Видео

Thanks for sharing. It's a shame ADATA never issued any recall or whatever. They must have known at this point. Kinda like Intel tried to sweep the situation with their 13th and 14th gen under the rug. ADATA was apparently more successful in doing so. Hope you didn't lose any important data.

Cheers for sharing

Thanks for sharing. Hydra is an OC sandbox. Yes, lots of things can work with a click of a button, but manual user intervention is often required. Impressed by your results, though.

Are you reading temperature from the same sensor? An SSD usually has a few. Not sure how it's done in Linux tbh.

@@savvylayman Using Psensor it says Sensor 1 and 2 on the SSD are from 64c to 70c while on youtube.

@@FutureEyesOnly-be3do That is really strange. Even if it is a system drive, I would not expect such high temperature while just watching RUclips. Does it have a heatsink with thermal pad properly installed? What kind of drive is it? Anyways, maybe the readings are wrong?

@@savvylayman Just found the problem. VMD in the bios was enabled causing everything to get hot for some reason. When i shut it off and rebooted temps dropped off to a very normal level.

Thanks for asking. It's actually smarter to go with your own limits because you can fine-tune them to match your VRM / cooling configuration. I myself optimized limits on my R9 5900X to squeeze as much performance as possible but with temperature not exceeding 80 °C under full-core (non-AVX) load, such as video rendering.

Thanks for your comment. What makes you think this was the case? What's your assumption regarding such behavior?

i get this with curveoptimizer enabled on every setup, when i disable it ,everything is fine ,must be a windows problem , cuz i hadn't this problem before idk :)

@@Lolliswonderland I see. I think this is caused by so-called clock stretching. Due to too aggressive CO curve, the CPU attempts too boost too high for the given voltage. The supplied voltage is not too low to immediately crash the system, but low enough so the CPU has to drop the clocks significantly. I think this whole procedure can take a fraction of a second. In such case, I would recommend to lower the negative volttage offset.

@@savvylayman so you think the auto setup makes clock stretching by -1-20 negativ on all cores in the curveoptimizer on 2 Systems with different boards and cpu's?

@@Lolliswonderland Don't know to be honest. I would need to see your system in person. By the way, are these Asus mobos? If so, make sure you disable PBO Fmax Enhancer: ruclips.net/video/csGkp4pVuO0/видео.html

Thanks for letting me know about the new CoreCycler version and kudos to your CPU for being able to handle serious undervolting.

Perhaps it made some sense in Ryzen 3000 series. But who still owns a R5 3600 these days (kudos to those who do)? The setting should definitely be disabled by default.

@@savvylaymanbro what? There are people still out there with ryzen 2600’s plenty of people are still rocking 2000/3000 cpu’s

@@slaydog5102 If you're happy with your R5 2600, you have earned my respect.

Kingston KF432S20IBK2/32: www.amazon.fr/Kingston-dordinateur-Portable-KF432S20IBK2-32/dp/B097QJ25NY/ref=sr_1_1?__mk_fr_FR=%C3%85M%C3%85%C5%BD%C3%95%C3%91&crid=199DT1V1X5QKJ&dib=eyJ2IjoiMSJ9.zrgqFqOgFdMhBacjSYJPPhqLIHbCbwHUMp1PHnTSQ-dzfrSvtC6fUiUXXh7CtqNMKgpAjJu9gKZISwH0e_G_tUAxqtP87mGeyiGTRlPFnLy2nlCika61GVUy2bwHboXvZvopb8WZTvSmAtXLS-1A05s8qSADLFsXrYkWyXlg2m5SFtqrjBRMkm4HnM4GUQXzfbXq6oJnWTVZXDIbplVfcazbcWNEnpjkcw-D5-lupLu7MvDKMq1en2Zlw8T_0stvaKEIciQZ1Hp1kwjLzNkC51CUJgT4TjlG9dbHMH8FiL4.KHNZmqkd-UDIdInDBrhj5XVYuge9wLF_wpJC_xc61wM&dib_tag=se&keywords=Kingston+KF432S20IBK2%2F32&qid=1720010027&sprefix=kingston+kf432s20ibk2%2F32%2Caps%2C498&sr=8-1

Bonjour, j'espère que le lien vous aidera. Désolé, je ne parle pas français, j'ai donc utilisé le traducteur AI pour m'aider. Je vous souhaite bonne chance.

@@savvylayman merci vous êtes mon sauveur. Au fait je remarque aussi mon pc fait de temps en temps des freezes de 1 second. Sait tu ce qui cause ça ?

Did you struggle with the initial start or with the diagnostics? I mean it's an overclocking sandbox tool and I have to admit it's not completely user-friendly.

Do you have a heatsink installed on that SSD? I mean it's not a tragedy but those download temps are a bit toasty. Maybe try to copy huge files here and there to see if it gets any worse. If it does, you may want to consider options. But unless you copy huge files very regularly, it's probably fine.

