Bitcoin Forum
June 13, 2021, 09:03:09 AM *
News: Latest Bitcoin Core release: 0.21.1 [Torrent]
 
   Home   Help Search Login Register More  
Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 14 [15] 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 »
  Print  
Author Topic: AMD Mem Tweak XL - Read/modify timings/pp/straps on the fly  (Read 57192 times)
dragonmike
Hero Member
*****
Offline Offline

Activity: 1218
Merit: 556


VEIL - PRIVACY WITHOUT COMPROMISE


View Profile
April 16, 2019, 12:22:46 PM
 #281

That said, on your (even modded Polaris) GPU, drop RFC for example.. just try it ;-)

Cheers!


Dropping RFC from 260 to 250 results in much lower hashrate on pool for my Vega rig.
At least at 1107 MHz HBM clocks on CN_Turtle.

So far I'm quite happy with --CL 16 --RAS 30 --RCDRD 14 --RP 12 --RC 44. Reported pool hash is spot on - and ultimately, that's what matters.
It was said for Polaris cards ))
It was said "even modded Polaris" which to me implies that it is supposed to work on Vegas too. But maybe it just got lost in translation.

On another note, what's the idea behind increasing values like FAW or REF?



▄▄                                  ▄▄
 ███▄                            ▄███
  ██████                      ██████
   ███████                  ███████
    ███████                ███████
     ███████              ███████
      ███████            ███████
       ███████▄▄      ▄▄███████
        ██████████████████████
         ████████████████████
          ██████████████████
           ████████████████
            ██████████████
             ███████████
              █████████
               ███████
                █████
                 ██
                  █
veil|     PRIVACY     
     WITHOUT COMPROMISE.       
▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂▂
|   NO ICO. NO PREMINE. 
   X16RT GPU Mining. Fair distribution.   
|      The first Zerocoin-based Cryptocurrency       
   WITH ALWAYS-ON PRIVACY.   
|



                   ▄▄████
              ▄▄████████▌
         ▄▄█████████▀███
    ▄▄██████████▀▀ ▄███▌
▄████████████▀▀  ▄█████
▀▀▀███████▀   ▄███████▌
      ██    ▄█████████
       █  ▄██████████▌
       █  ███████████
       █ ██▀ ▀██████▌
       ██▀     ▀████
                 ▀█▌




   ▄███████
   ████████
   ███▀
   ███
██████████
██████████
   ███
   ███
   ███
   ███
   ███
   ███




     ▄▄█▀▀ ▄▄▄▄▄▄▄▄ ▀▀█▄▄
   ▐██▄▄██████████████▄▄██▌
   ████████████████████████
  ▐████████████████████████▌
  ███████▀▀▀██████▀▀▀███████
 ▐██████     ████     ██████▌
 ███████     ████     ███████
▐████████▄▄▄██████▄▄▄████████▌
▐████████████████████████████▌
 █████▄▄▀▀▀▀██████▀▀▀▀▄▄█████
  ▀▀██████          ██████▀▀
      ▀▀▀            ▀▀▀
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction. Advertise here.
1623574989
Hero Member
*
Offline Offline

Posts: 1623574989

View Profile Personal Message (Offline)

Ignore
1623574989
Reply with quote  #2

1623574989
Report to moderator
kamisama233
Newbie
*
Offline Offline

Activity: 38
Merit: 0


View Profile
April 16, 2019, 01:51:47 PM
 #282

Rx Vega 56 hynix memory  
1250/875mv   916/925mv
tartl cn 20.12 - 20.20  kh/s  L24+24
CN/r   2160   1450/900
60 min no errors.
WinAMDTweak.exe --rcdrd 19 --rcdwr 4 --rc 35 --rp 13 --rrds 4 --rrdl 5 --rfc 148 --REF 15600
Cool!!!

cool!  cnr 1450/900? is it 1450 core and 900 memory? how about the voltage? is it can run for long time?
exsexs1
Newbie
*
Offline Offline

Activity: 31
Merit: 0


View Profile
April 16, 2019, 02:39:28 PM
 #283

Thanks @Eliovp for this tool.
I have a problem:
http://www.kepfeltoltes.eu/images/2019/03/254AMDTweak.jpg
http://www.kepfeltoltes.eu/images/2019/03/254AMDTweak_tn.jpg

Tweak shows crazy GPU values, but this is GPU0 in miner (not GPU 5 like a screenshot) and low hashrate. I think tweak not applied the values.

My rig:
Win10, 18.6.1 driver, 6xVega56@64 Sqmsung RAM.
Zorg33
Jr. Member
*
Offline Offline

Activity: 146
Merit: 2


View Profile
April 16, 2019, 03:47:05 PM
Last edit: April 16, 2019, 04:11:16 PM by Zorg33
 #284

On another note, what's the idea behind increasing values like FAW or REF?
A good reading on this matter:
https://www.micron.com/about/blog/2013/december/addressing-the-tfaw-effect-for-ddr3
rasic
Newbie
*
Offline Offline

Activity: 1
Merit: 0


View Profile
April 16, 2019, 04:59:50 PM
 #285

Any specific command to enable it in HiveOS?

I'm trying the standard ./amdmemtweak with -- commands... I see the targeted values change when using --current option, but it has no effect on the HR, like nothing happens.

Cheers,
GR
plox99
Newbie
*
Offline Offline

Activity: 9
Merit: 0


View Profile
April 16, 2019, 05:42:11 PM
 #286

Does anyone try on VEGA 56 non-ref Sapphire, Samsung memory?

I tried different values from topic and no result on CnR
Anwil
Newbie
*
Offline Offline

Activity: 13
Merit: 5


View Profile
April 16, 2019, 05:54:02 PM
 #287

Rx Vega 56 hynix memory  
1250/875mv   916/925mv
tartl cn 20.12 - 20.20  kh/s  L24+24
CN/r   2160   1450/900
60 min no errors.
WinAMDTweak.exe --rcdrd 19 --rcdwr 4 --rc 35 --rp 13 --rrds 4 --rrdl 5 --rfc 148 --REF 15600
Cool!!!

cool!  cnr 1450/900? is it 1450 core and 900 memory? how about the voltage? is it can run for long time?
Everyone is different, need to test.. voltage pick up your.
pbfarmer
Member
**
Offline Offline

Activity: 340
Merit: 29


View Profile
April 16, 2019, 07:45:28 PM
 #288

On another note, what's the idea behind increasing values like FAW or REF?

Don’t know why anyone would want to increase FAW from stock.  As for REF, see my last comment here:

https://bitcointalk.org/index.php?topic=5123724.msg50611707#msg50611707
wacholek
Newbie
*
Offline Offline

Activity: 55
Merit: 0


View Profile
April 16, 2019, 08:28:17 PM
 #289

Hi Guys,

Zorg33 said to me share this set, because it's impossible, but working good for me:
--CL 12 --RAS 28 --RCDRD 12 --RCDWR 5 --RC 44 --RP 10 --RRDS 3 --RRDL 3 --RTP 5 --FAW 20 --CWL 7 --WTRS 4 --WTRL 9 --WR 16 --WRRD 1 --RDWR 19 --REF 17550 --MRD 8 --MOD 15 --PD 8 --CKSRE 10 --CKSRX 10 --RFC 250

Vega56 @ 64 sapphire samsung ram, Win10, 18.6.1 and 18.5.1 classic 1408/900core 1100/900mem

Turtle 22+
ETH 51+
Cr-Lite 10K khs (1600/990 1100/900, but crashed, 9600kh 1408/900 1100/900 super stable after 24h)




This is by far the best settings I tired.
4x Vega 64 with waterblocks.
Core 1300@0.950V and 1100@0.95V Normally it was 19.24kH/s/Vega. After the timing set is about 21.85kH/s/Vega.
Max temps: Core 40*C, HBM mem 53*C.
Team Red Miner v0.4.3 L22+22. HWinfo GPU chip power 157W. I don't have a power meter.
almost 4h no errors. https://ibb.co/R7CWKjw
Mashy81
Jr. Member
*
Offline Offline

Activity: 224
Merit: 1


View Profile
April 16, 2019, 08:32:13 PM
 #290

Hi Guys,

Zorg33 said to me share this set, because it's impossible, but working good for me:
--CL 12 --RAS 28 --RCDRD 12 --RCDWR 5 --RC 44 --RP 10 --RRDS 3 --RRDL 3 --RTP 5 --FAW 20 --CWL 7 --WTRS 4 --WTRL 9 --WR 16 --WRRD 1 --RDWR 19 --REF 17550 --MRD 8 --MOD 15 --PD 8 --CKSRE 10 --CKSRX 10 --RFC 250

Vega56 @ 64 sapphire samsung ram, Win10, 18.6.1 and 18.5.1 classic 1408/900core 1100/900mem

Turtle 22+
ETH 51+
Cr-Lite 10K khs (1600/990 1100/900, but crashed, 9600kh 1408/900 1100/900 super stable after 24h)




This is by far the best settings I tired.
4x Vega 64 with waterblocks.
Core 1300@0.950V and 1100@0.95V Normally it was 19.24kH/s/Vega. After the timing set is about 21.85kH/s/Vega.
Max temps: Core 40*C, HBM mem 53*C.
Team Red Miner v0.4.3 L22+22. HWinfo GPU chip power 157W. I don't have a power meter.
almost 4h no errors.
<a href="https://ibb.co/R7CWKjw"><img src="https://i.ibb.co/t3b56Lr/Vega.jpg" alt="Vega" border="0"></a><br /><a target='_blank' href='https://pl.imgbb.com/'>prywatne serwery plemiona</a><br />

You would be using a  lot of power. You should be able to drop your core and voltage a lot. 950mv is soo high for 1300 core.
wacholek
Newbie
*
Offline Offline

Activity: 55
Merit: 0


View Profile
April 16, 2019, 08:38:08 PM
 #291

Hi Guys,

Zorg33 said to me share this set, because it's impossible, but working good for me:
--CL 12 --RAS 28 --RCDRD 12 --RCDWR 5 --RC 44 --RP 10 --RRDS 3 --RRDL 3 --RTP 5 --FAW 20 --CWL 7 --WTRS 4 --WTRL 9 --WR 16 --WRRD 1 --RDWR 19 --REF 17550 --MRD 8 --MOD 15 --PD 8 --CKSRE 10 --CKSRX 10 --RFC 250

Vega56 @ 64 sapphire samsung ram, Win10, 18.6.1 and 18.5.1 classic 1408/900core 1100/900mem

Turtle 22+
ETH 51+
Cr-Lite 10K khs (1600/990 1100/900, but crashed, 9600kh 1408/900 1100/900 super stable after 24h)




This is by far the best settings I tired.
4x Vega 64 with waterblocks.
Core 1300@0.950V and 1100@0.95V Normally it was 19.24kH/s/Vega. After the timing set is about 21.85kH/s/Vega.
Max temps: Core 40*C, HBM mem 53*C.
Team Red Miner v0.4.3 L22+22. HWinfo GPU chip power 157W. I don't have a power meter.
almost 4h no errors. http://<a href="https://ibb.co/R7CWKjw"><img src="https://i.ibb.co/t3b56Lr/Vega.jpg" alt="Vega" border="0"></a><br /><a target='_blank' href='https://pl.imgbb.com/'>prywatne serwery plemiona</a><br />
<a href="https://ibb.co/R7CWKjw"><img src="https://i.ibb.co/t3b56Lr/Vega.jpg" alt="Vega" border="0"></a><br /><a target='_blank' href='https://pl.imgbb.com/'>prywatne serwery plemiona</a><br />

You would be using a  lot of power. You should be able to drop your core and voltage a lot. 950mv is soo high for 1300 core.
You're right, but for now I would like to test this settings for a while and than I will decrease voltage. However I did observe some stability issue at 0.900V but it was at different timings an on LC bios which I found to be not very good for low power tuning.
rednoW
Legendary
*
Offline Offline

Activity: 1514
Merit: 1003


View Profile
April 16, 2019, 09:07:58 PM
 #292


This is by far the best settings I tired.
4x Vega 64 with waterblocks.
Core 1300@0.950V and 1100@0.95V Normally it was 19.24kH/s/Vega. After the timing set is about 21.85kH/s/Vega.
Max temps: Core 40*C, HBM mem 53*C.
Team Red Miner v0.4.3 L22+22. HWinfo GPU chip power 157W. I don't have a power meter.
almost 4h no errors. https://ibb.co/R7CWKjw

vega64 lc is happily mining 21.2kh/s with 1230/1090@800mv showing only 115 watt in HWinfo GPU chip power
pbfarmer
Member
**
Offline Offline

Activity: 340
Merit: 29


View Profile
April 16, 2019, 11:06:27 PM
Last edit: April 16, 2019, 11:17:55 PM by pbfarmer
 #293

First - quick results for cn-trtl, w/ efficiency-focused settings, using timings mostly borrowing from others here, w/ some minor tweaks:

Vega 64 air, ubuntu 18.04 + amdgpu-pro 18.50, TRM 0.4.3 (L18+18), 852 cclock (p0)/1107 mclock/818mv, power readings at the wall

stock timings:
--CL 20 --RAS 33 --RCDRD 16 --RCDWR 10 --RC 47 --RP 14 --RRDS 4 --RRDL 6 --RFC 260 (--REF 3900)
18.5 kh/s @ 135w (137 h/w)

modded timings 1
--CL 19 --RAS 28 --RCDRD 12 --RCDWR 5 --RC 44 --RP 12 --RRDS 3 --RRDL 3 --RFC 248
19.75 kh/s @ 137w (144 h/w)

modded timings 2
same as above, plus --REF 15600
20.71 kh/s @ 137w (151 h/w)


Second - notes on power...  I don't see any appreciable power differences - nor would I expect to.  Clocks and voltages are untouched, we simply have a bit more data being transferred.  Even the 2w difference I'm reporting here is conservative - taking natural fluctuations in my readings into account, my actual increase could be closer to <= 1w.  People seeing large power increases (at least on vega 64) seem to have something else going on.


Last - some conjecture / educated guessing re: THAT --REF THO!!!  I'm assuming --REF is the refresh frequency, in nanoseconds, and unlike most timings, a higher value (meaning less refreshing) is better.  Refreshes steal bandwidth, and AMD seems to have gone majorly conservative (aggressive?) on this, probably due to the super high temps of the HBM during normal/gaming use.  As leakage increases w/ temps, more refreshes would be required when running your GPU/HBM at high clocks/voltages.  Since (efficient) miners tend to run cooler, the crazy high default refresh rate is really unnecessary.  I found 4x the default to be around where returns quickly diminish, at least at my clocks - i can get maybe another 50 h/s (turtle) going 4.5x.  HOWEVER - if you run super aggressive for max h/r, or just aren't effectively cooled in general, you may want to dial this back, or you may start seeing mem errors / bad shares from corrupted data due to insufficient refreshing / leakage.



After applying timing mods to an 8x vega 64 rig, I have a better picture of power use, and my initial observation does not hold.  On this 8x64 rig, I now see an avg of ~8-9w increase w/ my timings above, though that's still only a ~6% increase (for ~+12% h/r.)  However, I could see this delta possibly increasing for those running higher cclocks.

Also, tried dropping CL back from 19 to 20, and saw no appreciable change in h/r (maybe a 10h/s drop)

This now leaves me @ 144 h/w for cn-trtl.
kamisama233
Newbie
*
Offline Offline

Activity: 38
Merit: 0


View Profile
April 17, 2019, 01:13:39 AM
 #294

First - quick results for cn-trtl, w/ efficiency-focused settings, using timings mostly borrowing from others here, w/ some minor tweaks:

Vega 64 air, ubuntu 18.04 + amdgpu-pro 18.50, TRM 0.4.3 (L18+18), 852 cclock (p0)/1107 mclock/818mv, power readings at the wall

stock timings:
--CL 20 --RAS 33 --RCDRD 16 --RCDWR 10 --RC 47 --RP 14 --RRDS 4 --RRDL 6 --RFC 260 (--REF 3900)
18.5 kh/s @ 135w (137 h/w)

modded timings 1
--CL 19 --RAS 28 --RCDRD 12 --RCDWR 5 --RC 44 --RP 12 --RRDS 3 --RRDL 3 --RFC 248
19.75 kh/s @ 137w (144 h/w)

modded timings 2
same as above, plus --REF 15600
20.71 kh/s @ 137w (151 h/w)


Second - notes on power...  I don't see any appreciable power differences - nor would I expect to.  Clocks and voltages are untouched, we simply have a bit more data being transferred.  Even the 2w difference I'm reporting here is conservative - taking natural fluctuations in my readings into account, my actual increase could be closer to <= 1w.  People seeing large power increases (at least on vega 64) seem to have something else going on.


Last - some conjecture / educated guessing re: THAT --REF THO!!!  I'm assuming --REF is the refresh frequency, in nanoseconds, and unlike most timings, a higher value (meaning less refreshing) is better.  Refreshes steal bandwidth, and AMD seems to have gone majorly conservative (aggressive?) on this, probably due to the super high temps of the HBM during normal/gaming use.  As leakage increases w/ temps, more refreshes would be required when running your GPU/HBM at high clocks/voltages.  Since (efficient) miners tend to run cooler, the crazy high default refresh rate is really unnecessary.  I found 4x the default to be around where returns quickly diminish, at least at my clocks - i can get maybe another 50 h/s (turtle) going 4.5x.  HOWEVER - if you run super aggressive for max h/r, or just aren't effectively cooled in general, you may want to dial this back, or you may start seeing mem errors / bad shares from corrupted data due to insufficient refreshing / leakage.

nice test, can you test cnr algo, is it necessary  flash 56 to 64 for this algo?
N2DCRYPT
Jr. Member
*
Offline Offline

Activity: 148
Merit: 5


View Profile
April 17, 2019, 01:59:35 AM
 #295

Well that's fun... Vega FE's can crack the 2800 H/s threshold on CN-Heavy Algos Cheesy

https://i.imgur.com/IUZtjAL.jpg
ingyenfrag
Newbie
*
Offline Offline

Activity: 84
Merit: 0


View Profile
April 17, 2019, 07:04:20 AM
 #296

This is my lighter timing, hopefully stable (by:Zorg33):
--CL 19 --RAS 30 --RCDRD 12 --RCDWR 6 --RC 44 --RP 13 --RRDS 5 --RRDL 5 --RTP 4 --FAW 18 --CWL 6 --WTRS 4 --WTRL 9 --WR 15 --WRRD 1 --RDWR 18 --REF 17000 --RFC 248

These timings looks stable after a few days and give a pretty good hashrate improvement.


