Pages:
Author

Topic: BAMT version 0.5 - Easy USB based mining Linux with farm wide management tools - page 19. (Read 324176 times)

hero member
Activity: 700
Merit: 500
940/815 is not possible with my cards. i can only go 100 lower. so now they are running at 940/840 with 290mhs
but the temp is what i am worryed about... now with 100% fan temps are 62 to 77 so thats a bit high for actual time and temp outside. i do not want to know how much they are when its hot outside. my miner stands outside in my balkony. so thats why klima affects directly. but under windows fans where automatic and mostly on 60% but temp of cards was lower... only when it was extremely hot outside they went up to 84.
btw i am alway wondering how much they can have until pc restarts or cards die. will linux also stop mining when cards go to hot?

Weird, this 125 difference is working for my 6xxx.
If temperatures are worrying you, yes go with the bios flash.
hero member
Activity: 561
Merit: 500
940/815 is not possible with my cards. i can only go 100 lower. so now they are running at 940/840 with 290mhs
but the temp is what i am worryed about... now with 100% fan temps are 62 to 77 so thats a bit high for actual time and temp outside. i do not want to know how much they are when its hot outside. my miner stands outside in my balkony. so thats why klima affects directly. but under windows fans where automatic and mostly on 60% but temp of cards was lower... only when it was extremely hot outside they went up to 84.
btw i am alway wondering how much they can have until pc restarts or cards die. will linux also stop mining when cards go to hot?
hero member
Activity: 700
Merit: 500

Take your time, but you know, maybe you´ll start first by using that 125 difference.
I had my cards flashed to a lower mem and frankly is not THAT much improvement.
940/815 is not that bad.
  
hero member
Activity: 700
Merit: 500
...dynamically resizes the persistence partition to the size of the key.  It may work on 512MB key, it will definitely work on 1GB keys.
Excellent!
hero member
Activity: 616
Merit: 506
Thanks lodcrappo, that is good to know.
Have another question: how big the new BAMT will be?

Why I am asking this, the old BAMT that was meant for at least a 2GB stick was in fact too big for some 2gb sticks, which were in fact 1.8GB.
I always had to shrink the image prior to load it on my stick. So maybe you could make the image a biiiit smaller?
Something like 1.7GB or eventually 3.7GB whatever is the case.

bamt 0.6 has much smaller requirements and dynamically resizes the persistence partition to the size of the key.  It may work on 512MB key, it will definitely work on 1GB keys.
hero member
Activity: 561
Merit: 500


6xxx series cards are limited to a memclock speed that is 100Mhz less then the GPU clock unless you flash a hacked bois onto it.  The voltage is also locked.  Afterburner uses a hack to change the memory speed, and the voltage. It's a hack that is only available on windows.  BAMT is a linux OS.

So, you have 3 choices, use windows, use BAMT with higher memory speeds, or flash a new BIOS to the card.

thank you for that fast answer. Is there a howto for flashing bios? never did that before. windows is not so comfortable to use. the nice thing on bamt is that it just starts mining when the pc starts. so it makes it really simple.
hero member
Activity: 700
Merit: 500
Thanks lodcrappo, that is good to know.
Have another question: how big the new BAMT will be?

Why I am asking this, the old BAMT that was meant for at least a 2GB stick was in fact too big for some 2gb sticks, which were in fact 1.8GB.
I always had to shrink the image prior to load it on my stick. So maybe you could make the image a biiiit smaller?
Something like 1.7GB or eventually 3.7GB whatever is the case.
hero member
Activity: 616
Merit: 506
You also have factors such as one GPU effecting the behavior of another in the same rig
How?

Dozens of ways.  Beyond the obvious thermal interaction, all your GPUs share the same bus.  GPU 0 might run fine at 900Mhz itself, but in doing so it holds the bus just a tiny bit too long or pushes the PCIX controller chip just a little too far in one way or another, and the next GPU down the line now has less time/a weird state/god knows what.

As a very real example that took me weeks to sort out: for my test lab I have 5830s, 6870s, 6950s and a 7970 (thanks donors).  If I run the 5830s by themselves, they will mine stable at 980/300 for weeks.  However, if I put in a 6870 and o/c it to 940, *one of the 5830s will lock up*.  The 6870 never locks up itself.  Yet if I drop the 6870 to 920.. stable for weeks again.

When you over/underclock you push the hardware out of its normal tolerance, plain and simple.  This can cause that particular hardware to malfunction, or it can cause effects to other things that communicate with that hardware.

hero member
Activity: 700
Merit: 500
You also have factors such as one GPU effecting the behavior of another in the same rig
How?

Quote from: TheHarbinger
6xxx series cards are limited to a memclock speed that is 100Mhz less then the GPU clock
Not 125?
sr. member
Activity: 378
Merit: 250
Why is it so damn hot in here?
Hi together,

i am now using BAMT for my little miner at home. First of all: Its great and really easy to use!! Thanks for that.
But one little problem:
I have 4 6870 cards. Till now i had windows running and i clocked the cards to 940mhz and 340mem using afterburner.
i have set these settings in the bamt.conf. overclocking works, but the mem will not change. when i enter 340 or 550 or anything, it stays at 1050 when i look into the webstats.... so what could be the problem?

sorry if somebody asked it before, but as you see, my english is not the best, and there are too many sites to read them all.
i do not know its because of that, but till now i had 300mhs with each card. the fans didnt run on full speed, and the temps where about 60-78 degrees. but now i have only 290mhs and fans run on full speed but the cards often get 80 or more degrees celsius. so something must be different now...

i hope somebody can help me.
thanks and regards
mameise

6xxx series cards are limited to a memclock speed that is 100Mhz less then the GPU clock unless you flash a hacked bois onto it.  The voltage is also locked.  Afterburner uses a hack to change the memory speed, and the voltage. It's a hack that is only available on windows.  BAMT is a linux OS.

