Home Artists Posts Import Register
Join the new SimpleX Chat Group!

Downloads

Content

I'm glad to see you, thank you for your support. Today is a special day, the release of a long-awaited massive update. Please take a look at the list of changes. You'll also find a PDF document which contains a step-by-step guide and tips (the document has also been updated). No new diagnostics are required.

_______

Note that some anti-viruses will consider the program a virus, this is a false positive. You can read the report
here.
I advise you to add the folder to exceptions.

_______

The program is in a 7-ZIP archive.  
Download the archiver
.NET Framework 4.8.
Download

_______  


What's new?  


* Cezanne support (Ryzen 7 5700G and Ryzen 5 5600G).

* Added profiles for low-thread load with dynamic mode and stats.

* Added ability to customize CORE CO diagnostics (selects only the cores you want to test).

* Added a Clear standby cache feature with the ability to adjust the frequency of cache clearing ( improved gaming comfort ). 

* Added new trigger for game profile - GPU MEMORY trigger. Allows to improve system stability in non-game loads and reduce number of false triggers.

* Added full support for NON-SMT for all processors and all modes.

* Fixed memory leaks in some scenarios, but still require the latest versions of video drivers.

* Added ability to switch CO tables for GAME profile.

* Added the ability to manage Overboost.

* Added an option to limit the maximum frequency for the boost.

* Added "Realtime CO bottleneck" feature that allows you to estimate which core is the "weakest link" in any load and also allows you to tune CO tables more easily.

* Added ability to change CO in real time without deactivating profiles.

* Added ability to adjust frequency for GAME profile with GAME CO offset without modifying CO in tables.

* Modified the recommended basic settings.

* The second CO table is now the default for the GAME profile since the CO values for this table are more balanced.

* The function identifying the core that caused the failure has been updated, and the tips have been updated.

* Significantly improved stability of diagnostics and automatic recovery of diagnostics after rebooting the system.

* The core activity state is now defined with "thread usage" instead of "C0 core state". This allows HYDRA not to react to background system processes and to use requested cores more accurately during load times.

* Ramping frequency for most modes now requires less time, i.e. you get more performance.

* Corrected descriptions, added new pop-up tips.

* Improved stability of communication with SMU after an attempt to resend a command caused by a failure.

* Fixed conflict with CTR, now it is impossible to start HYDRA if CTR is already running.

* Changed maximum and minimum voltage tolerances for profiles.

* High core temperature during GAME profile creation. Fixed.

_______

Special thanks to @eXp.fr, @yxes,  @profitkaa, @Riwwelorsch, @Ann Martiné, @ACE, @arcanazen, @DayDreamer, @Gadfly, @rizlah , @&&Devin, @Bloax , @RdWing and @wanderinarcheologist  for help with development and testing.

Files

Comments

Rory O'Donnell

...excellent, will try soon

Xdrqgol

wohooo. testing it asap! @Yuri - i am on DAKR HERO and for me BIOS 3003 has always been best in all senses ... would it be a problem with Hydra, if i didn't updated the BIOS to the latest ?

Danut Rotaru

Had an issue where HYDRA did not start during diagnostic ( after a restart) Was that fixed ?

Frederico Torres

A new diagnostic is needed? Or can I proceed with the old one? Thanks for your work Yuri :)

Anonymous

Good news I’ll try this version today !

Anonymous

Hello, I have a general question as I could not find an answer in any other post. My CCD1 is at 4750 and CCD2 is at 4550 in the GAME (All) profile. If i have full load on both CCD´s it switches correctly to the GAME profile but both CCD´s stay at 4550. Should CCD1 not boost to 4750 instead? And overall thanks for this amazing work.

iNerd

you have possibility now in the new peo version to activate first ccd for gaming. On pdf is written this increase system instability but you can try. you can, i believe otherwize Yuri or others can correct me, increase positive offset for gaming under gaming tab…

iNerd

running diagnostics with ram oc for the first time. Waiting is hard… 😃

Anonymous

Awesome work! CTR was never very stable for but hydra seems to be so far. . . very excited

