Bitcoin Forum

Hey guys, I've had a few people say that my 400 MH/s hash rate must be a fairy tale, I wanted to post a pic eventually, but now seemed like a good time to post a pic of 1000 MHz core with a private beta phatk kernel , my highest hashrate yet!  

I'm too scared to go over this voltage, I don't want to push it too far


https://i.imgur.com/WJQdp.png


What is the current record for a 5850? (or at least, one on air/reference cooler)

I got my reference 5850 up to 1.200v and 1020 MHz, which yielded 404 MH/s, but the VRM got too hot for my liking, a bit over 100C! :o

According to the official comparison (https://en.bitcoin.it/wiki/Mining_hardware_comparison), 392 MH/s was the previous record, so you did beat it. Good job!

I *think* I would have beat it without the kernel optimization that bitless posted a few days ago (3% speedup for everyone), but that plus the new private 1% opt I'm trying tonight makes it feel like I'm cheating a bit

Even still, I know it would have done 400 MH/s with the old kernel that everyone else was using

Private Beta?  :o

What was the default voltage, OP?
I can only get 805MHz stable. @900 (without voltage increase it's not very stable :/ ~350 Mhash/s though)

I managed to overclock to 1000 Mhz with default voltage, but I'm guessing this isn't particularly stable. I went back down to 775 Mhz again after a few minutes, didn't exactly feel safe to overclock that much when I barely have any experience overclocking. :P

I doubt it would work for more than 5 minutes (or 10 min if you're lucky) at full load without a BSOD. 1 GHz is a a 30% overclock.

Do you know any O/C software which would allow to overclock HD 6850 and HD 6850? I also need voltage change option.
On msi afterburner I can't change voltages.
On ati tray tools - only HD 6850 (my 'main' GPU, used also for gaming) is detected, 5850 isnt.
On trixx - same as ati tray tools.
In CCC I cannot change voltages.

I know I could modify and flash a new BIOS, but I would like to avoid doing that.

How do we know you are using 5850? My miner working at about 285 can solve 6 hashes in under a minute sometimes (luck). Your 408 took 1.25mins to solve 6 hashes. Those numbers translate (I realize luck has a lot to do with it)?

I am able to get around 408 at 970 but my VRMs run too hot (since I have to up the voltage a bit).  I have an old Twin Turbo (non-pro) in place of stock cooling but the stock ref plate isn't very efficient for the VRMs.

http://img52.imageshack.us/img52/3585/403mhash.jpg

make popcorn on the heatsink ?

Haha pretty much.  I'm almost think I could craft a poop chute for the popcorn to fly out into a bowl  :P

I win.
http://i52.tinypic.com/1z522ag.jpg

Nice.  What clocks?

1050/300

1.2375V!  That's pretty serious.

How are you cooling your voltage regulators?

Is this a reference 5850?

Reference, mcw60 block, stock plate w/80mm fan mounted in vrm location.
Crippled one already awaiting new smd capacitor for repair. Didn't have fan mounted on it. Considering piggy backing some capacitors for longevity.

http://i55.tinypic.com/24v673l.jpg

Wow, that's quite impressive.  What kernel were you using for that?  If it wasn't the one that bitless updated several days ago, you can still get +3% performance out of it, and there is a newly optimized one that will be posted in a day or two that will add an additional 1%


Several questions:
1.) how do you check the VRM temp?
2.) could VRM temps be how I killed my brand new Sapphire 5830?  I was running at 1.2v with 100% fanspeed, and it died after two weeks
3.) I noticed you said you were going to repair a broken 5850, how do you know what went wrong with it and what to replace?  Do you think my 5830 could be fixed instead of trying to RMA it?

Thanks for beating the record I thought I set LOL

1. Not all pcb's have vddc phase temp. sensor's. You can always buy an IR gun cheap.
2. Could be, best to visually inspect the board.
3. The red square in the picture I posted shows the missing smd capacitor. It was originally still there cracked along with the solder. Card still works at stock clocks with 1v.

What exactly is wrong with the 5830, symptoms?

Live by the hash, die by the hash. You guys are asking for a fire.

That's a good idea, didn't really think about cooling the back side of the card to attack VRM heat.  I am using 2 120mm fans ziptied to my Twin Turbo over the stock plate; waiting on some thermal tape to come by so I can replace the flakey stock stuff.

Doesn't matter how high the fan speed is, you are driving 20% overcurrent to the GPU chip and overloading the VRM's. Not worth it for the small increase in hashing rate.

Also, you can't RMA oc'ed cards. Or you can, but you'll just end up paying shipping and not getting a new GPU in return because they test cards.
Overclocking and running non-stock voltages voids the warranty.

Please explain the testing manufacture's use to determine whether overclocking has occurred or not.



1.163v to 1.2v is not a 20% increase, more like 3%

That's likely a trade secret to deter RMA fraud. Still, you can google warranties being denied even by just removing stock coolers, or fans (that should just be 3-pin connectors not transmitting data at all).

They can probably measure if OV was applied to the voltage regulator or GPU chip, nobody's going to be sending them working cards so they obv. have tools to inspect dead ones physically


I have some 58xx cards that run at 1.0xxx instead of 1.1xx, was basing it on that.

You based everything off assumption's or wrong information.

You asked for the symptoms - it acted like the computer's PSU was dead, until I removed the 5830, then the computer booted normally.  (the lights would flick on for a second, and every fan blade in the system turned about 1/4 turn and then it went dead again)

Basically I'm assuming the overcurrent protection was kicking in.

As to Jack of Diamonds - if running a card at 1.2v instead of 1.163v with 100% fan really was the sole cause of it dying in 2-3 weeks, then I submit to you that it would have died on me in a month or two at stock voltage.  Also, it's not a "small increase in hashing rate" , the card was producing ~305 MH/s instead of 240-250

Hard to say what's wrong with the symptoms you describe without knowing more or having it in my hands. Take the card apart and visually inspect it closely. If there are no signs of damage then not much you can do unless you learn how to test individual components on the pcb. The fact that the card died so quick means even at stock it would have died within a short amount of time regardless. Poor quality on the manufactures part for that particular card, it happens to any though once in awhile. What means did you use to able extra voltage? TrixX?

