[4+ EH] Slush Pool (slushpool.com); Overt AsicBoost; World First Mining Pool

[Psyrick]

Yes, I gave my GPUs a 5 minute cool down period near the end of the round. I don't think that deserves the loss of nearly all rewards from almost 2.5 hours of mining.

Just trying to be objective and helpful here. But I guess you could...

I appreciate your helpfulness. My cooling system is about as windy as I can make it, but I like to overclock to within the limits of it's heat transfer capacity. I'm trying to avoid going the liquid cooling route. I often do seize the chance for a cool down at the end of a long round, but when they peak up to 98C and there's no telling how much longer the round is going to be, I feel I don't have a choice. Only takes a couple minutes to get them back down to 40. After a cool-down period of my own, I find myself not really liking the waters in the Guild's pool, and as mentioned, most other pools either have the same scoring system and/or higher fees. While invalid blocks and scored out rewards give me the blues, I'm still jumping right back into Slush's pool.

Why are you clocking your cards so high? Find a happy median between heat and hashing and stick with it. I have plenty of boxes that have ran non-stop for months without ever looking at them. Are you giving the cards too much voltage?

You do realize that letting your cards heat up and then cool down is WAY worse for them than just letting them run hot right? 98C is simply too hot, pull them back into the 80's and you'll be golden. Personally I don't run much hotter than 80C if at all possible and find the clock and voltage settings to make sure that happens regardless of ambient temps.

Tuned down the voltage. Results are looking better. System is stable if I leave it alone to hash, but GPU crashes if I move the mouse too fast.
Vcc = 1188 mV    Core Clock = 1150 MHz    Mem Clock = 1515 MHz
GPU1 Temp = 88C      GPU2 Temp = 83C
Current Hash Rate = 672 Mhash/s
Default Hash Rate = 540 Mhash/s

[kanta]

2 x Powercolor Radeon 7970 mounted in a CoolerMaster HAF 925 case with extra 120mm fans inside. 2 for push-pull CPU radiator and 2 blowing on the video card stack from the side. Don't know if an Accelero cooler would perform a great deal better, since the stock ones seem pretty well designed. I may remount the stock ones though. I work at a precision optical manufacturing company and I can use the lapping wheel to grind flat the contact surface to within microns. Could even polish it to within a quarter of a red photon, but I think that would be waste of time, as the thermal paste layer is never close to that thin, and the mirror surface might actually do a good job of reflecting thermal infrared back at the chip. Does anyone here have any experience with diluting silicone based heatsink compound with silicone lubricant or other liquid to reduce viscosity and drastically thin-out the layer? I once did this with GPU heatsink on my Nvidia GTX 570. I got it so well contacted, that you could lift and swing the thing from the bond with no movement. Unfortunately, I forgot to reinstall the aluminum mounting frame before the heatsink, so I had to take it back apart, and the next time around I didn't dilute the compound. So no results to compare.

In your case, the accelero coolers would not be ideal due to the way that they cool.  They exhaust the hot air out through all directions except the face of the card where the fans suck air in and of course the PCB which is solid, but all sides and the rear are exhausting air, which would cause hot air from each card to be pulled in as cooling air, so it's not a good solution.  The 925 is a decent case for gaming, but for your application, it doesn't move enough air.  You could either get water cooling blocks for your 7970s, which would be pretty expensive, but would work nicely, or you could run an open computer, no case, but you'd end up having to clean the computer components a lot more often than with the case in place. Remounting the stock heat sinks would be ideal because it's much lower cost than water cooling and would still allow you to use your case.

I would recommend you clean off all of the existing thermal compound and examine the surface of the heat sink and where it mounts on your GPU/VRMs to ensure they are nice and flat, with no irregularities.  Once you have verified this and made both sides of the mating surfaces nice and clean with some IPA (isopropyl alcohol), then apply a nice small amount of high quality TIM to the heat spreaders on the GPU and VRMs and place the heat sinks back on.  If you want to be doubly sure, remove the heatsink again and look at the TIM, it should be devoid of any bubbles or clean areas (everything should ideally have a thin layer of TIM on it, little hills or ridges are normal). Clean it off again, as the removal will create irregularities in the TIM (little hills or ridges) and reapply the same amount if it was good, and reattach the heatsink, replace the shroud, and button everything back up.

I highly recommend a high quality TIM like the noctua NT-H1 that comes with the NH-D14 coolers, although you can buy it seperately.  People like Arctic Silver AS5, but it doesn't perform as well as several other TIMs, although the overall difference between best and average is only 4 or 5 degrees in the last comparison I read, so if AS5 is the only stuff you can get at a local shop, it's fine.  Poor application or too much TIM will make a much bigger difference than which brand you choose.  Another factor is cleaning the heat sinks and case to ensure all the airflow is as unrestricted as possible.

As for your performance, it really makes me wish I had a second card, but I'd probably run into heat issues.  I'm running mine at 1300 mV, 1230 MHz on the core clock, 1600 on the memory clock 72C 24/7.  Never crashes during mining, and I only turn off the miner while I'm gaming (primary purpose of my rig).  Hash rate bounces between 715 MH/s and 694 MH/s depending on what else I'm doing, I'd love to nearly double that, but not at the cost of a second 460 dollar card and the possibility of having to ramp them down significantly due to the heat blowing all over the case

[autonomous42]

Psyrick, it would also be a great idea to reduce the memory clock to as low as you can before the card gets sick. I was able to lop off 5-6 degrees per card by doing that and you don't lose performance as bitcoin mining is not memory-intensive. And really consider the bench I posted.

[Psyrick]

Psyrick, it would also be a great idea to reduce the memory clock to as low as you can before the card gets sick. I was able to lop off 5-6 degrees per card by doing that and you don't lose performance as bitcoin mining is not memory-intensive. And really consider the bench I posted.

Thank you for that suggestion. I was thinking it might be good to turn down the memory clock, but I wasn't sure if I'd lose stability due to the memory falling behind pace with the core. Also, I will consider the bench.

By the way. I really do appreciate the discussion I have received over this. I am very glad to be a member of this forum. The helpfulness of lack of hatefulness is very refreshing as compared to /b/.

[Xian01]

I was thinking it might be good to turn down the memory clock, but I wasn't sure if I'd lose stability due to the memory falling behind pace with the core. Also, I will consider the bench.

 Not to worry. I don't even know if the memory on the GPU's is being touched while hashing SHA256. You can safely down-clock your video memory to 300 and have them happily mining away at much cooler temperatures.

 Now if you are hashing Litecoin/Scrypt, that's another matter entirely, and you'll need to keep your memory normally clocked AFAIK.

[Vasilis]

4:03:45    33732253    none    none    none
wtf?

[desired_username]

I've been mining with 50Ghs for the 11th day now. Can you please evaluate the below log of earnings?

I watched that auction for the 10gig blades - it was awesome. How are your 5 running? Are they stable with temp etc? So jelly.

Actually I have "only" 4. They run stable, I cannot complain. Speed varies between 12.5-13Ghs / blade

I haven't got downtime caused by the miners as of yet. The cooling is quite easy compared to GPUs, I use 3 Fans for cooling all 4 blades.

[kanta]

4:03:45    33732253    none    none    none
wtf?

I'm seeing the same thing, and I haven't turned off my computer tonight at all. The "my account" page has me constantly submitting shares, so I'm not sure what's up with that. Hopefully it gets corrected at some point, 4 hours and all

[visdude]

4:03:45    33732253    none    none    none
wtf?

I'm seeing the same thing, and I haven't turned off my computer tonight at all. The "my account" page has me constantly submitting shares, so I'm not sure what's up with that. Hopefully it gets corrected at some point, 4 hours and all

Here we go again (sigh).  All my miners were happily hashing away at full bore in this long round.  The rounds before it and the next one coming up (if rewarded correctly) would attest to that fact.

[MyOldGeForce]

I have tried to install different drivers for the cards, and I have tried installing a previous version that supports both cards. Neither worked. It seems to be the combination of the new card and anything but the new driver that makes GUIMiner not start.

Now, when I am running only the new card, Phoenix and Stratum Proxy, it stopped working tonight, a few hours after I started it. I restarted the miner this morning and it worked again. I hope it's a one time occurance, but I doubt it. It didn't happen with Phoenix and BCPool. Any guesses?


My estimsted rewards seem to be about right. I'm not exactly sure what they should be, but they are not in the 7 satoshi order of magnitude... Currently it's 0.00051612 and I've been running 200 MHash/s for 1.5 hours.

Maybe you should use next workaround. Remove one card from a computer. Run cgminer and if it works you will get kernel for that card. If it doesn't work change SDK and run again. If hashrate is too low use diffident SDK/driver and remove .bin file from cgminer folder. When you find the best combination backup .bin file just in case. Now remove the card and put in other card. Do the same but don't delete first card .bin file only second card .bin file if you don't like hashrate. Now put both cards in and run cgminer. It might work.

And just in case. When running windows you need to plug monitor or null plug into card. Did you?

I have not got any miner working on it's own without GUIMiner, I have tried a few different miners and lots of different settings in the command line, both own ideas and found on the internet. Always one of two things happen: either the program exits/crashes/doesn't start, or it runs but gets 0 hashes/s. This when GUIMiner works, I haven't even tried it when it doesn't.


Why wouldn't GUIMiner even start if it was only needing a monitor? I had a monitor connected to the other card at some time, but it was not connected to electricity as I didn't have any more outlets and didn't feel the need for two monitors enough to dig up an outlet splitter. Does the sensing need power to the monitor, or is it just two connected pins or a resistor between two pins or something?



The other day, I also noticed the problem of low payouts. For some time, it was down to one satoshi.

I still get the problem with GUIMiner stopping, just saying "connecting...", but it only happens around once every two days, so it's managable. I think it's my Internet connection that's unstable, and GUIMiner somehow can't recover when the connection goes down temporarily.

[kanta]

I believe this is the workaround (pics and all in there to make it easy)

http://bitclockers.com/forums/index.php?topic=7.0

[vs3]

4:03:45    33732253    none    none    none
wtf?

I'm seeing the same thing, and I haven't turned off my computer tonight at all. The "my account" page has me constantly submitting shares, so I'm not sure what's up with that. Hopefully it gets corrected at some point, 4 hours and all

Here we go again (sigh).  All my miners were happily hashing away at full bore in this long round.  The rounds before it and the next one coming up (if rewarded correctly) would attest to that fact.

Same here :
18082    2013-05-18 07:45:28    4:03:45    33732253    none    none    none    236720    25.24790000    87 confirmations left

Hopefully it will get corrected. Also - at around 3hrs into the round my score got reset - it was almost 0 at 3:15 and by 3:45 it was recovering...

[Uliss]

18082    2013-05-18 07:45:28    4:03:45    33732253    none    none    none    236720    

What does that mean?

[kanta]

4:03:45    33732253    none    none    none
wtf?

I'm seeing the same thing, and I haven't turned off my computer tonight at all. The "my account" page has me constantly submitting shares, so I'm not sure what's up with that. Hopefully it gets corrected at some point, 4 hours and all

Here we go again (sigh).  All my miners were happily hashing away at full bore in this long round.  The rounds before it and the next one coming up (if rewarded correctly) would attest to that fact.

Same here :
18082    2013-05-18 07:45:28    4:03:45    33732253    none    none    none    236720    25.24790000    87 confirmations left

Hopefully it will get corrected. Also - at around 3hrs into the round my score got reset - it was almost 0 at 3:15 and by 3:45 it was recovering...

I wonder about the score resetting thing, isn't that part of the weighting system to counter pool hopping?  I notice all the time, my score will drop from the millions to nothing and then begin climbing back up

[organofcorti]

4:03:45    33732253    none    none    none
wtf?

I'm seeing the same thing, and I haven't turned off my computer tonight at all. The "my account" page has me constantly submitting shares, so I'm not sure what's up with that. Hopefully it gets corrected at some point, 4 hours and all

Here we go again (sigh).  All my miners were happily hashing away at full bore in this long round.  The rounds before it and the next one coming up (if rewarded correctly) would attest to that fact.

Same here :
18082    2013-05-18 07:45:28    4:03:45    33732253    none    none    none    236720    25.24790000    87 confirmations left

Hopefully it will get corrected. Also - at around 3hrs into the round my score got reset - it was almost 0 at 3:15 and by 3:45 it was recovering...

I wonder about the score resetting thing, isn't that part of the weighting system to counter pool hopping?  I notice all the time, my score will drop from the millions to nothing and then begin climbing back up

No, the reset is to prevent scores reaching very large numbers. It is thought that if anyone's score reaches +Inf, the universe will implode. So be grateful for the reset.

[huanix]

4:03:45    33732253    none    none    none
wtf?

I'm just glad I wasn't the only one. It appears to be a valid block, so I'm sure slush will get it fixed.

[kanta]

Hmm, more weirdness.  Account page says 724 MH/s, constant shares going out (last update at 0) every time I look, yet instead of the normal 0.0017-0.002 that I get per block, it's 0.00000082. I'm sure it'll be taken care of, but it's weird that these problems have been cropping up over the recent past and not before

[Uliss]

18084    2013-05-18 10:21:35    1:29:33    12254159    10046    0.00000945    none    236736

18083    2013-05-18 08:52:02    1:06:34    9193635        7840    0.02110453    none    236729

-------------------

Miracles continue?

[PeBo]

Just give the pool some time to sort it out. Round 18082 is fixed already and was also showing none for me. Now:

Code:

18082	2013-05-18 07:45:28	4:03:45	33732105	2021	0.00148243

Sit back, relax...  

[TiborB]

4:03:45    33732253    none    none    none
wtf?

I'm seeing the same thing, and I haven't turned off my computer tonight at all. The "my account" page has me constantly submitting shares, so I'm not sure what's up with that. Hopefully it gets corrected at some point, 4 hours and all

Here we go again (sigh).  All my miners were happily hashing away at full bore in this long round.  The rounds before it and the next one coming up (if rewarded correctly) would attest to that fact.

Same here :
18082    2013-05-18 07:45:28    4:03:45    33732253    none    none    none    236720    25.24790000    87 confirmations left

Hopefully it will get corrected. Also - at around 3hrs into the round my score got reset - it was almost 0 at 3:15 and by 3:45 it was recovering...

I wonder about the score resetting thing, isn't that part of the weighting system to counter pool hopping?  I notice all the time, my score will drop from the millions to nothing and then begin climbing back up

No, the reset is to prevent scores reaching very large numbers. It is thought that if anyone's score reaches +Inf, the universe will implode. So be grateful for the reset.

There is definitely a difference between reset and renormalization. While reset means zeroing out the scores, renormalization means scaling down the scored and modifying C accordingly. The first is simpler to implement and faster to execute while the second is mathematically more correct. Another options to protect against an arithmetic overflow would be using a logarithmic scale, but that is not how the exp scoring method is implemented.

BTW, I really liked reading http://organofcorti.blogspot.hu/2012/04/41-slushs-pool.html and http://organofcorti.blogspot.hu/2012/09/43-slushs-score-method-and-miner.html and suggest those articles to anyone with doubts and questions about how exp scoring works.

Cheers,
   T