Rory O'Donnell

One suggested improvement, for users with more than 1 monitor - have option to prevent Hydra from being minimized. I have hydra displayed on the left monitor, when I diagnose, the CB23 opens on the center monitor but also my hydra window is minimized. It's very grrr :-)

iNerd

Yuri you're amazing! all game stucked at 4760mhz constantly. no stuttering (just had first run, everything set up as i always do). Feel smooth, also browsing through os or applications is smooth, feels not so overpowered. maybe i had a lucky hand with my settings but i'm really happy until now! Thanks!

iNerd

one extra information, for me most important: with my aio for the 5900x and airflow for gpu 3080Ti i'm about 10-12 degrees lower on CPU and about 6-7 degrees on GPU. And with better performance. Amazing!

Anonymous

hello i have a probleme

Anonymous

***HYDRA 0.9D PRO by 1usmus*** 08/26/2021 10:53:05 Microsoft Windows NT 6.2.9200.0 AMD Ryzen 5 5600X 6-Core Processor GIGABYTE B550M AORUS PRO-P BIOS ver. F13 SMU ver. 56.53.00 TABLE ver. 3672325 DRAM speed 3733 MHz 10:53:43: Test 1 started! 10:53:43: 1T-6T Basic statistics CORE / CPPC / FREQ / FREQ EFF / VID / POWER / TEMP C01 129 F 4458 F 3376 V 1.352 W 8.73 T 55.7 C02 133 F 4648 F 4580 V 1.352 W 12.06 T 55.08 C03 140 F 4648 F 4634 V 1.352 W 13.14 T 60.91 C04 140 F 4649 F 4648 V 1.352 W 14.44 T 60.78 C05 137 F 4648 F 4640 V 1.352 W 11.64 T 57.35 C06 125 F 4032 F 1098 V 1.352 W 3.46 T 45.9 Start VID 1400 Start FREQ CCD#1 4550 Vdroop: 0.86% 4575 - PASSED 1.218 4600 - PASSED 1.224 4625 - PASSED 1.245 4650 - PASSED 1.259 4675 - PASSED 1.284 4700 - PASSED 1.311 4725 - PASSED 1.324 4750 - PASSED 1.357 10:54:05: Cinebench stopped! 10:54:05: Test 1 finished! 10:54:05: Test 2 started! 10:54:05: ALL THREADS (AVX1) Basic statistics CORE / CPPC / FREQ / FREQ EFF / VID / POWER / TEMP C01 129 F 4564 F 4564 V 1.352 W 15.09 T 67.01 C02 133 F 4564 F 4564 V 1.352 W 15.16 T 63.04 C03 140 F 4564 F 4564 V 1.352 W 15.38 T 68.22 C04 140 F 4564 F 4564 V 1.352 W 15.43 T 68.54 C05 137 F 4564 F 4564 V 1.352 W 15.44 T 66.29 C06 125 F 4564 F 4564 V 1.352 W 15.45 T 69.68 Start VID 1275 Start FREQ CCD#1 4300 Vdroop: 1.41% 4325 1.108 - PASSED 4350 1.116 - PASSED 4375 1.124 - PASSED 4400 1.131 - PASSED 4425 1.141 - PASSED 4450 1.157 - PASSED 4475 1.174 - PASSED 4500 1.189 - PASSED 4525 1.207 - PASSED 4550 1.227 - PASSED CCX#1: 4550MHz 10:54:28: Cinebench stopped! 10:54:28: Test 2 finished! 10:54:28: Test 3 started! 10:54:28: ALL THREADS (AVX2) Basic statistics CORE / CPPC / FREQ / FREQ EFF / VID / POWER / TEMP C01 129 F 4348 F 4348 V 1.212 W 14.42 T 64.44 C02 133 F 4348 F 4348 V 1.212 W 14.14 T 59.86 C03 140 F 4348 F 4348 V 1.212 W 14.73 T 65.37 C04 140 F 4348 F 4348 V 1.212 W 14.96 T 67.57 C05 137 F 4348 F 4348 V 1.212 W 14.83 T 64.97 C06 125 F 4348 F 4348 V 1.212 W 14.92 T 68.46 Start VID 1125 Start FREQ CCD#1 3800 Vdroop: 1.69% 3825 0.969 - PASSED 3850 0.976 - PASSED 3875 0.982 - PASSED 3900 0.989 - PASSED 3925 0.996 - PASSED 3950 1.003 - PASSED 3975 1.01 - PASSED 4000 1.016 - PASSED 4025 1.026 - PASSED 4050 1.033 - PASSED 4075 1.041 - PASSED 4100 1.049 - PASSED 4125 1.057 - PASSED 4150 1.064 - PASSED 4175 1.069 - PASSED 4200 1.078 - PASSED 4225 1.084 - PASSED CCX#1: 4225MHz 10:54:48: Stress test stopped. 10:54:48: Test 3 finished! 10:54:48: Test 4 started! 10:54:48: ALL THREADS (GAME) Basic statistics CORE / CPPC / FREQ / FREQ EFF / VID / POWER / TEMP C01 129 F 4608 F 4607 V 1.348 W 11.98 T 58.47 C02 133 F 4608 F 4607 V 1.348 W 12.2 T 54.72 C03 140 F 4608 F 4607 V 1.348 W 12.38 T 58.96 C04 140 F 4608 F 4607 V 1.348 W 12.52 T 59.76 C05 137 F 4608 F 4607 V 1.348 W 12.3 T 57.38 C06 125 F 4608 F 4607 V 1.348 W 12.09 T 59.73 Start VID 1350 Start FREQ CCD#1 4500 Vdroop: 0.96% 4525 1.197 - PASSED 4550 1.215 - PASSED 4575 1.233 - PASSED 4600 1.25 - PASSED 4625 1.27 - PASSED 4650 1.287 - PASSED 4675 1.306 - PASSED CCX#1: 4675MHz 10:55:02: Stress test stopped. 10:55:02: Test 4 finished! Phoenix ready! 10:55:02: CO (CORE) test started! 10:55:11: Step: 1 CORE#1 BASE FREQ: 4700MHz REAL FREQ: 4700MHz 10:55:11: Saving intermediate values... 10:55:19: Test#1 CORE#1 CO: -11 DELTA: -4 TEMPERATURE: 68°C 10:56:00: Saving intermediate values... 10:56:10: Test#2 10:56:52: Saving intermediate values... 10:57:02: Test#3 10:57:47: Step: 2 CORE#1 BASE FREQ: 4700MHz REAL FREQ: 4725MHz 10:57:47: Saving intermediate values... 10:57:55: Test#1 CORE#1 CO: 17 DELTA: 28 TEMPERATURE: 68°C 10:58:37: Saving intermediate values... 10:58:47: Test#2 10:59:28: Saving intermediate values... 10:59:38: Test#3 11:00:24: Step: 3 CORE#1 BASE FREQ: 4700MHz REAL FREQ: 4750MHz 11:00:24: Saving intermediate values... 11:00:32: Test#1 CORE#1 CO: 44 DELTA: 27 TEMPERATURE: 68°C 11:01:13: Saving intermediate values... 11:01:23: Test#2 11:02:04: Saving intermediate values... 11:02:14: Test#3 11:03:00: Step: 4 CORE#1 BASE FREQ: 4700MHz REAL FREQ: 4775MHz 11:03:00: Saving intermediate values... 11:03:08: Test#1 CORE#1 CO: 80 DELTA: 36 TEMPERATURE: 69°C 11:03:50: Saving intermediate values... 11:04:00: Test#2 11:04:41: Saving intermediate values... 11:04:51: Test#3 11:04:53: Stress test stopped. At this frequency CORE#1 is unstable, the frequency is reduced. CORE#1 BASE FREQ: 4700MHz REAL FREQ: 4750MHz 11:04:53: Saving intermediate values... 11:05:01: Test#1 CORE#1 CO: 75 DELTA: 31 TEMPERATURE: 70°C 11:05:43: Saving intermediate values... 11:05:53: Test#2 11:06:34: Saving intermediate values... 11:06:44: Test#3 Results CO testing 11:07:26: Saving intermediate values... 11:07:26: Saving intermediate values... 11:07:26: Saving intermediate values... CORE| CPPC | BASE | RES | TEMP | CO C01 | 129 | 4700 | 4750 | 70 | 75 C02 | 133 | 0 | 0 | 0 | 0 C03 | 140 | 0 | 0 | 0 | 0 C04 | 140 | 0 | 0 | 0 | 0 C05 | 137 | 0 | 0 | 0 | 0 C06 | 125 | 0 | 0 | 0 | 0 11:07:39: Step: 5 CORE#2 BASE FREQ: 4700MHz REAL FREQ: 4700MHz 11:07:39: Saving intermediate values... 