Yes, TriXX

Saphhire just RMA'd my 5870 that I toasted. I didn't touch voltages but I certainly OC'd(980) the hell out of it. Best part is that it was only at 70c with .5amp fans blowin on the backs and front of the cards.

These "you can't rma an OC'd xxxx" is just a myth the manufacturers keep putting out because I doubt they CAN test the card. I'm sure it's some chinese guy who throws the pc part into a machine, the machine goes "it's busteded" and he go's ok! And sends a new one back.

It is possible to find out whether the card was overclocked, but as far as I know, only if the voltage(s) were increased.
The problem is that it isn't easy and it's time-consuming, so most people can successfully RMA their cards even after extreme O/C.

This is what I don't understand.  If you overclock your card but you don't touch the voltage and you keep the card cool (say 50*C), what causes it to "toast"?

I know that one of my 5850's goes over 1000MHz on stock voltage (1.0875V) and is able to sustain it for hours (I haven't done a 24 hour test yet).  Everyone else on the forum likes to raise the voltage before bringing the clocks that high and I wonder what I'm missing.

I don't want to kill my card, indeed I'd like it to last a long time, but I feel like it would be wasteful to keep the card at stock clock (725MHz) or even just at 850MHz or so.

 i get 410 MH/s easy at 1.175v

What's that utility you have there that lets you adjust clock speeds and mem speeds? (program that says sapphire in the screengrab) As I could really use the help for my HD 6950 card

If I can't unlock shaders,what would the optimal settings be for my card for mining?

Thank you

Thats Sapphire Trixx (not limited to sapphire cards only) , it allows for overvolting on some odd out cards. The interface isn't as good as afterburner however.

hi

i can not reach over 1000 MHz  on Trixx how do you do to have more?

It's GPU specific. Some can be pushed as high as 1075mhz, and others will start freezing even at 910.

If it's a 'sour GPU' so to speak, then it simply can't go past a certain point no matter how much voltage is applied.
Though usually the premium ones are binned for the top of the food chain (best 5870 cores to the 5970, best 6970 cores to the 6990)

ok with my actual HD 5850 I reah 395MH/s with 1000/300 at 1200mV (80°C)

I wanted to push a little and reach 400MH/s

I used AMDOverdriveCtrl and Catalyst 11.6 (I'm on Linux).

Edit: You can also use aticonfig directly, all that's needed is Catalyst 11.6.  To set the clock rate to 1020 say on your first card use:
aticonfig --adapter=0 --odsc=1020,0

You can probably reach 400MH/s by playing with the software rather than pushing the card further.  I've found setting which will give me 413 MH/s for a 1000MHz 5850 but you might have to pull that back a bit if your rig is not a dedicated miner.  Let me know if you want details.

Do you have the latest miner kernel code?  There have been some improvements in the last few weeks.

I use GUIMiner (last version) with those option :
-v -w 256 VECTORS BFI_INT AGGRESSION=13 FASTLOOP=false

Is there something to change?

It may behoove you to read this thread: http://forum.bitcoin.org/index.php?topic=25860.0 (http://forum.bitcoin.org/index.php?topic=25860.0)
You can use Phoenix with GUIMiner now, I believe, and this modified phatk kernel really does make a difference on my cards, YMMV of course.

The settings you are using are very good for these cards using phatk (I can only assume GUIMiner uses phatk).  Now that I see that these settings are good I ask whether or not you are using SDK 2.1.  For me SDK 2.1 is a fair bit faster than SDK 2.4 (with either Catalyst 11.5 or Catalyst 11.6).  If you are currently using SDK 2.4 then I suggest you give 2.1 a go.

Another thing to consider is whether or not the read-out from GUIMiner is taking a moving average before displaying the hashing speed.  I know phoenix does this by default and I have to pass it '-a 1' to make it stop.  If you manage to remove the moving average filter from the readout then you may see your hash rate jumping up and down, perhaps by a few MH/s.  Try to reduce this jitter as much as possible by closing other running programs and/or reducing desktop effects/resolution if your card is driving the GUI.

Finally, because your clock is as high as 1000MHz you may find you gain another megahash by increasing the RAM frequency to 350 or so.

-w 128 gives me better results with my 5850.

if have better result with -w 256 ^^

If I use SDK 2.1 I have to use to an other version of AMD Catalyst isn't it?

http://www9.picfront.org/token/fNnh/2011/07/06/1970233.png

Sapphire 5850 Xtreme @ Stock Voltage, 1000,300MHz
this is one extremely lucky boy  ;D

Interesting.  When I change to 128 my rate drops by a full 10 MH/s per 5850.  Perhaps there are other settings which need to be changed in tandem to yield an improved rate.

Could you let me know your other settings?  RAM speed, SDK version, Catalyst version, kernel type/version/patches, kernel parameters.

If your comments can improve my rate I'll tip in BTC.

Now that's what I'm talking about!  Very nice!

Can you increase the RAM speed a little without losing stability?  If so you might score another MH/s.  I'm thinking around 350-360 MHz.

With lots of tweaking and testing the best I've gotten from a 1000MHz Sapphire 5850 Xtreme @ stock voltage is 413.3 MH/s (fluctuating between 413.1 and 413.5).  If you can beat that then please let me know for some BTC.

Can you push the core clock any higher than 1000 MHz?  Are you bound by Trixx or by stability?

I'm running linux headless, so I use aticonfig and adjust my base-clocks at bios level.

I will try to do some optimizations, I will let you know ;)

can you told me how do you reach 413MH/s?

I will but I'll give whopper a day or two to try and beat me and claim some BTC.  It's not the same challenge if I give him all of the information I have.

I will say that I am using Linux headless and driving everything with ssh so whopper and I are on a level playing field.  I don't know about adjusting base-clocks at bios level so I hope to learn some more myself with this little challenge.

You me to pay you to know how I can gain 13 MH/s  :o

All right if you want but for now I have only 0.00006084 BTC it is all I can give...