Zorg33 said to me share this set, because it's impossible, but working good for me:
--CL 12 --RAS 28 --RCDRD 12 --RCDWR 5 --RC 44 --RP 10 --RRDS 3 --RRDL 3 --RTP 5 --FAW 20 --CWL 7 --WTRS 4 --WTRL 9 --WR 16 --WRRD 1 --RDWR 19 --REF 17550 --MRD 8 --MOD 15 --PD 8 --CKSRE 10 --CKSRX 10 --RFC 250

Now I'm trying these timings. Tiny bit better hashrate. Let's see the stability. Power consumption is about the same as the CL19.


I've tried The CL 16 strap --> -CL 16 --RAS 30 --RCDRD 14 --RCDWR 14 --RC 44 --RP 14s --RRDS 3 --RRDL 5 --RTP 5 --FAW 20 --CWL 7 --WTRS 4 --WTRL 9 --WR 16 --WRRD 1 --RDWR 19 --REF 17550 --MRD 8 --MOD 15 --PD 8 --CKSRE 10 --CKSRX 10 --RFC 244
but it gives a worse hashrate @ CNR, but still better than stock timings.
wacholek
Newbie
*
Offline Offline

Activity: 55
Merit: 0


View Profile
April 17, 2019, 07:21:47 AM
 #297


This is by far the best settings I tired.
4x Vega 64 with waterblocks.
Core 1300@0.950V and 1100@0.95V Normally it was 19.24kH/s/Vega. After the timing set is about 21.85kH/s/Vega.
Max temps: Core 40*C, HBM mem 53*C.
Team Red Miner v0.4.3 L22+22. HWinfo GPU chip power 157W. I don't have a power meter.
almost 4h no errors. https://ibb.co/R7CWKjw

vega64 lc is happily mining 21.2kh/s with 1230/1090@800mv showing only 115 watt in HWinfo GPU chip power
I will get to that eventually.

For now I'm testing other settings.
On normal timings the best intensity was L24+24. at the same core/mem freq I had consistent 19,24kH/s on all cards +-0.05kH/s.
On new timings this intensity was not that stable. After that L22+22. After 4,5h of hashing first HW error showed up and speed was 21.85 kH/s but with the spread od 0.4kh/s betwen cards.
After that I have tested L20+20 but that was ruled out because of decrease in hashrate of about 0.5kH/s/cards.
Now I'm using L18+18 and got result of 21.83 kH/s +-0.03kH/s between cards.
Using Adrenalin 19.4.1.
rednoW
Legendary
*
Offline Offline

Activity: 1514
Merit: 1003


View Profile
April 17, 2019, 08:03:14 AM
 #298

guys. this crazy high --ref values you are using can't be good. check your pool-side hash ...
pbfarmer
Member
**
Offline Offline

Activity: 340
Merit: 29


View Profile
April 17, 2019, 08:21:36 AM
 #299

guys. this crazy high --ref values you are using can't be good. check your pool-side hash ...

Why?  if temps are low enough, data is more stable, and refreshes can be less frequent.  Stock setting is almost certainly tuned to gaming usage under expectation of super high temps, plus a healthy margin.  And if the setting is too high, it should manifest in mem errors and/or invalid results due to bit flips - no need to look at pool rate.

FYI, my pool rate is as expected (w/ --REF 15600)
Zorg33
Jr. Member
*
Offline Offline

Activity: 146
Merit: 2


View Profile
April 17, 2019, 08:36:28 AM
 #300

--CL 12 --RAS 28 --RCDRD 12 --RCDWR 5 --RC 44 --RP 10 --RRDS 3 --RRDL 3 --RTP 5 --FAW 20 --CWL 7 --WTRS 4 --WTRL 9 --WR 16 --WRRD 1 --RDWR 19 --REF 17550 --MRD 8 --MOD 15 --PD 8 --CKSRE 10 --CKSRX 10 --RFC 250

What memclk?
I got instacrash with this. Also you can try to lower FAW to even 12. After some discussion it seems lower FAW is better.
Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 14 [15] 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 »
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!