Hmm, propably yes, you should set -r 1 - it will execute my reboot.sh magic script which acts like FORCE REBOOT (it will work even if gpu will be hanged) SRR yes, me and my friend are working hard on it  We are pretty ok with schedule. We changed a little layout of the board to be more user friendly and put SimpleRigResetter name on it. We received yesterday new printed board. Now we will put all transistors on it and programm it with our firmware. If it will have no mistakes and no bugs, then we will order big quantity of printed boards and start production process So the date of shipping is still 15-30 January I didnt want to release preprder until we are certain that the board has no bugs and we also will show how it will look like in changed layout. Anyway next 7-10 days i will open preorders most propably and start selling first batch Here is how it looks like printed board without elements of your PRODUCTION UNIT Pico IP Great stuff  |
|
|
|
I did some tweaks on both of my Tahiti (HD7950) cards and I get around 200H/s per card. Makes 400H/s in total. I'm pretty happy with these results.
Thanks Claymore for this great program. Keep up the good work!
can you post your tweaks? I did some overclocking to make sure both cards would yield the same results in Catalyst 15.11. I increased the GPU and memory clock of one of the cards. As you can see this gives me 200H/s per card. check : http://imgur.com/a/7D1EtI've been thinking of trying two more things to increase the H/s: - updating catalyst to the latest version - Trying this: http://www.overclock.net/t/1554360/tahiti-memory-timings-patch-for-hynix-vramIt seems there is a patch for the Tahiti Memory Timings which seems to increase the speed. Now, this is all a bit technical for me and I don't really know what it does exactly, but the results look promising. This only works for the Tahiti cards though (HD7950), like mine. Yes, the mem timings patch does work indeed, but.. It depends on your card whether it is OK for you or not. This patch has different memory straps than my stock BIOS and Stilt timings that seem to be too much for my card. I have re-branded 7950 i.e. R9 280 and I am running it at 1000/1500, undervolted. With timings patch it went from 185H/s to 195, but was not stable any more. Tried other clocks and voltages too, but no luck, it was able to mine couple of hours at best before stopping. Then I decided to do timings tweaks to my stock BIOS by myself and result is also 195H/s, but 100% stable. I was able to get 200 H/s with the same clocks and even tighter timings, but then it was again unstable. So I left it at 195, tested also other algos, real world apps and Furmark, all good. It would do over 200H/s with higher GPU clock, but I like to keep it cool and quiet. With 1000/1500 it works at ~65C with 36-39% fan speed. With similar timings mod I was able to speed up R7 370 1175/1500 from 140H/s to 155 H/s and R9 290 1000/1250 from 275 H/s to 295 H/s. I tested different driver versions with these mods and for me latest 16.12.1 drivers work the best. They are very slightly faster than 16.11.2 and 16.2.1 I used before, just 1-2 more H/s. Maybe it depends on BIOS mods, but for me 15.12 is 4-7 H/s per card slower than 16.12.1 Great research. Thanks for that! I will not take the risk for just a few more H/s. As I have read it seems as if 200H/s is a great speed, and I will let it be as it is. It's as stable as it can be atm, and I should be glad with such result, knowing these cards are second handed. if you do it "good" you'll get a nice increase Custom timing (thx to wolf for the assist) give me an increase of 10-15h and 30+Mh on a 470 4G. (clocks 1165/2050) <-- yeah..  Next phase, rop shedding. |
|
|
|
They have been done, i have done this months ago, shared the roms here as well. Check the op.