No.  I simply want to give whopper some time to try and improve on my rate.

I can only get it up to 412.0 MH/s  ::)
not incorporating the recently made kernel improvements posted at this board

Can you get it higher with the kernel improvements?  No improvements have helped my rate since the MA tweak.  Feel free to use the recent improvements (from earlier today) to beat 413.3 MH/s.  If you have 412.0 MH/s then I'm hopeful you'll beat it and I'll find out how to improve my rates.

Looks like we need SDK 2.4 for this, but I will stay with 2.1 at the moment

You and I are clearly of one mind.

As you probably know from the latest patch thread
http://forum.bitcoin.org/index.php?topic=25860.40
I've already tested SDK 2.4 a little and found that at high core clock speeds SDK 2.1 is still slightly superior for the Sapphire HD5850 Xtreme.

If you are done with trying then let me know and I'll fill gellimac in on the details of my configuration.

Hehe I guess so  ;D

Maybe it will change with the new SDK 2.5, so we can squeeze out a little more Hashes  ;)

Yes, I look forward to when that is released.  As far as I can tell there's a early version out but it's only usable on Windows.

By the way, how are you using phatk?  Are you driving it through phoenix?  If so be aware that, by default, the readout is an average of the last 10 rates and this can be disabled with the flag '-a 1'.  I find the moving average is only useful if your MH/s is leaping around by more than 1 MH/s say which shouldn't be the case here.

I'm running poclbm

For me, it's the opposite. I lose 10 MH/s if I set -w 256.

I'm actually using Phoenix, so my settings look a bit different:

Ok.  In which case '-a 1' does something entirely different and no averages taken by default.

What is your RAM clock and AGGRESSION at the moment?

Also, you mention you are running headless.  I assume you are therefore using custom '/etc/X11/xorg.conf', '/home/user/.xinitrc', et cetera, to activate the cards and give the drivers and poclbm something to point at but without asking the cards to drive any kind of desktop.  If so then I must admit I'm having some trouble getting xorg.conf just right.  I've got it to a point where if I plug a monitor into one of the cards there will be no signal but I'm not convinced it's not doing something anyway.  Any tips would be appreciated and I'll happily share my files if you're interested.

Very interesting.  On my system (980MHz core with 360MHz ram):

python phoenix.py -u http://<user>:<pass>@<host>:<port>/ -a 1 -q 1 -k poclbm VECTORS BFI_INT AGGRESSION=11 WORKSIZE=256 DEVICE=1
401 MH/s (+/- 2 MH/s)

python phoenix.py -u http://<user>:<pass>@<host>:<port>/ -a 1 -q 1 -k poclbm VECTORS BFI_INT AGGRESSION=11 WORKSIZE=128 DEVICE=1

391 MH/s (+/- 2MH/s)

Perhaps this is something to do with RAM clock speed.  I'll investigate.

Edit: Before getting to the RAM I noticed that this playing around with poclbm and AGGRESSION has frozen my card.  This is the second time today!  I really have to pull it back to 975 MHz unfortunately.  I'll do my testing on my good core (still running well after 36 hours at 1020 MHz).

I dont use the poclbm kernel, I'm using poclbm miner with phatk kernel ;) Run "git pull", it was included recently
As of the clocks, I'm on 1000,350 now. There is no aggression in poclbm, I use -f0 which is the hardest setting, -w256 and -v

Running headless means I simply run "xinit" which spawns an empty desk with one xterm window in case of ubuntu server. I didn't change anything config related, except for having "export DISPLAY=:0" included in my bashrc file.

Ah.  I see.  I was confused because phoenix comes with two kernels, called poclbm and phatk.  I guess poclbm also refers to the front end.  I'll also try some 'harder' aggression settings again (was using AGGRESSION=13 until recently).

I have only one possible tip:  Add a file '.xinitrc' to your home folder and simply put in the one line 'cat'.  This will simply wait for input and could be considered a slight improvement on running an xterm.  This almost certainly won't affect hashing rate but might affect stability a tiny bit.  Either way, it seems neater to me.

Are you doing anything special in xorg.conf?  I don't really know what I'm doing in this file.  I don't think I've made an impact on jitter, hash-rate, or stability yet but it's hard to tell with stability.

Our setups are very similar, so much so that the missing 1.3 MH/s puzzles me.  I guess this could be a difference in the way phoenix and poclbm calculates the value to be displayed from the data fed to them by phatk.  the 10 MHz RAM difference has practically no effect on the hashing speed and varying the AGGRESSION among the 'hard' levels (13 and up) only makes a difference of about 0.6 MH/s.

I simply let aticonfig generate the xorg.conf for me, nothing changed there. (aticonfig --initial -f --adapter=all)

I guess that little extra MH/s isnt really important. At this point it would be better to look for some real-world effecting factors, including better cooling, network latency (better pools) to reach a higher real hashrate

Absolutely.  I'm just following a half-way scientific approach, fixing some factors and then trying to maximise others (in this case hash rate).  Besides, maxing the hash rate is fun.

When my power meter finally arrives I'll be looking very carefully at power consumption and may try to max MH/J before returning to a compromise which takes into account the very high value of mined BTC versus the value of the consumed power where I live.  I've planned ahead to this end a little by investing in a fairly effinient (90%) and exceptionally quiet power supply.  Undervolting is a consideration here too.

As for the pools, I've always been a solo miner and my circumstances are such that there is no reason for me to change.  A friend of mine is pool mining in Windows with BTC guild and has had much trouble in the last two weeks.  multiminer solves some of these issues but he's not tried it yet (and I have no reason to).

Another very important factor for me is noise.  My miner has to live in my room and I value my sanity far more than the BTC it produces.  I replaced the stock coolers on my cards with Zalman VF3000As (Caution: these coolers are not designed for these cards).  Combining this with no overvolting and I'm able to run these cards at high clocks and with 40% fans (which are very quiet).  When the stock coolers were still attached I needed to run them at 80% for reasonable temperatures and this was very noisy indeed (be wary of 100% on the stock coolers, many people have had problems with the fans dying on these cards).

