Topic: MSI 7870 - dead or still usabele? please post all who have same card (Read 1027 times)

1) Sapphire 7870 with cgminer 3.7.2, Catalyst 13.5/Win7 64, Sapphire Trixx
2)Trixx: GPU Clock 925, Memory Clock 1375, VDCC 1.050, Fan 50% Fixed
3)
"intensity" : "19",
"worksize" : "256",
"kernel" : "scrypt",
"lookup-gap" : "2",
"thread-concurrency" : "15232",
"shaders" : "1280",
"gpu-engine" : "925",
"gpu-fan" : "50",
"gpu-memclock" : "1375",
"gpu-memdiff" : "0",
"gpu-powertune" : "0",
"gpu-vddc" : "1.050",
"temp-cutoff" : "89",
"temp-overheat" : "85",
"temp-target" : "0",
"api-port" : "4028",
"expiry" : "120",
"gpu-dyninterval" : "7",
"gpu-platform" : "0",
"gpu-threads" : "1",
"hotplug" : "5",
"log" : "5",
"no-pool-disable" : true,
"queue" : "1",
"scan-time" : "60",
"scrypt" : true,
"temp-hysteresis" : "3",
"shares" : "0",
"kernel-path" : "/usr/local/bin"
}

Getting slightly less than 400KH/s (398,7) with temperatures lower than 60 Celsius.

Try -g 2 and a thread concurrency of 8192, keep the intensity and worksize and take all the other stuff out of the code and try that.

--thread-concurrency 8192 --intensity 12 --worksize 256 -g 2

Thanks for reply guys!

My PSU 1.450 KW, so it should be enough for such card, for now its only one on new rig.

Voltage overclock option never touched.

Using cgminer, tried both latest AMD drivers and older version. No difference

Tried with intensity 20, then 12: both giving almost same hashrate and same result: after 1 minute screen blink, cgminer stop working and system giving message "video adapted has stop work and driver repaired after crash".
And screen sometimes repaired same time, sometime card appears off and screen appears black, repaired only after restart-clear CMOS on motherboard.

Tried different settings, samples:
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_USE_SYNC_OBJECTS 1
cgminer --scrypt -o stratum+tcp://127.0.0.1:15522 -u  -p  --intensity 12 --worksize 256 -g 1 --thread-concurrency 5120 --auto-fan --auto-gpu --gpu-engine 1100 --gpu-memclock 850 --temp-target 72

or

cgminer --scrypt -o stratum+tcp://127.0.0.1:15522 -u  -p  --intensity 12 --worksize 256 -g 1 --thread-concurrency 15232  --auto-fan --auto-gpu --gpu-engine 1000 --gpu-memclock 1200 --temp-target 72


But what also funny is: now only "GPU Reactor" lights shining, other lights that normally shining on the card are off.
Monitor never ON with that card now.

Looks like its dead
Config file will post later today.

But strange: temperature never was higher then 60 C.
And no one stable long time mining has success with that MSI HAWK. Just for minute or two, then "screen crash, cgminer stopped".

Maybe just factory defect or something. Now I have feeling it was incorrect decision to buy that MSI card ...

i have a Radeon 7850 and i get 400 KHash/s?

Please post the following:

1) What program are you mining with?
2) What launch flags are you using (if any) when you launch your miner?
3) What does your config file look like for your mining program?

Black-blank screen and card crash can be a few issues.

1. Temperature crash.
2. Intensity too high, resulting in driver crash.
3. Undervolt or overclock too extreme. Card unable too perform.
4. Drivers (or parts of old drivers).
5. PSU not enough power.

Try reinstalling the newest drivers and run the cleaning tool first...

I have (2) Sapphire HD7870 GHz Edition cards. Both are running on stock settings and getting 325-350 KH/s. I was running at about 275KH/s until I installed the latest APP SDK and played with my CGMiner settings a bit. I have not had a crash running with my settings. I only crashed when trying to max out engine clock and memory clock.

What are your settings?

Hi All!

I just bought that card: MSI 7870 HAWK and....
1) This give me with any of drivers only 276 (MAX) KHs
2) After 1 minute of mining scren is black and card crash
3) after restart i can repair the card using cleaning CMOS

Anyone have an ideas or had same issues?
P.s - for tests only one card installed

Thanks!