11:07:47: Test#1 CORE#2 CO: -12 DELTA: -2 TEMPERATURE: 67°C 11:08:28: Saving intermediate values... 11:08:38: Test#2 11:09:19: Saving intermediate values... 11:09:29: Test#3 11:10:15: Step: 6 CORE#2 BASE FREQ: 4700MHz REAL FREQ: 4725MHz 11:10:15: Saving intermediate values... 11:10:23: Test#1 CORE#2 CO: 17 DELTA: 29 TEMPERATURE: 68°C 11:11:04: Saving intermediate values... 11:11:15: Test#2 11:11:56: Saving intermediate values... 11:12:06: Test#3 11:12:51: Step: 7 CORE#2 BASE FREQ: 4700MHz REAL FREQ: 4750MHz 11:12:51: Saving intermediate values... 11:13:00: Test#1 CORE#2 CO: 41 DELTA: 24 TEMPERATURE: 67°C 11:13:41: Saving intermediate values... 11:13:51: Test#2 11:14:32: Saving intermediate values... 11:14:42: Test#3 11:15:28: Step: 8 CORE#2 BASE FREQ: 4700MHz REAL FREQ: 4775MHz 11:15:28: Saving intermediate values... 11:15:36: Test#1 CORE#2 CO: 70 DELTA: 29 TEMPERATURE: 67°C 11:16:17: Saving intermediate values... 11:16:27: Test#2 11:17:09: Saving intermediate values... 11:17:19: Test#3 11:18:04: Step: 9 CORE#2 BASE FREQ: 4700MHz REAL FREQ: 4800MHz 11:18:04: Saving intermediate values... 11:18:12: Test#1 CORE#2 CO: 103 DELTA: 33 TEMPERATURE: 68°C 11:18:54: Saving intermediate values... 11:19:04: Test#2 11:19:45: Saving intermediate values... 11:19:55: Test#3 11:20:09: Stress test stopped. At this frequency CORE#2 is unstable, the frequency is reduced. CORE#2 BASE FREQ: 4700MHz REAL FREQ: 4775MHz 11:20:10: Saving intermediate values... 11:20:18: Test#1 CORE#2 CO: 78 DELTA: 8 TEMPERATURE: 66°C 11:20:59: Saving intermediate values... 11:21:09: Test#2 11:21:50: Saving intermediate values... 11:22:00: Test#3 Results CO testing 11:22:42: Saving intermediate values... 11:22:42: Saving intermediate values... 11:22:42: Saving intermediate values... CORE| CPPC | BASE | RES | TEMP | CO C01 | 129 | 4700 | 4750 | 70 | 75 C02 | 133 | 4700 | 4775 | 66 | 78 C03 | 140 | 0 | 0 | 0 | 0 C04 | 140 | 0 | 0 | 0 | 0 C05 | 137 | 0 | 0 | 0 | 0 C06 | 125 | 0 | 0 | 0 | 0 11:22:56: Step: 10 CORE#3 BASE FREQ: 4775MHz REAL FREQ: 4775MHz 11:22:56: Saving intermediate values... 11:23:04: Test#1 CORE#3 CO: -17 DELTA: 4 TEMPERATURE: 69°C 11:23:45: Saving intermediate values... 11:23:55: Test#2 11:24:36: Saving intermediate values... 11:24:46: Test#3 11:25:32: Step: 11 CORE#3 BASE FREQ: 4775MHz REAL FREQ: 4800MHz 11:25:32: Saving intermediate values... 11:25:40: Test#1 CORE#3 CO: 5 DELTA: 22 TEMPERATURE: 69°C 11:26:21: Saving intermediate values... 11:26:31: Test#2 11:27:13: Saving intermediate values... 11:27:23: Test#3 11:28:08: Step: 12 CORE#3 BASE FREQ: 4775MHz REAL FREQ: 4825MHz 11:28:08: Saving intermediate values... 11:28:16: Test#1 CORE#3 CO: 24 DELTA: 19 TEMPERATURE: 69°C 11:28:37: Stress test stopped. At this frequency CORE#3 is unstable, the frequency is reduced. CORE#3 BASE FREQ: 4775MHz REAL FREQ: 4800MHz 11:28:37: Saving intermediate values... 