Topic: 5970 mining thread - page 5. (Read 25623 times)
Yep that was the one. Should have bookmarked it. IIRC, for some vector settings, hashrate increased with clocks, but the fastest performance was achieved with other settings and very low memory clocks.
Does dropping the memory clock affect the overall hash rate either up or down? In my own testing it seems to not affect it one way or the other, but I haven't been very scientific.
It depends on the miner app and settings. Ive seen some fancy chart that showed hashrate plotted against memory clock, and in most cases, hashrate actually decreased (only very very marginally, a few %) with increasing vram clock rates, though it wasnt quite a straight line. This is counter intuitive, but apparently had some solid theoretical explanation, something about caching algorithm, but I dont quite remember how it worked.
Cant seem to find that graph or thread, if someone knows it, Id love to see it again.
I believe the graph you were referring to, was based off of different "Vectors" V2 vs V4, and so forth. It showed the different hashrates at each memory & GPU clock, per vector.
I can't seem to find it either, but it was nice :/
So I dropped my memclocks and bumped my intensity. Got a minor increase in shares per minute (which matters more than hashrate). I'm tempted to bring GPU 1 up another 5 MHz to match GPU 0, but when I ran it like that before it crashed after a couple days.
GPU 0 and 1 are my 5970. I'm really temped to buy a second card now that my paycheck cleared.
GPU 0 and 1 are my 5970. I'm really temped to buy a second card now that my paycheck cleared.
Code:
cgminer version 2.1.0 - Started: [2011-12-30 04:16:36]
--------------------------------------------------------------------------------
(5s):1081.5 (avg):1082.7 Mh/s | Q:59366 A:54757 R:1627 HW:0 E:92% U:14.45/m
TQ: 5 ST: 6 SS: 1 DW: 2356 NB: 396 LW: 2803 GF: 245 RF: 8
Connected to http://goat1.zapto.org:8337 with LP as user redemerald1
Block: 00000ada8101be812ceac86e64822a87... Started: [19:20:00]
--------------------------------------------------------------------------------
[P]ool management [G]PU management [S]ettings [D]isplay options [Q]uit
GPU 0: 74.5C 4364RPM | 382.5/381.8Mh/s | A:19264 R:592 HW:0 U:5.08/m I:9
GPU 1: 70.5C | 380.2/379.6Mh/s | A:19106 R:626 HW:0 U:5.04/m I:9
GPU 2: 69.5C 1569RPM | 317.0/321.3Mh/s | A:16387 R:409 HW:0 U:4.32/m I:9
--------------------------------------------------------------------------------
GPU 0: 382.8 / 381.8 Mh/s | A:19264 R:592 HW:0 U:5.08/m I:9
74.5 C F: 85% (4354 RPM) E: 830 MHz M: 160 Mhz V: 1.049V A: 99% P: 0%
Last initialised: [2012-01-01 06:45:45]
Intensity: 9
Thread 0: 193.8 Mh/s Enabled ALIVE
Thread 3: 188.7 Mh/s Enabled ALIVE
GPU 1: 380.4 / 379.6 Mh/s | A:19105 R:626 HW:0 U:5.04/m I:9
71.0 C E: 825 MHz M: 160 Mhz V: 1.049V A: 99% P: 0%
Last initialised: [2012-01-01 14:44:38]
Intensity: 9
Thread 1: 191.6 Mh/s Enabled ALIVE
Thread 4: 188.6 Mh/s Enabled ALIVE
GPU 2: 316.5 / 321.3 Mh/s | A:16385 R:409 HW:0 U:4.32/m I:9
69.5 C F: 43% (1575 RPM) E: 1000 MHz M: 160 Mhz V: 1.162V A: 99% P: 0%
Last initialised: [2011-12-30 19:46:43]
Intensity: 9
Thread 2: 161.8 Mh/s Enabled ALIVE
Thread 5: 154.2 Mh/s Enabled ALIVE
[E]nable [D]isable [I]ntensity [R]estart GPU [C]hange settings
Does dropping the memory clock affect the overall hash rate either up or down? In my own testing it seems to not affect it one way or the other, but I haven't been very scientific.
It depends on the miner app and settings. Ive seen some fancy chart that showed hashrate plotted against memory clock, and in most cases, hashrate actually decreased (only very very marginally, a few %) with increasing vram clock rates, though it wasnt quite a straight line. This is counter intuitive, but apparently had some solid theoretical explanation, something about caching algorithm, but I dont quite remember how it worked.
Cant seem to find that graph or thread, if someone knows it, Id love to see it again.
Does dropping the memory clock affect the overall hash rate either up or down? In my own testing it seems to not affect it one way or the other, but I haven't been very scientific.
I haven't noticed any difference in hash rate with lower memory clocks. I think a reduced memory clock can allow for a higher engine clock which can increase hash rates.
Does dropping the memory clock affect the overall hash rate either up or down? In my own testing it seems to not affect it one way or the other, but I haven't been very scientific.
I ran intensity 9 on one GPU and 7 on the other for a few weeks. The one with 7 consistently had more accepted shares. Anyone else try this?
Does anyone have any hard numbers about memclock energy savings? I'm running 300 since that is the default bottom, but cgminer will let me go to 150. I'll have to plug in my kill-o-watt and take some readings
Does anyone have any hard numbers about memclock energy savings? I'm running 300 since that is the default bottom, but cgminer will let me go to 150. I'll have to plug in my kill-o-watt and take some readings
My rig w/ 3x 5970s uses 870w @ 825/160. It uses 885w @ 825/300. It uses 935W @ 825/1000. That was about 6 months ago. It is possible different driver versions have changed that somewhat. 160 seemed as stable and same speed as 300 so I have kept all the rigs @ 160. 15W isn't much (5w per card) but I guess every little bit helps.
I wonder if dropping the mem clocks to a lower value would also decrease the temps of VRMs...in theory, I would say yes. Agree ?
In theory however temp is linear w/ power right? Thus the smaller and smaller amounts of power reduction means smaller and smaller decreases in thermal output and thus temps. It can't hurt though.
Memory has it's own power phase (vrm). Dropping memory clocks will reduce the heat load on the heat plate and over all reduce temps for the gpu vrm's and even gpu core.
I ran intensity 9 on one GPU and 7 on the other for a few weeks. The one with 7 consistently had more accepted shares. Anyone else try this?
Does anyone have any hard numbers about memclock energy savings? I'm running 300 since that is the default bottom, but cgminer will let me go to 150. I'll have to plug in my kill-o-watt and take some readings
Does anyone have any hard numbers about memclock energy savings? I'm running 300 since that is the default bottom, but cgminer will let me go to 150. I'll have to plug in my kill-o-watt and take some readings
Use -I 8 for 5xxx cards and -I 9 for 6xxx cards. I believe that was the previous note in CGMiner.
Now it says: "NOTE: Running intensities above 9 with current hardware is likely to only
diminish return performance even if the hash rate might appear better. A good
starting baseline intensity to try on dedicated miners is 9. Higher values are
there to cope with future improvements in hardware."
Hm. I went back to running intensity 9 with memclocks down to 160. I'm not sure if intensity 7 card doing better was just luck. since my hashrate went from 1071 to 1083.
I ran intensity 9 on one GPU and 7 on the other for a few weeks. The one with 7 consistently had more accepted shares. Anyone else try this?
Does anyone have any hard numbers about memclock energy savings? I'm running 300 since that is the default bottom, but cgminer will let me go to 150. I'll have to plug in my kill-o-watt and take some readings
Does anyone have any hard numbers about memclock energy savings? I'm running 300 since that is the default bottom, but cgminer will let me go to 150. I'll have to plug in my kill-o-watt and take some readings
My rig w/ 3x 5970s uses 870w @ 825/160. It uses 885w @ 825/300. It uses 935W @ 825/1000. That was about 6 months ago. It is possible different driver versions have changed that somewhat. 160 seemed as stable and same speed as 300 so I have kept all the rigs @ 160. 15W isn't much (5w per card) but I guess every little bit helps.
I wonder if dropping the mem clocks to a lower value would also decrease the temps of VRMs...in theory, I would say yes. Agree ?
In theory however temp is linear w/ power right? Thus the smaller and smaller amounts of power reduction means smaller and smaller decreases in thermal output and thus temps. It can't hurt though.
I ran intensity 9 on one GPU and 7 on the other for a few weeks. The one with 7 consistently had more accepted shares. Anyone else try this?
Does anyone have any hard numbers about memclock energy savings? I'm running 300 since that is the default bottom, but cgminer will let me go to 150. I'll have to plug in my kill-o-watt and take some readings
Does anyone have any hard numbers about memclock energy savings? I'm running 300 since that is the default bottom, but cgminer will let me go to 150. I'll have to plug in my kill-o-watt and take some readings
My rig w/ 3x 5970s uses 870w @ 825/160. It uses 885w @ 825/300. It uses 935W @ 825/1000. That was about 6 months ago. It is possible different driver versions have changed that somewhat. 160 seemed as stable and same speed as 300 so I have kept all the rigs @ 160. 15W isn't much (5w per card) but I guess every little bit helps.
I wonder if dropping the mem clocks to a lower value would also decrease the temps of VRMs...in theory, I would say yes. Agree ?
I ran intensity 9 on one GPU and 7 on the other for a few weeks. The one with 7 consistently had more accepted shares. Anyone else try this?
Does anyone have any hard numbers about memclock energy savings? I'm running 300 since that is the default bottom, but cgminer will let me go to 150. I'll have to plug in my kill-o-watt and take some readings
Does anyone have any hard numbers about memclock energy savings? I'm running 300 since that is the default bottom, but cgminer will let me go to 150. I'll have to plug in my kill-o-watt and take some readings
Use -I 8 for 5xxx cards and -I 9 for 6xxx cards. I believe that was the previous note in CGMiner.
Now it says: "NOTE: Running intensities above 9 with current hardware is likely to only
diminish return performance even if the hash rate might appear better. A good
starting baseline intensity to try on dedicated miners is 9. Higher values are
there to cope with future improvements in hardware."
I ran intensity 9 on one GPU and 7 on the other for a few weeks. The one with 7 consistently had more accepted shares. Anyone else try this?
Does anyone have any hard numbers about memclock energy savings? I'm running 300 since that is the default bottom, but cgminer will let me go to 150. I'll have to plug in my kill-o-watt and take some readings
Does anyone have any hard numbers about memclock energy savings? I'm running 300 since that is the default bottom, but cgminer will let me go to 150. I'll have to plug in my kill-o-watt and take some readings
My rig w/ 3x 5970s uses 870w @ 825/160. It uses 885w @ 825/300. It uses 935W @ 825/1000. That was about 6 months ago. It is possible different driver versions have changed that somewhat. 160 seemed as stable and same speed as 300 so I have kept all the rigs @ 160. 15W isn't much (5w per card) but I guess every little bit helps.
I ran intensity 9 on one GPU and 7 on the other for a few weeks. The one with 7 consistently had more accepted shares. Anyone else try this?
Does anyone have any hard numbers about memclock energy savings? I'm running 300 since that is the default bottom, but cgminer will let me go to 150. I'll have to plug in my kill-o-watt and take some readings
Does anyone have any hard numbers about memclock energy savings? I'm running 300 since that is the default bottom, but cgminer will let me go to 150. I'll have to plug in my kill-o-watt and take some readings
Some people have gotten the memclock as low as 160 I think. You have to mess around with settings to see what works best.
In my current series of experiments I have memclock at 150.
810 on the engine clock was looking good until this morning when GPU 3, which seems to be my problem child, hung. So now I've got it at 800 and the other five at 810. We'll see...
lower memclock = lower energy use
It also increases hashrate slightly I believe, but it definitely lowers energy use which is huge.
Some people have gotten the memclock as low as 160 I think. You have to mess around with settings to see what works best.
DeathandTaxes has done alot of screwing around with 5970's and can probably add more to this discussion than i can.
It also increases hashrate slightly I believe, but it definitely lowers energy use which is huge.
Some people have gotten the memclock as low as 160 I think. You have to mess around with settings to see what works best.
DeathandTaxes has done alot of screwing around with 5970's and can probably add more to this discussion than i can.
Thanks for all the input. I'm messing around with it and learning all kinds of things. One question-- how does memclock affect things in terms of energy usage and hashrate?
Here's the relevant portion of my cgminer.conf:
"intensity" : "9",
"gpu-engine" : "810",
"gpu-memclock" : "150",
"gpu-fan" : "0-85",
"temp-cutoff" : "95",
"temp-overheat" : "85",
"temp-target" : "75",
"temp-hysteresis" : "3",
"auto-fan" : true,
with which I'm getting these Mh/s on the six cores in three 5970s:
369.4, 370.5, 370.5, 370.4, 370.4, 370.3
"intensity" : "9",
"gpu-engine" : "810",
"gpu-memclock" : "150",
"gpu-fan" : "0-85",
"temp-cutoff" : "95",
"temp-overheat" : "85",
"temp-target" : "75",
"temp-hysteresis" : "3",
"auto-fan" : true,
with which I'm getting these Mh/s on the six cores in three 5970s:
369.4, 370.5, 370.5, 370.4, 370.4, 370.3
@btc_artist
my cgminer mining script looks like this if you were using one 5970
./cgminer -o http://miningpool:port -u username -p password -I,9,9 --gpu-fan 85 --gpu-engine 800 --gpu-memclock 300
This gets about 700-725 MH/s for a single card average about 710 MH/s or so. This is for a stock Diamond card from newegg.
I am testing some cards that I redid the TIM and thermal pads on and have those at --gpu-engine 825 --gpu-memclock 300
Those are getting an about 730-750 MH/s for an average of 740 Mh/s or so.
For three cards I'm using same as above, but I,9,9,9,9,9,9.
my cgminer mining script looks like this if you were using one 5970
./cgminer -o http://miningpool:port -u username -p password -I,9,9 --gpu-fan 85 --gpu-engine 800 --gpu-memclock 300
This gets about 700-725 MH/s for a single card average about 710 MH/s or so. This is for a stock Diamond card from newegg.
I am testing some cards that I redid the TIM and thermal pads on and have those at --gpu-engine 825 --gpu-memclock 300
Those are getting an about 730-750 MH/s for an average of 740 Mh/s or so.
For three cards I'm using same as above, but I,9,9,9,9,9,9.
Yeah, I use Linuxcoin also, the final version on a usb stick. I follow the headless instructions from the Wiki.
If I already have powered extenders, can I just not plug in the molex? Could that be screwing things up?
Anyway, thanks for the input, I just have to play around some more I guess, got frustrated after 5 hours yesterday... At least I got three GPU's running, lol.
If I already have powered extenders, can I just not plug in the molex? Could that be screwing things up?
Anyway, thanks for the input, I just have to play around some more I guess, got frustrated after 5 hours yesterday... At least I got three GPU's running, lol.
Hey jimm, you're running four 5970's on each MSI 890FX-GD70, correct? You're also using two seasonic 750watt gold psus.
I just reset my rigs and used a lian li power supply adapter to get both psus running. I have been able to get three cards going on each rig for now, but I couldn't get a fourth going. I used pcix16 slots 1,3,5,6 like the MSI manual suggested and also tried 1,4,5,6.
I'm using the molex powered x16 extenders from cablesaurus as well.
Any suggestions or tips?
Thanks.
Oh and I have one power supply powering two cards and the CPU with one molex rail connected to two extenders.
The second psu is/was connected to the other two cards and one molex rail to power the other extenders.
I just reset my rigs and used a lian li power supply adapter to get both psus running. I have been able to get three cards going on each rig for now, but I couldn't get a fourth going. I used pcix16 slots 1,3,5,6 like the MSI manual suggested and also tried 1,4,5,6.
I'm using the molex powered x16 extenders from cablesaurus as well.
Any suggestions or tips?
Thanks.
Oh and I have one power supply powering two cards and the CPU with one molex rail connected to two extenders.
The second psu is/was connected to the other two cards and one molex rail to power the other extenders.
I didn't know MSI had a suggestion on which slots to use.
I do not use molex powered extenders. not needed, just use regular ones if you have them I try not to use molex plugs unless I am forced to .
as for seeing all four cards, most times all four show up OK. but for stubborn rigs I just keep moving things around and redoing the drivers until it works. sometimes I have to reset the bios on mobo.
what operating system? all of my 4X5970 rigs are linuxcoin on a stick
Jump to: