids w/Rockminer Fork but after 12 hrs on this (4.8.0) I'm at 1.68% Invalids, I bet Kano will appreciate it!

What's your bat look like?

My speed is still slow with this one.

[img]http[Suspicious link removed] --icarus-options 115200:1:1 --rock-freq 310 -o stratum+tcp://stratum.westhash.com:3334 -u XXXX -p XXXX --no-submit-stale

Don't set "--icarus-options" they are not needed in official cgminer.

[Ntrain2k]

^^^ 6 New R-Box's, Official Main Branch CGMiner identifies these as LIR. ^^^

Damn, I'm gonna have to give this a try. Looks like she's up to speed now!

The rejects at the poolside are MUCH BETTER. In comparison Kano.is WAS showing an average 5.23% Invalids w/Rockminer Fork but after 12 hrs on this (4.8.0) I'm at 1.68% Invalids, I bet Kano will appreciate it!

What's your bat look like?

My speed is still slow with this one.



Crap. Still slow.


--rock-freq 310 -o stratum+tcp://stratum.westhash.com:3334 -u XXXX -p XXXX --no-submit-stale

Crap. Still slow.



--rock-freq 310 -o stratum+tcp://stratum.westhash.com:3334 -u XXXX -p XXXX --no-submit-stale

[-ck]

Crap. Still slow.



--rock-freq 310 -o stratum+tcp://stratum.westhash.com:3334 -u XXXX -p XXXX --no-submit-stale

Try without clock settings first (default is 270) and on another pool next just for testing purposes.

[notlogic]

Tonight was the first night I had time to work with my new R-Box since last weekend. Using 4.7.1 it has been running around 91 GH/s cgminer, but around 73 GH/s thanks to a nearly 20% error rate. (I previously thought my errors improved when I moved to GHash, but I just wasn't accustomed to their different way of displaying errors). I get these results at pretty much any frequency.

I tried 4.8 today and, at first, I was thrilled with the results. At 270-290 MHz I get only about 2-3% errors, which is fantastic. It runs in the low 80's GH/s, which means slightly better results than 4.7.1.  However, after 10-15 minutes it will tank down to about 40-50 GH/s. I have no idea why it's doing that, but it only happens with 4.8.

Tomorrow I'm going to try using my laptop in case it's something to do with my computer. I tried BFG briefly, but couldn't get it to find my device. I'll give that another shot tomorrow, too.

(please take all of this with a grain of salt, as I'm very new to ASIC mining)

[hurricandave]

BFGMiner and CGMiner use different drivers, you will not likely get BFGMiner to recognize any device with Zadig drivers installed and would possibly have to even delete Zadig before getting the SiliconLabs CP210x USBtoUart drivers to work, not always but some people have had issues with these two coexisting on the same computer. As far as hash rate falling off, make sure to watch what the temps are reporting as well as making sure something else in the house isn't trying to consume your bandwidth from your router.

[bitsolutions]

^^^ 6 New R-Box's, Official Main Branch CGMiner identifies these as LIR. ^^^

Damn, I'm gonna have to give this a try. Looks like she's up to speed now!

The rejects at the poolside are MUCH BETTER. In comparison Kano.is WAS showing an average 5.23% Invalids w/Rockminer Fork but after 12 hrs on this (4.8.0) I'm at 1.68% Invalids, I bet Kano will appreciate it!

What's your bat look like?

My speed is still slow with this one.



Crap. Still slow.


--rock-freq 310 -o stratum+tcp://stratum.westhash.com:3334 -u XXXX -p XXXX --no-submit-stale

Crap. Still slow.



--rock-freq 310 -o stratum+tcp://stratum.westhash.com:3334 -u XXXX -p XXXX --no-submit-stale

Can you see what happens if you disconnect the normal r-boxes(LIN devices)?

[hurricandave]

update: Now at 24hrs running 6 New R-Box V.2 with CGMiner 4.8.0 in --load-balance  mode on Kano.is and BTCGuild

[AJRGale]

...Speaking of these New R-Box's, there's a version 1 and 2 yes? anyone got details on the differences between them? physical and software/firmware?
Im wondering what one i have now, if its Version 1 that may be my problem...

also, getting 61.86/49.53GHs on Kanos pool, not the 100/90 im hoping for..

[hurricandave]

No, its not a version 1 and 2 of the "new" R-Box but people try to differentiate between the 32-37 GH/S original R-Box and the "Newer" R-Box which is rated at 100GH/S but for some reason was given the same name without any "Official" designation to help people tell the difference other than "New" which is often confused with "New to Me" equipment.

[AJRGale]

No, its not a version 1 and 2 of the "new" R-Box but people try to differentiate between the 32-37 GH/S original R-Box and the "Newer" R-Box which is rated at 100GH/S but for some reason was given the same name without any "Official" designation to help people tell the difference other than "New" which is often confused with "New to Me" equipment.

I see, you call the New Rbox as ver2, i just call them NRBox... how about we all them by how many cores they have? r-box = 4 core, new rbox = 12 core?

...since i really hate the name "New R-box"

not going into politics

[-ck]

update: Now at 24hrs running 6 New R-Box V.2 with CGMiner 4.8.0 in --load-balance  mode on Kano.is and BTCGuild

Looks good.

