HYDRA 1.1B PRO and HYDRA 1.1C PRO
Added 2022-01-23 14:47:31 +0000 UTCGuys, there's an important question to be addressed.
At the moment I have encountered a problem which is related to Overriding best cores (applying a custom mask of cores to active processes). The problem is that if I return the standard affinity back to the process, the system will completely freeze (for some applications). Now I'm trying to help me 2 superprogrammers, but there is no result.
In this regard, I offer options on how we can solve this problem.
1) On Monday I will publish version 1.1B with Overriding best cores disabled. At the end of the month (on the 31st) I will publish 1.1C which will include a workaround for Overriding best cores (maybe that will help).
2) Monday night I'll publish a version of 1.1B with Overriding best cores active but limited to the maximum affinity mask (for testing). At the end of the month (on the 31st) 1.1C with a number of tweaks.
Which way shall we go?
+
Maybe someone you know knows people who can understand the monitoring of system processes. Any help could be very helpful.
Comments
I've been running 1.1C on a 5950x/MSI X-570 mobo. I completed the diagnostics and began running XMRig. My system appeared stable for a few hours then I had an APIC ID:14 error (unstable frequency). I went ahead and reduced Core#8 by 10 in CO Table #2 and I also increased CCD#2 by 10 in SSE CO Offset for MT per the log recommendations. Restarted the system, ran fine for a few hours, suddenly it failed again with APIC ID:14 error. I haven't adjusted my AVX1 values yet since they are all zeroes. Any suggestions about what to try next ?
whoosh
2022-02-10 08:50:55 +0000 UTCGoing for option 1, of course. There's no need for overriding the BEST CORES! You got less than 2 hours to release ;)
Alexander Wünsche
2022-01-24 21:15:28 +0000 UTC1
2022-01-24 15:30:59 +0000 UTC1
2022-01-24 14:49:25 +0000 UTC1
Mumpitz23
2022-01-24 11:35:08 +0000 UTC1
elpuertorro
2022-01-24 10:30:45 +0000 UTC1
2022-01-24 09:45:45 +0000 UTC1
2022-01-24 09:42:02 +0000 UTC1
2022-01-24 09:12:59 +0000 UTCOption 1 would be preferable for me.
Tom Parkinson
2022-01-24 08:41:43 +0000 UTCPrefer option 1
2022-01-24 05:59:55 +0000 UTCOption 1, I need stable version.
newlibrarian
2022-01-24 00:52:39 +0000 UTCOption 1 i would say. Currently i experience freezes when it tries to boost higher. Also i got better single core results with 1.1.9 bios, but better multicore results with hydra. I hope it makes sense to you, because i couldnt understand why this happens.
2022-01-23 21:24:00 +0000 UTCGot unstable system since Hydra too. :( Better a more stable version instead of a full featured one.
Jeliel
2022-01-23 20:45:18 +0000 UTCOption 1
2022-01-23 20:36:17 +0000 UTCOption 1
2022-01-23 19:51:55 +0000 UTCoption 1. and great work my dude.
Doug Wilkinson
2022-01-23 19:25:20 +0000 UTCI agree with option 1
2022-01-23 19:04:53 +0000 UTCHi! Option 1 sound a better option, and great job 👏
2022-01-23 18:59:06 +0000 UTCoption 1 in my opinion
2022-01-23 18:58:58 +0000 UTCOption 1 sounds good
2022-01-23 18:35:10 +0000 UTCThank you for all the work on that - I would prefer option 2
2022-01-23 18:07:51 +0000 UTCOption 1
JohnTheFarmer
2022-01-23 17:48:13 +0000 UTCI think Option 2, for me. But it is the Workload how decides and it is the Number of tweaks. Good Job Yuri
2022-01-23 17:47:52 +0000 UTCOption 1
2022-01-23 17:39:58 +0000 UTCOption 1
2022-01-23 17:07:35 +0000 UTCOption #1
Joni Ollanketo
2022-01-23 16:29:20 +0000 UTCOption #1
2022-01-23 16:09:55 +0000 UTCAgreed, I spent about a month getting my system stable wih Hydra trying different CO values and base frequency. CTR 2.1 RC5 I was using before was absolutely stable for me.
2022-01-23 16:09:09 +0000 UTCOption 2, I think.
2022-01-23 15:44:45 +0000 UTC1. I'm still on CTR 2.1 RC6 as all the newer versions were unstable for me. Please work on stability instead of the new features. Make it rock solid so that one can install it and forget about it.
2022-01-23 15:39:24 +0000 UTCi think if other things is ready, only Overriding best cores not work, so we can wait that problem solved but test another fixes.
2022-01-23 15:38:34 +0000 UTCOption 1
Ricky Campbell
2022-01-23 15:34:02 +0000 UTCOption 1sound good but as stated above a 3rd option might be better. a better version is better than an interim. i rather wait
SubiXT
2022-01-23 15:32:56 +0000 UTCOption 1 looks better. But may I suggest 3th option? Just wait end of the month to get better version without wasting time for interim measures.
2022-01-23 15:24:22 +0000 UTCI'd prefer option 1 as well.
2022-01-23 15:22:52 +0000 UTCOption 1 sounds like the best way to move forward.
2022-01-23 15:09:34 +0000 UTCOption 1
2022-01-23 15:04:02 +0000 UTCOption 1
2022-01-23 14:55:12 +0000 UTCi think my choice is number 1. i can wait when problem is solved and get much possible stable version. :)
2022-01-23 14:54:43 +0000 UTCOption 1
Beech Horn
2022-01-23 14:53:42 +0000 UTC