11:28:45: Test#1 CORE#3 CO: 3 DELTA: -2 TEMPERATURE: 68°C 11:29:27: Saving intermediate values... 11:29:37: Test#2 11:30:18: Saving intermediate values... 11:30:28: Test#3 Results CO testing 11:31:10: Saving intermediate values... 11:31:10: Saving intermediate values... 11:31:10: Saving intermediate values... CORE| CPPC | BASE | RES | TEMP | CO C01 | 129 | 4700 | 4750 | 70 | 75 C02 | 133 | 4700 | 4775 | 66 | 78 C03 | 140 | 4775 | 4800 | 68 | 3 C04 | 140 | 0 | 0 | 0 | 0 C05 | 137 | 0 | 0 | 0 | 0 C06 | 125 | 0 | 0 | 0 | 0 11:31:23: Step: 13 CORE#4 BASE FREQ: 4750MHz REAL FREQ: 4750MHz 11:31:23: Saving intermediate values... 11:31:31: Test#1 CORE#4 CO: -18 DELTA: 0 TEMPERATURE: 69°C 11:32:12: Saving intermediate values... 11:32:22: Test#2 11:33:04: Saving intermediate values... 11:33:14: Test#3 11:33:59: Step: 14 CORE#4 BASE FREQ: 4750MHz REAL FREQ: 4775MHz 11:33:59: Saving intermediate values... 11:34:07: Test#1 CORE#4 CO: 4 DELTA: 22 TEMPERATURE: 69°C 11:34:49: Saving intermediate values... 11:34:59: Test#2 11:35:40: Saving intermediate values... 11:35:50: Test#3 11:36:36: Step: 15 CORE#4 BASE FREQ: 4750MHz REAL FREQ: 4800MHz 11:36:36: Saving intermediate values... 11:36:44: Test#1 11:36:45: Stress test stopped. At this frequency CORE#4 is unstable, the frequency is reduced. CORE#4 BASE FREQ: 4750MHz REAL FREQ: 4775MHz 11:36:45: Saving intermediate values... 11:36:53: Test#1 CORE#4 CO: 5 DELTA: 1 TEMPERATURE: 69°C 11:37:35: Saving intermediate values... 11:37:45: Test#2 11:38:26: Saving intermediate values... 11:38:36: Test#3 Results CO testing 11:39:18: Saving intermediate values... 11:39:18: Saving intermediate values... 11:39:18: Saving intermediate values... CORE| CPPC | BASE | RES | TEMP | CO C01 | 129 | 4700 | 4750 | 70 | 75 C02 | 133 | 4700 | 4775 | 66 | 78 C03 | 140 | 4775 | 4800 | 68 | 3 C04 | 140 | 4750 | 4775 | 69 | 5 C05 | 137 | 0 | 0 | 0 | 0 C06 | 125 | 0 | 0 | 0 | 0 11:39:31: Step: 16 CORE#5 BASE FREQ: 4700MHz REAL FREQ: 4700MHz 11:39:31: Saving intermediate values... 11:39:39: Test#1 CORE#5 CO: -4 DELTA: 4 TEMPERATURE: 73°C 11:40:20: Saving intermediate values... 11:40:30: Test#2 11:41:11: Saving intermediate values... 11:41:21: Test#3 11:42:07: Step: 17 CORE#5 BASE FREQ: 4700MHz REAL FREQ: 4725MHz 11:42:07: Saving intermediate values... 11:42:15: Test#1 CORE#5 CO: 25 DELTA: 29 TEMPERATURE: 74°C 11:42:56: Saving intermediate values... 11:43:06: Test#2 11:43:48: Saving intermediate values... 11:43:58: Test#3 11:44:43: Step: 18 CORE#5 BASE FREQ: 4700MHz REAL FREQ: 4750MHz 11:44:43: Saving intermediate values... 11:44:51: Test#1 CORE#5 CO: 51 DELTA: 26 TEMPERATURE: 73°C 11:45:33: Saving intermediate values... 11:45:43: Test#2 11:46:24: Saving intermediate values... 11:46:34: Test#3 You have unique processor, 2 CCD's! 11:47:16: CMD_REJECTED_PREREQUISITE 93 Phoenix deactivated!