Since the bulk of the rejects still remain duplicates, I've expanded on the duplicate detection and filtering code which should decrease rejects further. The driver should detect any remaining duplicates as (extra) hw errors instead of submitting them now. Try the latest git for linux or updated windows binary:
http://ck.kolivas.org/apps/cgminer/temp/cgminer.exe

There are no performance enhancements in this update, just the duplicate filtering.

[hurricandave]

update: Now at 24hrs running 6 New R-Box V.2 with CGMiner 4.8.0 in --load-balance  mode on Kano.is and BTCGuild

Looks good.

Since the bulk of the rejects still remain duplicates, I've expanded on the duplicate detection and filtering code which should decrease rejects further. The driver should detect any remaining duplicates as (extra) hw errors instead of submitting them now. Try the latest git for linux or updated windows binary:
http://ck.kolivas.org/apps/cgminer/temp/cgminer.exe

There are no performance enhancements in this update, just the duplicate filtering.

One big difference I noticed is that in balanced mode with 4.8.0 I will always see messages that pool 1 detected new block followed by pool 0 requested work restart. Then about 45-60 seconds later I get the opposite - pool 0 detected new block and its followed by pool 1 requested work restart. So its kinda hard to tell how many shares are counted or not as rejects if/when it actually does the double restart. I know there is absolutely no way to sync the two pools but it may have had a slightly negative bearing on my numbers posted.
edit: For some reason load balance tends to favor BTCGuild in share count and hash rate over the connection to Kano.is, not by much, but over time it adds up.

[-ck]

One big difference I noticed is that in balanced mode with 4.8.0 I will always see messages that pool 1 detected new block followed by pool 0 requested work restart. Then about 45-60 seconds later I get the opposite - pool 0 detected new block and its followed by pool 1 requested work restart. So its kinda hard to tell how many shares are counted or not as rejects if/when it actually does the double restart. I know there is absolutely no way to sync the two pools but it may have had a slightly negative bearing on my numbers posted.
edit: For some reason load balance tends to favor BTCGuild in share count and hash rate over the connection to Kano.is, not by much, but over time it adds up.

It depends on which pool gives you the notification first, that's normal behaviour for cgminer. It will use the work that is valid from the new block in preference to the pool that still thinks it's on the old block. It may just be that you have a lower latency connection to btcg from where you are. Kano's pool is west coast USA.

[hurricandave]

Ok, that makes sense, I'm in Texas and I know BTCG is West of me but your a little further down the line.

[notlogic]

BFGMiner and CGMiner use different drivers, you will not likely get BFGMiner to recognize any device with Zadig drivers installed and would possibly have to even delete Zadig before getting the SiliconLabs CP210x USBtoUart drivers to work, not always but some people have had issues with these two coexisting on the same computer. As far as hash rate falling off, make sure to watch what the temps are reporting as well as making sure something else in the house isn't trying to consume your bandwidth from your router.

Thanks for the driver advice, that helped me get BFGminer running.

With my New R-Box I've now tried cgminer versions 4.3.3, 4.5, 4.7, 4.7.1, and 4.8. I've also tried BFGminer 4.10(I think that's the version). I've tried these at multiple frequencies, with different USB cables, on different computers, and on different pools.

The best results I've gotten are with pre-4.8 cgminer versions running 270 to 290 MHz. These results are typically about 90 GH/s reported by cgminer with lots of errors, pool-side usually comes out to low-70's GH/s with 15-25% error rate. cgminer 4.8 starts out around 70-80 GH/s, then settles around 45 GH/s with 2-3% errors. BFGminer reports mid-70's GH/s pre-error, but an effective mid-40's GH/s with a 35+% error rate.

I could be wrong... but I think I have a faulty box. I emailed the dealer tonight and I'm hoping they offer to exchange.

[AJRGale]

BFGMiner and CGMiner use different drivers, you will not likely get BFGMiner to recognize any device with Zadig drivers installed and would possibly have to even delete Zadig before getting the SiliconLabs CP210x USBtoUart drivers to work, not always but some people have had issues with these two coexisting on the same computer. As far as hash rate falling off, make sure to watch what the temps are reporting as well as making sure something else in the house isn't trying to consume your bandwidth from your router.

Thanks for the driver advice, that helped me get BFGminer running.

With my New R-Box I've now tried cgminer versions 4.3.3, 4.5, 4.7, 4.7.1, and 4.8. I've also tried BFGminer 4.10(I think that's the version). I've tried these at multiple frequencies, with different USB cables, on different computers, and on different pools.

The best results I've gotten are with pre-4.8 cgminer versions running 270 to 290 MHz. These results are typically about 90 GH/s reported by cgminer with lots of errors, pool-side usually comes out to low-70's GH/s with 15-25% error rate. cgminer 4.8 starts out around 70-80 GH/s, then settles around 45 GH/s with 2-3% errors. BFGminer reports mid-70's GH/s pre-error, but an effective mid-40's GH/s with a 35+% error rate.

I could be wrong... but I think I have a faulty box. I emailed the dealer tonight and I'm hoping they offer to exchange.

same boat as you... less the ability to call on the dealer :/

since you can get bfgminer working with yours, can you gimme a run down on what each chip is showing? im looking for worse hashrate, and higher heat on some chips..

when i get my main gaming rig back up, im going to get mine running on bfgminer again

[notlogic]

I've since uninstalled the drivers for BFGminer and gone back to cgminer. However, I did check individual chips when I had BFGminer. The lowest error rate was low-teens, the highest was 53%.