lhr unlock 100

[thumerman]

Hi, my video card unexpectedly unlocked when I decided to lower the memory frequency a little, after a reboot everything went off. How can this be explained.

[Faltskog]

@Faltskog What are the settings for 90 MH/s on 3080Ti?

Easy: Absolute core clock: 1140 Mem clock: +1300/1500 (windows) +2600/3000 (hiveos) No power limit

i have this cfg @zhanko73

[gmzombie]

Has anyone else gotten to get this unlock to work with other cards besides the 3070's? im trying with 3060s and i will with m y 3060ti when i get the 3060s done. im using windows and just cant seem to get an invalid share. it crashes the miner rather than giving an invalid share. im using trex 24.8 driver 510.06 and windows 10. also tried with windows 11.

[nCr78]

@trexminer Would it be possible to save a debug / protocol log from your protocol dump (or something similar) for when someone manages to replicate this? We can send you the 1-2 logs that will be produced and you can check for a pattern? It could be that certain invalid shares do something to the card because of the reason they became invalid (?)

[basking-in-the-sun2000]

Has anyone else gotten to get this unlock to work with other cards besides the 3070's? im trying with 3060s and i will with m y 3060ti when i get the 3060s done. im using windows and just cant seem to get an invalid share. it crashes the miner rather than giving an invalid share. im using trex 24.8 driver 510.06 and windows 10. also tried with windows 11.

@thumerman (the OP) got it working on a 3060 ti. Could be that it is card dependent, and not all will go fhr.

The oc is no guarantee that you will get the "right" invalid share. Had a week with several invalid shares and no fhr kicking in. Also, if your oc is too high you will crash the rig, if not your card won't kick in. I did notice that having high oc doesn't mean it will give you an invalid fast. When it worked, it took all night to get a couple.

[thumerman]

I have a theory about this for everyone, we all know that now there is a problem with a shortage of semiconductors in the world and video card manufacturers began to install chips that were defective earlier, this is evidenced by the fact that when generating a dag file, an error occurs about a defective segment like @basking -in-the-sun2000 is the same for me. If you do not have a similar error, then unlocking for you will not work with a probability of 99.9%. For those who have such an error, you can work the card to the fullest in various overclocking modes. Now my card is working at full capacity in another miner in dual mode, the last time I unlocked it was after trying to lower the PL.

[gmzombie]

ok interesting. currently im going to go back to your driver 497.09 as im on 510.06 in windows 10. i can get the damaged dag crc error by adjusting my mem clock at 1424 or more. tested backwards to see and that is the line in the sand where it will give me a crc error for an evga 3060 xc. unfortantly i icant get the msi ventis 2 fan 3060 to error even at 1500 in windows. lol it just wants to push out more hash i know it. so im going to goto the older driver. one thing i noticed more times than not when i get that crc error on the dag the very first quick share is almost full speed. ovbiously it goes down but its good to see it lol. anyways i know we can get there with this. i will ask anyone who is trying to list their card brand and model that way we know which ones have been tested and what values you used to get the crc error. am i missing something?

[Faltskog]

I have a theory about this for everyone, we all know that now there is a problem with a shortage of semiconductors in the world and video card manufacturers began to install chips that were defective earlier, this is evidenced by the fact that when generating a dag file, an error occurs about a defective segment like @basking -in-the-sun2000 is the same for me. If you do not have a similar error, then unlocking for you will not work with a probability of 99.9%. For those who have such an error, you can work the card to the fullest in various overclocking modes. Now my card is working at full capacity in another miner in dual mode, the last time I unlocked it was after trying to lower the PL.

Can you detail every step by step, plus pictures to help us out? or like a list of the conditions you have to do or have to achieve such FHR? It happens only but only with Samsung chips? Or even i can achieve FHR with my 3080ti with Micron chips?

I would really love to know more about this.

[jemshit]

Need to keep an eye on this: https://github.com/BySergeyDev/NvidiaRTX-LHRv2Unlocker

Doesn't give anything other than SS right now, but lets see

[AlbyGNinja]

Doesn't give anything other than SS right now, but lets see

Sound "magic" (aka scam) But let's see

[basking-in-the-sun2000]

Hadn't happened in a while, but went fhr again. Not sure if it was related, but just added cpuminer-opt-rplant as a miner. Noticed the lhr value was at 72, even when I had set it at 71.

Even when I had gotten some during the last few weeks, didn't give it much importance that there was an invalid share. That was until I noticed the lhr value. Haven't been able to figure out what makes an invalid share a "good" one and others not.

Also reduced the core last week from 1470 to 1170. I have no idea what triggers the card doing fhr even when it is lhr.

Sadly I lost being able to mine alph by going fhr

-----------------------20220301 00:19:15 ------------------------                                                                       
Mining at eu.kroxy.k1pool.com:4444 [37.187.226.146], diff: 4.00 G                                                                       
GPU #4: RTX 3060 Ti - 39.01 MH/s, [LHR 68.5<>]  [T:55C, P:112W, F:38%, E:331kH/W], 19/19 R:0% I:0%
GPU #5: RTX 3070    - 61.98 MH/s, [LHR 100.0<>] [T:56C, P:116W, F:59%, E:539kH/W], 27/27 R:0% I:0%
GPU #7: RTX 3060 Ti - 40.00 MH/s, [LHR 70.0<>]  [T:57C, P:104W, F:63%, E:360kH/W], 19/19 R:0% I:0%
Hashrate: 140.99 MH/s, Shares/min: 3.817 (Avg. 2.098), Avg.P: 344W, Avg.E: 410kH/W                                                      
Max diff share was found by GPU #5, diff: 398.17 G                                                                                      
Uptime: 31 mins 1 sec | Algo: ethash | T-Rex v0.25.6                                                                                    

-----------------------20220301 00:19:15 ------------------------                                                                       
Mining at pool.woolypooly.com:3107 [147.135.70.208], diff: 1.07 G                                                                       
GPU #4: RTX 3060 Ti - 335.42 MH/s, [LHR 68.5<>]  [T:55C, P:112W, F:38%, E:2.84MH/W], 36/36 R:0% I:0%
GPU #5: RTX 3070    -    0.00 H/s, [LHR 100.0<>] [T:56C, P:116W, F:59%, E:0.00H/W]
GPU #7: RTX 3060 Ti - 303.96 MH/s, [LHR 70.0<>]  [T:57C, P:104W, F:63%, E:2.74MH/W], 24/24 R:0% I:0%
Hashrate: 639.38 MH/s, Shares/min: 3.812 (Avg. 1.963), Avg.P: 344W, Avg.E: 1.86MH/W                                                     
Uptime: 31 mins 1 sec | Algo: blake3 | T-Rex v0.25.6                                                                                    

[lowzyyy]

at this point its card dependent. My 3080 ti cant even produce invalid share. I can only crash the driver and restart the pc.

[basking-in-the-sun2000]

I know this is a moot point, but maybe with nvidia fighting back, it might provide a glimmer of a solution for other cards. I know some people kept thinking my 3070 was fhr, but it isn't. lol Nor was it a start of mining spike in the hashrate (did mine for over 24h at fhr). Hive just updated their code to identify the cards for ye of little faith ;) Still, it might be a bug that only happens in one of every thousand cards?

Maybe something to keep in mind for when the current unlock doesn't work with new cards