Greetings, 5850 owners. ;D I am dropping in to tell you the true tale of the non-reference XFX 5850 (this one: http://www.newegg.com/Product/Product.aspx?Item=N82E16814150477&cm_re=xfx_5850-_-14-150-477-_-Product (http://www.newegg.com/Product/Product.aspx?Item=N82E16814150477&cm_re=xfx_5850-_-14-150-477-_-Product)) that I purchased for $100 on Craigslist. The first night I ran it, the fan broke so that it was no longer speed-adjustable, and stayed at 1100 RPM at all times. This made it quite unsuitable for mining, of course. I communicated with XFX's customer services representatives several times, but was unable to obtain the necessary RMA. I ended up buying the Arctic Cooling Twin Turbo Pro, which fits perfectly and works like.....well, it just works. So well does it work, in fact, that I slowly increased the core clock speed, and it did not freeze until 1035 MHz (410 MH/s)! Right now, I am running it at 1000 MHz (398 MH/s), just to be safe. Note that this is at STOCK VOLTAGE, and has not exceeded 64 degrees C at any time! I think I got a lucky card. :o

Miner Settings: Phoenix, phatk (latest optimized version), worksize=128, aggression=11

Wow!  My best card froze at 1025 MHz.  Perhaps the XFX cards are a bit better at overclocking than the Sapphire Xtremes, perhaps you have a lucky card, perhaps you have a more stability inducing configuration than my own.  I'm assuming your XFX's stock voltage is indeed 1.0875V.  I'd be very interested to learn about your configuration, are you using Windows or Linux?

Inspired by this I've clocked my good card up to 1025 MHz again to see if it lasts longer this time.  I wonder if keeping the card very cool affects stability, I'll keep my card under 50*C and see if it can hold 1025 MHz for 24 hours.  I'm only 5 mins in but already 423.7 MH/s feels good :).  How long did your card last at 1030MHz before you decided to increase the clock?

Here is that WORKSIZE=128 again.  I simply cannot figure this out; WORKSIZE=256 is always about 10 MH/s faster for me no matter what my other settings are.  I guess there's an outside chance that this is a Windows/Linux difference.  If you check up the thread you'll find whopper and I competing for the highest hash rate for a 1000MHz card and ended up agreeing that 412-413 is currently achieveable.  Perhaps you can give our settings a go and realise 410MH/s once more!

At the moment I'm using:
RAM=360 MHz, SDK 2.1, Catalyst 11.6, basic phatk with just the MA patch (3%), AGGRESSION=16, WORKSIZE=256, VECTORS, BFI_INT, FASTLOOP=false.

although swapping out SDK 2.1 for SDK 2.4 and 'basic phatk with MA patch' for 'latest phatk patches' is only 0.6 MH/s slower for me at these core speeds and may be much easier for you to try.

what agression is for?

can we put AGRESSION 999 for exemple?

Aggression tells the miner how much work you want to do per OpenCL instruction.  The formula is nonces = 2^(16 + AGGRESSION).

So AGGRESION=14 will be about 1'000'000'000 nonces per OpenCL instruction (which I believe takes about 2.5 seconds at 1GHz core on a 5850 - hence the rather slow rate updates).  AGGRESSION 15 will be twice as much work per OpenCL instruction and your rate will change once every 5 seconds or so.  I chose AGGRESSION=16 because 2^32 is a nice round number and I get about 0.2 MH/s compared to AGGRESSION=14.  For a dedicated rig I don't think increasing the AGGRESSION to 16 is going to hurt (might even help) stability but for a miner you want to use for other tasks it is completely ridiculous and 11 or below is more suitable.

I use AGGRESSION to reduce jitter (reported hash rate variance).

With help from Diapolo I've modified the latest phatk kernel to work with SDK 2.1 and the result is another 2.1 MH/s for the 1GHz 5850s.  Now I'm able to achieve 415.4 (+/- 0.1) MH/s.  See http://forum.bitcoin.org/index.php?topic=25860 for details.

I had momentarily forgotten about changing the worksize, thanks for the reminder! :) Changed to 256, and now hashing at 405 MH/s, will do some more tweaking later today.

I got my 5850 Toxic 1Gb to 418Mhash @ 1055 Core, 375 Mem and 1.25V.
That was on XP though and after I changed to W7 the diriver started to crash at 1025 Core..

Note that at that voltage it's less efficient than a 5870 but I just wanted to test how high it could go :3 (Temps were never above 65C though, Vapor-x <3)

No worries.  I hope you manage to get a little more out of that card with tips mentioned on this thread.

I'm currently trying to improve further.  I believe it's possible to get quite a bit more because Tx2000 is able to get a staggering 413 MH/s at only 970 MHz!  Granted, this is on Windows, and I have a suspicion that he's using Catalyst 11.7, but if he's not then there's surely some trick he's using for the extra 12 MH/s.

My good card has been running at 1025 MHz for over 12 hours now and is perfectly fine.  I think being kept very cool is good for it's stability (at these clocks 60*C is far too high).  I'm getting impatient so time to go to 1030 MHz.

Yes, increasing voltage can hurt power consumption in a big way.  There's certainly room for improvement in software though.  A 1055 core should be able to manage more than 430 MH/s with a little tweaking.

I'm impressed with the cooling though.  65*C on a 5850 1055/375@1.25V is very good.

I always have highest speeds with 300 or 302 MHz ram, but yes the worksize and ram speed are related.  Someone actually made a graph if you search for the original thread that said "Lowering ram speed increases hashing rate" or similar, go all the way to the last post and work backwards to find the graph.

Well teukon,

I grew a pair and tried the magic 1000 core and I was successful.  But it wasn't enough to satiate my curiousity so I tried a wee bit more and here are my results.


http://img836.imageshack.us/img836/2845/1010429.jpg


While 1000 was stable at 1.149v, I had to up to 1.162v to hit 1010MHz on the core.  The VRMs were a bit scary though, nearing 115c.  This was only for a few minutes too; probably won't feel comfortable until I get that new thermal stuff in.


429.7   ;D

Cool.  Thanks for this.  I assume the 429.7 MH/s figure is for 1010MHz and that your 1000MHz figure is therefore more like 425.6 MH/s.  Even so this is a huge improvement on my 415.5 MH/s so well done indeed.

I'm fairly certain than Catalyst 11.4 is the magic ticket here but as a Linux user I'm relying on Catalyst 11.6 to overclock my cards past 900 MHz.  More power to Windows users!

I anyone knows how to clock past the BIOS limits in Linux while using Catalyst 11.4 then let me know for some BTC (and no, WINE/DOS/Windows/Bios Flash do not count).

Yea, it was pretty much around that.  I may try some alternate kernels to see if I can't milk a bit more from it but I think it's pretty much topped out.

Do you mean this graph?
http://forum.bitcoin.org/index.php?topic=4292.msg148234#msg148234

I experience similar changes to my hashing rate as I vary the RAM.  The card type and core clock speeds are also important factors.  For the 1GHz 5850 I found 350-360 MHz to be optimal (worksize 256).

I think this is looking pretty much maximal but could be wrong.  As you're using Windows perhaps you could try the Catalyst 11.7 preview if you haven't already.

Guess it's not maximal  :D

Tried diapolo's modified kernel as you mentioned and I was able to push it a bit more through phoenix miner


432.5

http://img823.imageshack.us/img823/7314/432fi.jpg



I may try some other driver versions but seems like this is a good mix so far

Woah! I thought you were already using his latest kernel!  I can't recall whether or not this was SDK 2.1 or SDK 2.4 but I'm guessing the former (I'm fairly sure it would be the faster of the two here).

I guess this would be 428.3 MH/s at 1GHz.  Surely I wouldn't gain 12.8 MH/s just from going from Catalyst 11.6 to Catalyst 11.4.  Do you think that with Catalyst 11.6 you would sink all the way to 415.5 MH/s? surely not.  I guess there's something else going on here allowing you such high hash rates but can't think what it would be.  If you have any ideas as to what I'm missing then the 0.4 BTC I offered is still available.  Hopefully Catalyst 11.7 will perform better than 11.6 and I very much look forward to that.

For now I'll focus on making my cards stable at high clock rates.  At stock volts I seem to be ok at 980 MHz (406.9 MH/s) and 1030 MHz (428.0 MH/s) on my two cards.  Temperature seems to make a big difference so I'm glad I've got good after-market coolers.  I had some fun putting the fans to maximum and opening the door and window to see how far my cards would go.  My good card held it's own for 10 minutes at 1045 MHz (stock voltage).  When I increased it to 1050 MHz it crashed instantly (42*C).  My card is definitely not stable at these higher clock rates.

Yea I'm not sure what it is to be honest, some hoodoo magic or something.  Certainly nothing I've done aside from overclock.  It's SDK 2.4; I'll see if I can try SDK 2.1 but I'm weary of the issues of switching SDKs.  I'll keep you updated as I try different things.

I wasn't aware that switching SDKs was difficult.  On Linux you are able to install both simultaneously and switching between them can be as easy as changing the destination of a symbolic link.  It's about as difficult as changing a kernel.  I'd quite like to find a way of switching between drivers easily on Linux but at the moment I can't find a way of doing it without a reboot!  If it's difficult/dangerous/unpredictable then don't worry.  I think the improvement will only be 2-3 MH/s or so.

Actually, thinking about the figures, it's quite possible that Catalyst 11.4 works better with SDK 2.4 and Catalyst 11.6 works better with SDK 2.1.  This could explain the 12.8 MH/s difference.  If I'm right about this you may actually lose 3-4 MH/s or so with SDK 2.1.  I might test this later but I'm currently running testing the stability of my cards.

how did you get that techpowerup gputool to display the data from your card, when i run it im told my clocks are at 14mhz and no temp info, want to get the vrm temps up, i grow concerned at my 1.243v overclock, it seems stable enough but id rather be sure

another question for the clockers out there, im using trixx and it maxes out at 1000mhz, the app doesnt seem let me go any higher. however at my reccomendation a friend bought the same 5850 cards as i have, at first he too could only clock to 1000mhz but then it suddenly allowed him to go higher (currently 1032)  does anyone know how i can make trixx take more mhz for me too, he tells me he didnt do anything at all to make it happen and i need it, i know itll give me more and i cant let him beat me, i just cant!

EDIT

for some reason a restart brought mine into working too, if i push it further it will go all the way to 1050mhz (428MH/s) but not stable enough and i fear for the vrm temps as the monitors ive found dont seem to work. i think modding some cooling for the vrms is the answer if i want more out of it so thats this weekends task, but as always it has to be applied fast as i hate downtime, currently running 1015mhz getting 415MH/s

new record?

http://img7.imageshack.us/img7/7902/sansrewb.png (http://img7.imageshack.us/i/sansrewb.png/)

Uploaded with ImageShack.us (http://imageshack.us)

I cant get over 1000mhz one 1 5850 , but I've got 3 of them at 915hmhz...

I can't change the voltage of any of them :(

you know what, i hate you :P but i shall persevere, once i cool those vrms ill be pushing more

you guys don't actually get it. why would you want to invest like 50% more power consumption for a mere 30 mhash/s? is this a benchmark? try benchmarking actual mining efficiency.

Unfortunately not but that is a damn good hash rate but then 1055 MHz is a very high clock.  I feel that it should be possible to get more MH/s at that clock rate but is your card stable at that rate?  How long did you run the miner for?

I decided to overvolt my good 5850 from stock (1.0875V) to 1.25V and see what it could do.  It managed to hold 1110 MHz under load for 3 hours (finally dying when I got adventurous and tried to start mining on my other card) and managed 461.1 MH/s with very few stales.  I posted about this here (http://forum.bitcoin.org/index.php?topic=27468.msg346776#msg346776).  You can get a screenshot here (http://img195.imageshack.us/img195/339/72screenshot2.png) and I have pictures of my rig here (http://forum.bitcoin.org/index.php?topic=7216.msg345510#msg345510).  Admittedly, I have the advantages of Linux without any GUI, very good aftermarket coolers, and lowish ambient temperatures where I am (peaking at 21*C in the middle of the day).

I was proud to have beaten the highest claimed 5870 record but it wasn't long until scifimike12 took his water cooled 5870s to an astonishing 501 MH/s (which ran for an hour before he turned it down).

Needless to say, I do not run at 461.1 MH/s all the time.  I'm not very concerned about the voltage or heat and I'm not convinced it's not profitable but there was simply far to much noise.  428 MH/s at stock volts and 50% fans is ok for now and I'm going to undervolt my cards and turn down the fans once my power meter finally arrives.

Without changing the voltage it's hard to get a 5850 to 1GHz.  The most important factors seem to be system stability (don't drive a GUI and mine with the same card) and low temperature (certainly less than 60*C).

Just incase your inability to increase the voltage on your cards is a software issue, I'm able to change the voltage by using Catalyst 11.6 and AMDOverdriveCtrl (editing the .odrv profile for higher voltage, a GUI slider may not go beyond 1.0875V).  This is on Linux but I can't imagine it's harder on Windows.  No BIOS flashing or Trixx should be required.  Older versions of Catalyst will suffice if you only want to decrease your voltage.  I cannot make my voltage increase beyond 1.3V with this method.

are you sure increase for 0.2V give 50% more power consumption?


are you talking about a 5870?
I am on a 5850. I run it for 10min it was limit stable I touch the mouse => BSOD. But now I mine at 426MH/s with a 1040MHz clock stable.

well, around that, yes. maybe 30%.

The more you add the quicker it adds up (exponentially)

So yes, at the levels shown in this thread, 0.2V increments will result in 20-30% bigger power draws

At 1.3V or above you are already crossing a negative threshold
(you will be paying more for electricity than you will earn from the extra few dozen mhash/s

But as long as people just want to see a 'big number' without caring how much electricity is required to produce it, go ahead..
Or the fact the VRM will die in a month.

No, I'm talking about a 5850, in particular a Sapphire HD5850 Xtreme.  This particular card is built with the 40nm process and looks like a 6950 but has the essential structure of a standard 5850.  It will go to 1030 MHz on stock voltage provided I keep it cool enough and will run happily at that speed for days at a time without freezing.  To get more than 1030 MHz I had to increase the voltage and the increase from stock (1.0875V) to 1.25V was sufficient to make 1110 semi-stable (it only lasted 3 hours).  Indeed, the card went to 1130 MHz but was not stable (lasted only 90 seconds).

I'm sure the low temperature (about 55*C) and lack of a GUI made a difference to my maximum stable overclock.  In particular, I can't tell you whether or not mouse movement would have crashed the card because there was no mouse to move.  I'm sure if I was mining in a GUI environment I'd have to drop 50 MHz off of my clock to get stability, and another 40 MHz if the GPU was over 70*C.

I posted on the 5870 thread rather than here because they were interested in maximum speed whereas here we were more interested at maximum speed at 1GHz core.

I do not deny the diminishing returns of increasing voltage but I don't understand why increasing voltage will increase power consumption exponentially and perhaps you can explain why?

If you increase the voltage (assuming no change in electrical resistance of the cards circuitry) then the current will increase in proportion (Voltage = Current * Resistance) and therefore the power consumption will increase quadratically (Power = Voltage * Current), not exponentially!  In real life the increased current will increases the temperature of the circuitry and therefore the electrical resistance so the growth of power with respect to voltage will actually be a little slower than quadratic.  Increasing clock rate will increase power consumption approximately linearly as far as I understand so if you increase both voltage and clock rate by 10% you might draw around 23% more power.

I simply cannot see how voltage and power can have an exponential relationship.

Alright, I'll say this one more time.

Running your cards at high voltages or temperatures for extended periods will likely cause the cards to die prematurely.  These records should be taken as a measure of potential and can be useful for comparison but for most circumstances will not be profitable for long periods (considering power costs, card depriciation, or both).

If you are concerned only with maximum profit then this thread is not for you (unless you have no power or hardware costs and limited time) but you might hope to reap the benefits of the work done here in the long run.

When my power meter finally arrives I'll be undervolting my cards and looking for high MH/J scores then.  I'll make a thread for this when I'm ready and perhaps you'll find it more useful but, just as here, it will be for research and NOT recommended for maximum profit.

i only wanted to point out how inefficient overvolting is. i'm not sure about the actual increase in power given a voltage increase, but even if it was linear, your core clocks won't scale linear and neither will your hashrates. so if the consumption increases quadratic, you're looking at an even greater loss. i'm pretty sure if people could monitor consumption without much effort, this would be the benchmark everyone would go for. too bad you can't measure consumption without buying tools, so we're stuck with benchmarks of raw hashing power. this is insane, though. if you like benchmarking and comparing your results, go for something else than mining. go do some weird 3dmark benchmarking sessions with LN2 and quad-fire etc.

I see what you're saying.  I don't want to mislead people into thinking that overclocking to the extreme is a good idea financially and I agree that many people would be interested primarily in profit given the tools (not all but most).  Maximising profit is far from easy but my feeling is that a mild undervolt is best in this regard (accounting for the high value of BTC compared to power costs for GPU miners and the cost/depreciation of the cards themselves).

I personally rather like using mining for benchmarking cards simply because it shows off the OpenCL power that the cards have.  In the long term I will want to use my cards for tasks such as Folding@Home, Mandelbrot fractal generation, Strong Go/Weiqi/Baduk engine, and am not interested in 3D acceleration, video acceleration, desktop effects, or GUI at all.  Mining is a good benchmark for the former because it is a test of low memory/bandwidth raw OpenCL processing which I can run without a GUI and supports a rather cool crypto currency.  Naturally, for Folding@Home high efficiency is most important and I'll get to that but for Mandelbrot and Go it's all about performance.

Are you saying the Sapphire 5850 Xtreme model is built on a smaller process (nanometers) than the normal 5850's ?  Because the stock voltage you listed for it is the same as my original reference 5850.

It's crazy how it's totally luck of the draw with GPU chips.  I have two reference 5850s, one that still has the stock red blower heatsink on it and will do 400+ MH/s, and one with a crazy aftermarket heatsink on it that keeps it at 55c, RAMsinks, and air blowing over the ram and vrm's, and it won't go any over 870 MHz nor will it accept any overvolting whatsoever (instant lockup)


Also as to the extra voltage being linear or exponential, according to ohm's law it seems like it should be linear, but I have also read elsewhere that power increases with the square of the voltage added (maybe due to excess heat being lost through the transistors?), so I don't know which is correct.  What I do know however, is once you hit the ceiling clock with stock voltage and begin adding extra voltage to go higher, it quickly gets into diminishing returns and the extra voltage needed goes exponential.


Too bad they stopped making 5850's and 5870's, they really were powerful little cards even by today's standards!

I'm sorry, I've been assuming that the 5xxx cards and 6xxx cards use a different size process.  A quick look online suggests they both use a 40nm process but I can't be sure.  What I do know is that the Sapphire HD5850 Xtreme cards are quite a bit shorter than the reference 5850s and are built more like the 6xxx family.

I appreciate that I've been extremely lucky with my cards.  Perhaps the manufacturer has a lot to do with it.  I do know that when running a GUI and with stock cooling the maximum stable clocks of the cards were around 900 MHz but I didn't do any testing after removing the GUI and replacing the coolers so I can't say which of these two had the most impact.  A card that only goes to 870 MHz even with good cooling, and doesn't accept even a small overvoltage, sounds like a very bad card to me.  I assume that if you undervolt the card you lose stability at 870 MHz.

Ohm's law is "Voltage = Current * Resistance" but there are two caveats to mention here.

1) Current is not Power.  Check out Joule's Law to see that both heat and power are proportional to the square of the current.
2) As temperature increases, resistance increases, so current will grow slower than linearly as voltage is increased.

I agree that the maximum stable clock seems to grow logarithmically as voltage is increased (and so voltage must be increased exponentially to increase maximum stable clock linearly).  However, just as 1 + x is a good approximation for exp(x) for x close to 0, small changes in voltage close to stock voltage will affect maximum clock rate approximately linearly.

Also, I'm fairly sure that increasing the clock rate increases the power consumption linearly.  Combining this with the above, if you increase both your voltage and core clock by 10% you should expect the power consumption (of the core) to increase by 33.1% (a cubic relationship).  I think some people choose to do this is because, for them, 10% more MH/s is actually more valuable than 33.1% more power + shorter card life.  I'm sure some people do it because they think it's profitable when it is not for them.  Some people do it because they are overclockers and are actually slightly more interested in configuring, pushing, and comparing cards than the bitcoins.  These people will invest in after-market coolers or water cooling (not profitable), do copious testing (more profitable to just get reasonable settings and let the card run), and will post what they learn online (not profitable as it increases the network hashrate).

I think you can still get the Sapphire HD5850 Xtremes (I think they were released even after some of the 6xxx family!)  They cost about 130 GBP each.  I can't vouch for their power in gaming but they certainly fare well with the 6xxx family for mining.  Indeed, I believe the 5970 is the fastest mining card (I'm fairly sure it's capable of 1GH/s underwater and I don't think the 6990 can do that) so the 5xxx family is doing something right.

hey gang ! I got 5830 here and am reaching 330s with sapphire Xtreme stk voltage and 1040  370 mems @ 78 C stable 24/7

Want to go buy 5850 or 5870 but am a bit scared due to not knowing whether I can OC them... So if I stk to sapphire Xtreme 5850 or saphire 5870 will I be ok ?

Thanks ! yup ... you were right pushing the envelope is my game ;) Diablo you rule for your mods to phatk ;)

Cheers,
Al

Private communication with the original builders of my rig, BlightyCoin, revealed that some Sapphire 5850 Xtremes can be overvolted to 1.3V on Linux with Catalyst 11.6 and AMDOverdriveCtrl but some will not go beyond 1.0875V.  Given that my cards do go beyond and their cards do not and because I believe I ordered one of the first rigs they assembled, I think that newer BIOSes will limit your overclocking potential in Linux.  I'd be very worried about overclocking/overvolting limits in your position as they were the greatest obstacle to configuring my rig.

If you are using Windows then you have Trixx but I hear there are a number of annoying limitations with this too so beware.

All I can advise is to read everything you can find on the subject, buying the card is almost certainly going to be a gamble but you can atleast get the odds in your favour.

Also, if you cannot overclock you can always undervolt instead so that the rig isn't completely useless but I guess this is little compensation.  My good card is undervolted by 0.1V and it still clocks 170MHz above stock so the untapped potential could be huge.

Best of luck!

hhmmm ... thanks for the warnings ! sooo .... you think it's safer to buy 5870 instead of 5850 or are they both suffering from same gambling problem ;)
yup trixx is my favorite way for I just get along better with it ... MSI Afterburner got a littlebit on my tits ... so I hit the button and rid myself of it !

ok ... I know from a buddy of mine that 6950 with 2 gb ram can be unlocked shaderwise for sure ... since that's going at about the same price as a 5870 where I am, would that be a safer bet ? I read here https://en.bitcoin.it/wiki/Mining_hardware_comparison that 6950 can also exceed 430 Mhas/s and that'll be more energy efficient as my 5830 at 330 despite double the price ... and only 30% or so gain in performance . what do you think ?

I know practically nothing about overclocking the 5870s although I did see a screenshot of someone using Windows who had their 5870 up to 1.35V and 1135 MHz so they were able to break my current 1.3V barrier.  Not knowing makes it a gamble, particularly with Linux as Sapphire don't seem to be aware that Linux exists.

If you can unlock the shaders then a 6950 looks like a good plan.  Unlocked it might be competition for a 5870 (the 5870s are incredible miners).  I still can't say anything about your ability to overclock or overvolt but if you're using Windows then the risk is small.

Also, if you are looking for energy efficiency then 430 MH/s might be a bit much for an unlocked 6950.  I see from the linked page that this is achieved by taking the card to 1.275V which is a pretty serious overvolt.  Remember that power increases quadratically with voltage and linearly with clock rate so a 20% overvolt on both should give you about a 73% power consumption increase (of the gpu - not necessarily the whole card).

with the new Version of TRIXX it is possible to OC more than one card

maybe it is thanks to the mail I send them to do a version for multi GPU support  :D

My record is 450MH/s but the GPU is running at about 95C even with an after market cooler, and is extremely unstable, so is not actually usable for mining more than about 30 secs.

way to necro :P and what clock was you running it @ for 95c on a after market cooler?

I'm tempted to bring my ref 5850 into my outside computer @ 100% fan and shatter all previous records. ... But it would require dismembering my dedicated miners to get at it.

http://www.nogleg.com/archive/5870.jpg

there's a 5870

that's like, hmm, 3 or 4 days of stability i think (rejects are cause of p2pool)

it can actually go higher, cause that's only at 1.212

all day long, oh yeah

well, but not much point.  since it takes like 95-100 amps.   i run it at 0.95v now

Wow!  I'm tempted back to the mining hardware forum for the first time in 6 months by the news of the first ASIC system and find someone has revived one of the threads I was most active in.

I have nothing relevant to add to the content of this thread anymore (I sold all my mining hardware late last year and bought at ton of BTC with the proceeds) but wish you all luck with record breaking attempts.

well, it was resurrected by someone on jan 27th, so i just posted some screenshot i had from like a month ago

the truth is, those #'s on mining comparisons web page are all easily broken, heh

Are u kidding me!
I can goo even higher, but as u can see I crashed once cause PSU is too week, than I stopped the other two cards, but I'm doing it thru teamviewer just to show you and I'm afraid it might freeze. When I'm there I'll post a screen with 1080/300MHz.
Just to clear out this is a Sapphire 5850 referent

http://www.picvalley.net/u/2962/21411439112302265341359766116wMM6MnLjQo6Lmf8qSb6K.JPG

1060MHz and you can't get past 437MH/s? What driver/SDK are you using? I'd bet you it's a software thing at this point.

Dude This is a 5850!
Didn't u read the OP title - "Record hashrate for a 5850!"
Maybe its my fault, I'll edit my post.

Ah yes you're right. I was looking up at zvs up there with his 5870, and thinking you should be getting similar numbers as him. I noticed you were about 15% lower than him, but that makes sense now.

Still, that is quite an OC!

I guess there have been some improvements over the last few months, based on king_pin breaking gellimac's old 434.4 MH/s record with relative ease.

I'm curious:

• How far can you take your card king_pin?
• Has anyone beaten my personal best of 461.1 MH/s (https://bitcointalk.org/index.php?topic=24949.msg361735#msg361735)?
• What's the current record?

For 24/7 I can't go more than 1080 at 1174mV, and that was when it was -10'C here, now it is +10'C so I doubt that I can do it stably.
I am curious too, I might try 1.2V but I'm a bit scared, and I doubt I'll get more than 1100MHz, and my PSU GX750 is fucked up it gives 11.8V with just 1 card.
I should try it on the AX1200 but i run 5870s on it.

btw all of this is useless now, because of efficiency - I run my 5850 at 1V and around 900/300MHz depending on the card

Nope dude I'm from Bulgaria, and next week here there will be temperatures of -7'C to -10'C so expect some more printscreens from me :)
No nitrogen just cold air. There are vids of my rigs on YouTube:

http://www.youtube.com/watch?v=OCTDN05LKao

http://www.youtube.com/watch?v=WX6uvrPuUYs

I'm using SDK 2.4 maybe SDK 2.1 may give me 1-2 more MH/s and I have a week CPU, so 1-2 MH/s lost there too.

For the record.  My rate of 461.1 MH/s was also set with air cooling.  I went as far as replacing the standard fans with some decent custom ones but no liquid was involved.

king_pin:  Hash rate aside, 1080 MHz at only 1.174V is impressive! (damned impressive if I'm correctly gathering that this card is simultaneously driving the desktop).  How stable is that?  The big advantage I had over many other miners was having no GUI (my screenshots are just ssh sessions in terminals on another machine) and I bet my good card would have struggled to do that if at all.

Honestly, I was uncomfortable taking my card beyond 1.15 V but decided to try it briefly for the purposes of setting records and learning more about my cards.  I eventually set my card to 0.9875V, 890 MHz core, 300 MHz RAM, which was giving me 373.1 MH/s, and got more involved with pools (about a month before Namecoin merged mining came out).  I preferred this as it was much kinder to my cards and was more profitable (with my second card set a little slower my rig drew 300W at the wall, so 2.43 MH/J).  I never had a crash with these settings.

No no she isn't the desktop one :)
But I have one 5850 which is the desktop card and is running happily at 985/300/1049mV so I think she is my most impressive card. :)

Ah good.  I was prepared to grill you for details on the 1080 MHz setup, but with no desktop generating noise and distraction like a 3-year-old on holiday, it's a plausible rate for a good card given the overvolt.

As for the desktop card: running two differently demanding tasks stably, 24/7, with a serious overclock is already a considerable achievement.  Doing this with a mild undervolt?... :D.  She's a gem; long may she live.

hmm, i've never owned a 5850..

i messed around with that 5870 some more since it actually got cold again here for a bit...  i started running into problems not w/ stability but with "hardware errors",   maybe 2 or 3%?   this around 1120 clock

have anyone here ever use this http://www.biostar.com.tw/app/en/vga/introduction.php?S_ID=103#spec
what d'ya think about this card?

Rumor has it all started when he resurrected a thread that dates back to the big bang, that then sucked him in ripped his entropy into pieces, spreading clones of him all over time and space. All of them doing the same exact thing, resurrecting old threads, and all without him knowing why or even where he even came from nor who he is.

wtf did i just read ???

A mix of the script of a doctor who episode, portal 2, and my imagination. :P