Links are working now, thanks. Do you have a thread for RX series / Polaris cards? My RX480's are dieing to get some static undervolting  Haha, no not really, i posted an RX rom somewhere with vddc offset, not sure where :p Probably not going to create a thread regarding this. But perhaps some day.. Greetings |
|
|
|
Of course I tried so, but still does not work Haha, server is down, will be back up asap.. Hrmph .. Are there any tutorials on hex editing roms? Having to use Trixx / AB etc to do a -100mV is a pain, surely that can be done statically in the vBIOS / rom. They have been done, i have done this months ago, shared the roms here as well. Check the op. |
|
|
|
Of course I tried so, but still does not work Haha, server is down, will be back up asap.. Edit: server backup. |
|
|
|
getting a bit more errors over here :p main.c: In function ‘load_file’:
main.c:286:41: error: ‘O_BINARY’ undeclared (first use in this function)
if (-1 == (fd = open(fname, O_RDONLY | O_BINARY))) {
^
main.c:286:41: note: each undeclared identifier is reported only once for each function it appears in
main.c: In function ‘dump’:
main.c:325:30: error: ‘O_BINARY’ undeclared (first use in this function)
if (-1 == (fd = open(fname, O_BINARY | O_WRONLY | O_CREAT | O_TRUNC, 0666)))
^
main.c: In function ‘solve_equihash’:
main.c:1024:3: warning: passing argument 3 of ‘examine_ht’ from incompatible pointer type [enabled by default]
examine_ht(round, queue, buf_ht, rowCounters[round % 2]);
^
main.c:530:6: note: expected ‘cl_mem’ but argument is of type ‘struct _cl_mem **’
void examine_ht(unsigned round, cl_command_queue queue, cl_mem buf_ht)
^
main.c:1024:3: error: too many arguments to function ‘examine_ht’
examine_ht(round, queue, buf_ht, rowCounters[round % 2]);
^
main.c:530:6: note: declared here
void examine_ht(unsigned round, cl_command_queue queue, cl_mem buf_ht)
^
main.c:957:57: warning: unused parameter ‘ctx’ [-Wunused-parameter]
uint32_t solve_equihash(cl_device_id dev_id, cl_context ctx, cl_command_queue queue,
^
main.c: In function ‘mining_mode’:
main.c:1407:18: warning: unused variable ‘status’ [-Wunused-variable]
cl_int status;
^
main.c:1392:50: warning: unused parameter ‘program’ [-Wunused-parameter]
void mining_mode(cl_device_id dev_id, cl_program program, cl_context ctx, cl_command_queue queue,
^
make: *** [main.o] Error 1
They should be all fixed now. I really need to set up an automatic build system... Yup, sa-solver build succesfully. However, mining itself isn't really working :p Found 1 OpenCL platform(s)
Connecting to zec-eu2.nanopool.org:16666
Solver 0.0: unexpected banner "Gateless Gate mining mode ready"
Solver 0.1: unexpected banner "Gateless Gate mining mode ready"
Solver 1.0: unexpected banner "Gateless Gate mining mode ready"
Solver 1.1: unexpected banner "Gateless Gate mining mode ready"
Stratum server sent us the first job
Mining on 2 devices
Hangs there :p |
|
|
|
getting a bit more errors over here :p main.c: In function ‘load_file’:
main.c:286:41: error: ‘O_BINARY’ undeclared (first use in this function)
if (-1 == (fd = open(fname, O_RDONLY | O_BINARY))) {
^
main.c:286:41: note: each undeclared identifier is reported only once for each function it appears in
main.c: In function ‘dump’:
main.c:325:30: error: ‘O_BINARY’ undeclared (first use in this function)
if (-1 == (fd = open(fname, O_BINARY | O_WRONLY | O_CREAT | O_TRUNC, 0666)))
^
main.c: In function ‘solve_equihash’:
main.c:1024:3: warning: passing argument 3 of ‘examine_ht’ from incompatible pointer type [enabled by default]
examine_ht(round, queue, buf_ht, rowCounters[round % 2]);
^
main.c:530:6: note: expected ‘cl_mem’ but argument is of type ‘struct _cl_mem **’
void examine_ht(unsigned round, cl_command_queue queue, cl_mem buf_ht)
^
main.c:1024:3: error: too many arguments to function ‘examine_ht’
examine_ht(round, queue, buf_ht, rowCounters[round % 2]);
^
main.c:530:6: note: declared here
void examine_ht(unsigned round, cl_command_queue queue, cl_mem buf_ht)
^
main.c:957:57: warning: unused parameter ‘ctx’ [-Wunused-parameter]
uint32_t solve_equihash(cl_device_id dev_id, cl_context ctx, cl_command_queue queue,
^
main.c: In function ‘mining_mode’:
main.c:1407:18: warning: unused variable ‘status’ [-Wunused-variable]
cl_int status;
^
main.c:1392:50: warning: unused parameter ‘program’ [-Wunused-parameter]
void mining_mode(cl_device_id dev_id, cl_program program, cl_context ctx, cl_command_queue queue,
^
make: *** [main.o] Error 1
|
|
|
|
@zawawa, compiling on ubuntu 14.04 and getting some errors here mate. (to many to list to be honest :p) Edit: just checked your thread, i see you're on it |
|
|
|
just a moment where etheruem is better but it will not last, but those price are off in my country, for example i have the 470 and 1060 at the same price, and a 1060 is better at zec than a 470 same hash but lower wattage
let's talk once again about real number, assuming an investment of 3k, with 1070 i would buy 6 x 1070, 1 full rig, which will do 1920 sol at 650watt
with the 470 i would buy 11 gpu because you will lose some money on the second rig due to additional component, these 2 rig will hash at 2200/1100watt
now take a look at whattomine, we are basically doing the same net profit per day at 0.1 cent electricity, so again amd being better it's a myth
actually a 1070 can hash at 370 sol/140 watt, at least some model can, so you have better hash(2220) and lower wattage(900), on zcash nvidia win, on etheruem not, but etheruem will be dead soon and i don't care
The only thing happening here is the fact that the AMD VS NVidia "war" is yet again filling up a thread with nonsense not related to the thread itself. We can discuss this subject endlessly, there will simply always be people who prefer AMD and vice versa. It's like the endless Iphone(ios) vs Samsung(android) discussion... @zawawa I'm following your progress up close. Some extra options for us Linux users is always very welcome. Greetings |
|
|
|
Been getting so many pm's regarding the "dead links" issue. mininghwcomparison.com <-- used to be owned by me, obviously not anymore. However, the files are still on my server and can still be accessed. I already changed the links in the op and i know i have posted several other roms here and there, so for those stumbling upon a dead link, just change the " http://www.mininghwcomparison.com/" part to " http://62.212.74.86/~mining/" and it'll work just fine. Greetings. |
|
|
|
Ha, it seems that it's a bit to tight for most of your nano's. Here's a softer one. Link: HereNo reduced core voltage, increased memory voltage (+25mV) Stock memory clock is 550 with the option to raise to 600max. Greetings! |
|
|
|
@Eliovp at this ROM only the voltages are modded, or any mem timings too?
Cause I can raise my Nano mem clocks up to 550+ MHz, but I can´t see any effect at ZEC mining (stock ROMs).
Yes, modified timings as well. Rom update for ZcashZcash Nano Rom Link: HereReduced core voltage, raised memory voltage (which is needed) Have fun! Hello, I suppose these roms doesn't work for a fury-x? Maybe I will try to mod a bios by myself ... No they do not. |
|
|
|
Rom update for ZcashZcash Nano Rom Link: HereReduced core voltage, raised memory voltage (which is needed) Have fun!
|
|
|
|
Here's an updated one. Sapphire 480 4G Hynix + Elpida. Memory at 1950. Link: HereGreetings |
|
|
|
@Eliovp: Tested your ROM with ETH and ZEC, stock clocks, undervolted and underclocked...
Conclusion: No changes, still getting exactly the same results as the other cards with stock BIOS.
Anyway, no complaints. Thanks for trying!
Cheers!
Haha, i honestly was thinking, no way, that's impossible. So i checked the rom i uploaded, of course, it's the wrong one  Will upload the right one as soon as i get at my office. Turning Idiotic mode off again... |
|
|
|
Here, Sapphire 480 4G, basic changes. I didn't change clocks or voltage settings. Use atflash without the -f flag. In other words, DO NOT FORCE THIS rom on your GPU, also if atiflash complains that the device ID is not the same, DO NOT FLASH. You can use this on both Hynix and Elpida. Link: Here"All the older files are still available, just change the " http://www.mininghwcomparison.com" part to " http://62.212.74.86/~mining/". Greetings So you changed only the memory timings? Thank you. I'll give it a try Yup |
|
|
|
Of course i took the time But as i've mentioned before, these were for private clients. I could however do something small (some small changes).. I'll post one today. Greetings That would be highly appreciated Here, Sapphire 480 4G, basic changes. I didn't change clocks or voltage settings. Use atflash without the -f flag. In other words, DO NOT FORCE THIS rom on your GPU, also if atiflash complains that the device ID is not the same, DO NOT FLASH. You can use this on both Hynix and Elpida. Link: Here"All the older files are still available, just change the " http://www.mininghwcomparison.com" part to " http://62.212.74.86/~mining/". Greetings |
|
|
|
Tried that Bios... having problem to even boot to Windows. I guess my Nano couldn't take such aggressive memory timings.  It's most likely the memory undervolt.. |
|
|
|
Hello, guys...
One quck question: is there ROM available for RX 480 4GB Sapphire Nitro (non ref).
I tried manually changing memory straps, but when i open bios in Polaris Bios editor, there are no such timings as 1500, 1750 etc...
Any recommendations?
Thanks
Apparently no one took the time to try to mod one of these as they are less popular than their 8 GB cousins. I have a few 4Gb and I'm happy with their performance but wouldn't mind to try one modded BIOS just for the fun of it Of course i took the time But as i've mentioned before, these were for private clients. I could however do something small (some small changes).. I'll post one today. Greetings |
|
|
|
|
Show Posts