Anonymous

it's never happen with ctr

Morgan Ring

Hi, Im still a little hesitant about downloading and using Hydra, it just seems way more complicated than CTR where I could just hit 'Diagnostics' and let it do it's thing then save and load up the profiles and away you go. For someone like me,(beginner) reading through the pdf is pretty daunting, having to change this or enable that etc. Anyways, it's definitely out of my comfort zone but wish you all the best!

1usmus

HYDRA is definitely a more complicated product, but you can get used to it. At this point, I'm debugging and calibrating the basic settings to make it easier for you to interact with the app. The principle of diagnostics is similar to CTR, we enter the desired voltages or select presets and then just run the diagnostics. You don't need to do more than that. There will also be a video tutorial.

Anonymous

DOPE! So I did the upgrade to Hydra 09D two days ago and the pc has been running superbly. I didn't re-run diagnostic, just inputted my original results from 09A and got the new one running. I love the settings that you've implemented in this release, Yuri. Thank you again for your genius and vision. Yes, Hydra is more difficult than CTR, but so what. This is for enthusiast, and it'll benefit you after you've taken the time to adjust to all the features. I love being able to change a lot of the parameters in order to maximize my particular cpu. Great work.

Anonymous

I am getting fairly frequent reboots caused by my strongest core that seem to occur when the processor goes from 3 cores active and basically idle. I run a batch file to convert some files to CHD format I have included a sample below: if exist "%%~na.cue" 1-chdman createcd -i "%%~na.cue" -o "%%~na.chd" if not exist "%%~na.chd" echo Problem detected, CHD not create, check file names and cuesheet if not exist "%%~na.chd" pause if exist "%%~na.chd" echo CHD created and verified, deleting old archive if exist "%%~na.cue" del "%%~na.cue" if exist "%%~na*.bin" del "%%~na*.bin" (It usually crashes after this delete step) if exist "%%~na*.zip" del "%%~na*.zip" I have the CO for this core now set at -111 for low thread rules, which is diagnosed to be +1 when the diagnostic is run. This is quite a drastic drop and has me wondering if it's actually doing any good. I have the same problem with CTR that seems to be fine until idle or very light loads occur such as opening a browser. Is there something I can do, and is my strongest core running at -111 doing some long term damage? Also, can you clarify if a negative CO offset in Hydra increases or decreases the voltage, as to get cores stable you usually add voltage, not reduce it when doing the old trial and error manual overclock.

Anonymous

Great job Yuri 👌 I am following you for few years now and I am glad you are just getting better at this 🙂

Anonymous

I got an IO exception error and it stopped :(

Anonymous

Seems stable now at -111, just need to know if that's safe to run constantly on a 5950x

Xdrqgol

this version for me is not working at all good... crashes randomly, black screen, 00 code on Dark HERO... Normal settings ( not OC ,XOC) Did a fresh diagnostic as i have had a fresh install of windows 10. I will wait better versions... i am still back to the BIOS CO which has always been super stable with 5096mhz effective clock which i could have never got in CTR or HYDRA...

Anonymous

This might have been my fault, new folder and forgot to add to virus exclusions. Made a new root folder for Hydra and excluded entire folder. Testing again tonight.

James Tiller

Running very well on OC settings, 5950x and Dark Hero. Games check, benchmarks check, zero crashes. Win11 and Win10

Anonymous

always get cinebench aplication error when i start diagnostic, why is that happening?

Anonymous

System crashes during sleep and hibernation, but works well at other times. The CO has been adjusted according to the prompts. Why is this?

Anonymous

Could you add a new feature: automatically deactivate profiles before sleep or hibernation?

1usmus

Please write more information about the message you receive or send a link to a screenshot.

1usmus

Glad to hear you're doing well, buddy. The next phase is to get more performance at the same frequency. The fix is already in the testing phase :)

Anonymous

Same here, Hydra is definitely AMAZING! 0.9D was the one for me, never had any crashes or instability unlike the ones i had in CTR. provides perfect performance and lowered about 10 degrees on the CPU!

Peter

I managed to complete a full diagnostics and it seems to work well. However, since the 0.9D update the application does not always automatically startup any more upon starting windows despite having it ticked ''on'' in the program.

Anonymous

I ran the diagnostic, it completed, and after reviewing the results I've noticed the CO values resemble what you'd find within Curve Optimizer much more than in the earlier versions... How do they correlate with Curve Optimizer values? 1:1? They don't?

Rory O'Donnell

I finally managed to get a full diagnostic run completed. Tried cb, black screen instantly

iNerd

I just got the this Error below: ***HYDRA 0.9D PRO by 1usmus*** AMD Ryzen 9 5900X 12-Core Processor ASUS ROG CROSSHAIR VIII DARK HERO BIOS ver. 3703 SMU ver. 56.53.00 TABLE ver. 3672069 DRAM speed 3800 MHz Information about the last failure Unstable frequency, APIC ID: 2 15:11:35: MODE RSMU CMD 5 ARG 0 If the failure occurred in a game: Decrease by 30 the values for (GAME CO offset, SETTINGS tab). If the failure occurred in a multi-threaded load (CPU usage > 70%): In the CO table #2 reduce the CO for CORE#2 by 10. Since i could identify the issue under this one: If the failure occurred in a low-threaded load (CPU usage < 70%): In the CO table #1 reduce the CO for CORE#2 by 20. i went under First Table and decreased only Core #2 by 20 (Core 2: -20 - enter). Run same process again - issue solved. I just instant stoped by here to say Thanks Yuri for the great tool and thinking on so much things we could run in. I'm on Hydra since release and did not needed to touch anything until now.

Anonymous

is this a hydra alpha release still or is this official now? sorry im new to the hydra stuff.

Anonymous

I'm using the asus darkhero board. However, if the power option is set to llc3, vdroop is 2.2%. On the other hand, llc5 causes -0.4%. Is it better to set it to llc5 in this case?

Anonymous

For people using MSI MEG X570 ACE or similar, LLC5 brings better overclocking results for me, at least CCD1 went from Gold to Platinum. You might want to try LLC5 instead of LLC4 ;)

Anonymous

is hydra compatible with threadripper PRO 3975WX?

Anonymous

Yuri, any chance you could provide some clarification on how CO as displayed in Hydra related to CO values you'd enter in the BIOS?

Anonymous

Just joined and installed Hydra after using the public release of CTR 2.1 very successfully. Wow, Hydra is amazing, the ability to quickly switch between voltage profiles and adjust the equivalent of CO on the fly while in OS is excellent! For all who are having crashing in Diagnostics, you need to make sure you set your PPT, TDC, EDC and temperature limits to a reasonable level so that you don't trip the CPU overtemperature protection through temperature spikes that could happen during diagnostics. If you currently have your PBO power limits set to Motherboard in the bios, please set it back to a reasonable level. I use PPT 200W, TDC 140A, EDC 160A. Once you finish diagnostics and Hydra is controlling your CPU fine, you can set your power limits back to what you had before.