So, you have 3 choices, use windows, use BAMT with higher memory speeds, or flash a new BIOS to the card.
hero member
Activity: 561
Merit: 500
Hi together,

i am now using BAMT for my little miner at home. First of all: Its great and really easy to use!! Thanks for that.
But one little problem:
I have 4 6870 cards. Till now i had windows running and i clocked the cards to 940mhz and 340mem using afterburner.
i have set these settings in the bamt.conf. overclocking works, but the mem will not change. when i enter 340 or 550 or anything, it stays at 1050 when i look into the webstats.... so what could be the problem?

sorry if somebody asked it before, but as you see, my english is not the best, and there are too many sites to read them all.
i do not know its because of that, but till now i had 300mhs with each card. the fans didnt run on full speed, and the temps where about 60-78 degrees. but now i have only 290mhs and fans run on full speed but the cards often get 80 or more degrees celsius. so something must be different now...

i hope somebody can help me.
thanks and regards
mameise
hero member
Activity: 616
Merit: 506
Lodcrappo,

Once a machine stops responding to cgsnoop, mgpumon 'remembers' the last speed that rig had and still uses it to calculate the total speed of the farm. Is there a specific reasoning behind this?

not really, just lazy programming
sr. member
Activity: 349
Merit: 250
Lodcrappo,

Once a machine stops responding to cgsnoop, mgpumon 'remembers' the last speed that rig had and still uses it to calculate the total speed of the farm. Is there a specific reasoning behind this?
hero member
Activity: 616
Merit: 506
underclocking/undervolting will cause instability issues just as much as overclocking.  screwing with your GPU == potential instability.

Well, bollocks.  Angry

Okay, as and when each rig next screws up, I will revert their gpus back to 725 engine and 1.05 voltage. But can I at least keep memclock at 300? Because the stock 1000 really is a stupid setting for a miner.

It is a matter of degrees, not absolutes.  If your GPU locks up in a matter of minutes, you're probably a long way from stable.  If it takes hours, maybe 10Mhz.  If it takes days, maybe another 5 or 10.  Some GPUs will run for a couple weeks or so but eventually lock up at XXXMhz, but be stable for months and months at XXX-5 Mhz.  To make it even more fun, two identical GPUs (same manufacture/model/etc) will find happiness at different clock and/or mem speeds.  You also have factors such as one GPU effecting the behavior of another in the same rig.

It can take a great deal of experimentation to squeeze that last little 2-3% out of a rig.  Rarely worth it.  Some people get lucky, a lot of people brag about mh/s speeds without waiting a month to see if they are actually stable, and even more people waste their time and money chasing the dream.

hero member
Activity: 492
Merit: 503
That's a well known problem, most likely caused by network-manager problems.  I've removed network-manager from the rigs I have that cause me that problem.

Nuh-uh. In another thread I was advised to replace network-manager with wicd.

I did.

My rigs just hate me.
hero member
Activity: 956
Merit: 1001
underclocking/undervolting will cause instability issues just as much as overclocking.  screwing with your GPU == potential instability.

In unrelated news, one of my other, better-behaved rigs has been down seven hours (why so long? Because I was ASLEEP, a state we humans often need to be in when the big light in the sky goes away) because all of a sudden it just COULDN'T DETECT ANY NETWORKS ANYMORE. Soon I'm just going to take these rigs to the vet and ask her to put them down humanely.


That's a well known problem, most likely caused by network-manager problems.  I've removed network-manager from the rigs I have that cause me that problem.
hero member
Activity: 492
Merit: 503
underclocking/undervolting will cause instability issues just as much as overclocking.  screwing with your GPU == potential instability.

Well, bollocks.  Angry

Okay, as and when each rig next screws up, I will revert their gpus back to 725 engine and 1.05 voltage. But can I at least keep memclock at 300? Because the stock 1000 really is a stupid setting for a miner.

Thanks pehoko for recommending testing the USB sticks. That's certainly another avenue of pain.

In unrelated news, one of my other, better-behaved rigs has been down seven hours (why so long? Because I was ASLEEP, a state we humans often need to be in when the big light in the sky goes away) because all of a sudden it just COULDN'T DETECT ANY NETWORKS ANYMORE. Soon I'm just going to take these rigs to the vet and ask her to put them down humanely.

(ETA and my first draft of this post got bollixed because my router conked out. Not my old crappy router, the shiny new wireless-N one I got to replace it LAST WEEK, the one that cost 60 quid. I swear my whole house is just one big fucking Ancient Indian Burial Ground.)
full member
Activity: 232
Merit: 100
Test your clear formated flash memory with software for read/write errors. I had the same problem and now I am with new drive and waiting another for the same problem. In my opinion maybe 2-4 months is normal long life for cheap hardly used bamt flash drive.
Pages:
Jump to: