Bitcoin Forum

For some reason my 5830 hashing power keeps fluctuating from 200 to 330.  Not sure why this is happenning.  Never happened before.  I have two cards a 5770 and 5830.00 Thanks for any help.

Need more information.

software
OS
ATI driver version
what miner are you using...
as well as extra paramaters

hardware
motherboard
powersupply

Overclocks?

temperatures of the cards...

thats a good start

330mh/s is pushing a 5830 pretty hard.  If your temps are hitting ~90C that could easily be thermal throttle.

330 is what i've come to expect from sapphire cards, but i run in a "forced open air" rack i built with the cards on risers and a boxfan cooling two rigs so they hit the 60s on a bad day, that does sound like a temperature problem though what is your fan running at have you tried setting it to 80% fixed?

yes, full specs as per MadHacker post pls.

330 Mh/s 24/7 here.  open air @ around 60% fan (cgminer controls fans/clocks). see sig. the sapphire 5830 xtremes go like stink.. does weeks on end with the only shutdowns being thunderstorms.

5830 problems

wow, that's a lot of problems.

Dude :D 1040/354 on a sapphire 5830 !! Your a cruel cruel man lol

its stock volts, so... DIG HARDER is my motto lol

it will run at 1050/354 for a couple days before crashing, but I want ROCK solid, so 1040 it is. that rig is in the basement and I check it with teamviewer every few days.

I ran mine at 1045/300 for two weeks when i first got it!

I run mine at 1040/350 (338 mhash)... ran them at 1045 for a while but they would usually crash after about a day or two. But then again I always make sure they have sufficient cooling.

I can't get mine beyond 960, at 960 it runs steadily for days, but any higher and it crashes, if I try 1000 it crashes within minutes, but the temp never gets very high

what voltage are you giving it?

YOu can definately get more mhashes our of your 5830. I'm getting 336.7 on mine running 1035/330. GUIMiner, phoenix, -k phatk2 VECTORS AGGRESSION=14

teehee, +rep

There are lots of factors in play that determine your max stable clocks.  On one of my boards, anything over 960 causes hard locks.  An identical board runs the same cards at 1010/320 with rock solid stability, and could probably go higher.  Yet another different board tops at 970.

OP needs to give more info if he wants any useful answers.   

I cannot seam to consistently get any card above 950.  Even my vapor X rig will go down after a day at 960. but at 940 it runs like a deer.

myself i have 6 cards in 1 machine...
all running at 950 except 1.. it runs at 940.
in my other machine i have them 4 of them clocked to 910.
i could probably clock them higher but then i have to figure out which one doesn't like it
but the lost compute time while its down irks me

What temperatures are you guys getting on the VRMs? I have no problems keeping the GPU, a 5850, cool (below 50C, using accelero twin turbo cooler) and it will overclock almost arbitrarily even when undervolting atm, but not without VRM temps heading towards 90C and thats more than Im comfortable with.

my sapphire xtreme 5830 does not have a VRM trmp sensor (that I know of anyway, and GPUZ does not report anything that looks like a VRM temp). its a non ref cooler; VRMs have their own HS.

To actually read the question posed and attempt to discover a solution: is this a dedicated mining rig? The first thing I would do is look at the CPU usage to see if you are being bitten by the CPU bug on all non-11.6 ATI drivers. If the mining CPU usage is 100%, than any other CPU usage on the computer that takes priority will idle the processes of the miners.

Secondly, using flash to watch videos (Youtube, etc) will put the GPU into 2D clock mode and make it run at half the speed. You need to disable hardware acceleration on any video player software; in flash, you right-click on the window and uncheck the hardware acceleration box.

Beyond that, more information is needed - where are you even seeing the hashrate change in software? I've noted phoenix to report varying hashrates every refresh, and with the -a 100 option you can average the hashrate display to remove the ups and downs.

In GPU-z its called VCC Phase I think. But not all cards support it. If yours doesnt, that would have me worried:

I use an aftermarket cooler on my 5850. I use the standard base plate to keep the ram and VRMs cool. With the stock cooler, even at 100%, I had much higher GPU temps and comparable (but still slightly higher) VRM temps than what I have now.

At stock voltage and stock speed my VRMs go to ~75C, while my GPU itself is no hotter than 45C. Overclocked to 900 MHz (from 725) and still at stock voltage,  VRMs are closer to 90C, GPU temp is still only 50C.  90C seems too high for me, even if they are supposedly rated for 120C,  so I undervolt and clock my card at 850.

THose of you overclocking and particularly overvolting  should really worry about those VRMs I think. Its not difficult to keep the GPU cool, but  I wouldnt be surprised if you were hitting way above 100C on the VRMs. I read somewhere on this forum you can buy IR thermometer for a $10, Id consider doing just that if your card doesnt support VRM monitoring.

nothing like that with the sapphire xtreme 5830. GPUZ reports a gpu temp off by itself, and gpu temps1, 2, and 3 a bit lower in the list. all track to within 3* C, so I doubt any of them are a VRM temp.

the VRMs have a pretty beefy separate heatsink, and its an open case with an external 120 volt ~200 CFM 120mm aimed at it.  so dig till ya drop baby! seems lots of this particular model 5830 are running in the 1000+ range and so far so good.

I do consider that particular card disposable though. no resale value to speak of and no use besides mining to me. guess we will see.