Anonymous

I use LLC3. Didn't have issues even with vdroop of 2.2%

James Tiller

just a check-in with some more testing, 5800X on ASRock x570m pro4, 5900X on Strix b550i. On OC diagnostics. All running balls to the walls after several days, no issues or crashing so far. The 5900X took around 6 hours and crash/rebooted on final test but running perfectly after restart.

Gabriel

Is the recommendation to leave all voltages auto in BIOS and only set PBO power and LCC?

Anonymous

Leave CPU core voltage to Auto, and have LLC set based on the guide. I still use vSOC and other voltages manually.

Anonymous

Had to read this twice to understand you meant the BIOS settings not Hadra and was wondering how Hydra should override BIOS :D Nonetheless PBO Auto still seems to work best for me, at least in Diagnostics. Haven't reached PPT Limit of 140 W ever in diagnostics, only EDC is allways maxed, so I guess your hint with EDC 160 A might be a good advice if you don't want to go with Auto. As an example with LLC5, 200 W, 170 A, 170 A I'm losing 25-75 MHz per Core in Diagnostic Mode at least on the higher rated Core (Gold/Platinum Sample). For daily use, mostly Games modified PBO helps me, since I'm running into limits otherwise and my CPU will throttle.

Anonymous

If I am CPU Mining I get this error with XMRig: failed to start winring0 driver error 183 does Hydra use winring0 driver?

Anonymous

If I am GPU Mining I am unable to disable gaming mode with the enabled and dynamic check marks unchecked. Is this expected behavior? I thought if I unchecked these it would allow me to return to normal profile behavior, since unchecking does not work I instead just change the GPU memory trigger points.

Anonymous

Coming from CTR, just gave it a shot, and it worked great out of the box: normal voltage-> diagnostic-> a bit of CO offset on two profiles -> DONE ! Keep up the good work Sir.

Anonymous

Drewling for 1.0 release!

Anonymous

So far this has been working really good for the most part. But it seems like every time my system tries to wake from sleep it causes a crash. I've tried adjusting what was recommended to get around this, but it still crashes about 80% of the time waking from sleep.

James Tiller

one point oh......one point oh....one point oh... :)

Anonymous

Once installed, if you exit the application does the system go back to PBO or remain on active profile until restarted?

Anonymous

Recently, Hydra will display 100000 in the values for the cores and the system temp will climb to 70C. The issue goes away when I restart Hydra.

Anonymous

The system will go to an overclock profile. I do not PBO is involved as I have turned that off. I think it may be the OC or XOC profile kicking in.

Anonymous

Same here. I keep lowering the CO values for the low thread count cores the diagnostics indicated failed (reduce by 20 each time), but there seems to be no end in sight. I keep getting crashes coming out of sleep and I keep reducing whatever core CO value as directed. At this point I'm really far off from the original values the diagnostics came up with for some of the cores.

Anonymous

Since installing the new AMD Radeon driver 21.9.1, I've been getting these warnings from Hydra every few minutes: "MODE RSMU CMD 5 ARG 0" or "Hash sum of the command does not match the reference. Re-sending the command". In this driver update at Radeon you can now also tune your AMD CPU, but I haven't changed it and left it to standard. Can you help me?

Anonymous

means they probably adjusted timings and it's not compatible yet with the current release . Personally i dont always update to the latest amd drivers unless its a major bug fix or a critical one. im still on 21.6.1 with specific PCIE versions updated. 21.9.1 is giving a lot of people that do a lot of AVX related benching or work.

Anonymous

y wont hydra.exe open i have archiver and framework

Anonymous

Just tried Hydra and i have to say it works fine so far, weird is that i have a lower score (like +- 100) in Cinebench R20 and R23 then with CTR but i have higher scores in CPUZ specially in the single thread score. Also my 3dMark cpu score went up and games feel smoother.