@@savvylayman ​I went digging to see what was up. There is no heatsink over that specific slot. I'll need to erase both drives, turn off the raid, and reinstall windows to relocate that drive to a different slot with a heatsink. It's living as is for now because it only gets super hot when downloading huge files like a big game on Steam. Thankfully, I'm not doing that often.

@@cylerner I do recommend moving the SSD to another slot where you can install a heatsink. There may be a very small latency penalty if the PCIe lanes come from the chipset instead of the CPU (at least on AMD platfom), but I would say it's negligible. Good luck and cheers for your reply.

See, I found a good (and only) use for it - as a prop. :o)

Thanks for asking. There is no difference in temperature at all. I even did another upgrade to 32 GB and still works like a charm: ruclips.net/user/shortsejCSIPfx0pw?feature=share

​@@savvylaymansur le même pc ? Peut tu m'envoyer le lien

@@gocheikhcamara4461 ​ Kingston KF432S20IBK2/32: www.amazon.fr/Kingston-dordinateur-Portable-KF432S20IBK2-32/dp/B097QJ25NY/ref=sr_1_1?__mk_fr_FR=%C3%85M%C3%85%C5%BD%C3%95%C3%91&crid=199DT1V1X5QKJ&dib=eyJ2IjoiMSJ9.zrgqFqOgFdMhBacjSYJPPhqLIHbCbwHUMp1PHnTSQ-dzfrSvtC6fUiUXXh7CtqNMKgpAjJu9gKZISwH0e_G_tUAxqtP87mGeyiGTRlPFnLy2nlCika61GVUy2bwHboXvZvopb8WZTvSmAtXLS-1A05s8qSADLFsXrYkWyXlg2m5SFtqrjBRMkm4HnM4GUQXzfbXq6oJnWTVZXDIbplVfcazbcWNEnpjkcw-D5-lupLu7MvDKMq1en2Zlw8T_0stvaKEIciQZ1Hp1kwjLzNkC51CUJgT4TjlG9dbHMH8FiL4.KHNZmqkd-UDIdInDBrhj5XVYuge9wLF_wpJC_xc61wM&dib_tag=se&keywords=Kingston+KF432S20IBK2%2F32&qid=1720010027&sprefix=kingston+kf432s20ibk2%2F32%2Caps%2C498&sr=8-1

Thanks for asking. In the past, there was not much overclocking possibilities available for the 3D chips. For that reason, Hydra didn't support them either. This is what Hydra's online guide says regarding supported CPUs these days, "All desktop CPUs based on Zen 3 and Zen 4 architectures." docs.google.com/presentation/d/1_j6vB61koOVwEHLrvqt7lh1RYw4XRuoTIeEDtYmnFj0/edit#slide=id.g13b36094eab_3_0

Thanks for asking. It's not great, but I wouldn't worry about it too much.

When I was younger, I used to think spending premium on a PSU was a waste of money. I've learned my lesson, too.

That's a really good bin then. My personal guess is this was supposed to be a more premium CPU, but it failed some validation tests, so AMD disabled some cores and lowered the overall power budget. Good for you!

@@savvylayman yes, it has 2 disabled cores, the specs resembles a R7 5700.

Thanks for asking, @Dont-Evan-Bother. Samsung 970 EVO is a PCIe gen 3 drive, whereas Seagate Lightsaber is a PCIe gen 4 drive. Gen 4 drives do run hotter due to higher compute load on their ASIC, which results in a significantly higher throughput. I suppose you measured the difference of 30 °C on the ASIC controller, correct? This is usually the highest temperature reported by the drive. On 970 EVO, it is called Drive Temperature 3 in HWiNFO. May I ask what temperature you actually read? Even if the drive reaches 90 °C for short periods of time, I wouldn't worry it too much. However, the drive's lifespan could decrease if it runs this hot for 24*7. Concerning the heat sink touching the backplate of 7900 XTX, it's not ideal. That GPU dumps tons of hot air right onto the NVMe SSD. What motherboard do you use? My Asus mobos based on X570 chipset support PCIe 4.0 specs across all NVMe slots. If this is the case with your motherboard, I would recommend moving the drive to a lower slot. Yes, there is a small latency penalty but nothing significant. Oh, and my assumption is your case provides enough airflow.

Evidence?

@@daghtus on the 5800x it doesnt at least

Well it is highly effective on my 7800x3d, and for people who say you can not go past -30 offset is delusional. Temporarily setting it to -40 gives me higher clocks

Do you want to go for eco mode or full power? Because 7000 series have pretty generous limits set out of the box. What cooler do you use?

@@savvylayman at the moment i didnt change my cooler yet im using wraith stealth cooler and im setting thermal throtle to 85C + i undervolted all cores to negative 22

@@cjentuskuzan4027Thanks for sharing. With stock cooler, I would definitely go with eco mode. www.pcworld.com/article/1352253/how-to-enable-eco-mode-with-ryzen-7000.html For serious overclock, I recommend at least an AIO with 240 mm radiator. By the way, negative 22 curve for all cores seems too aggressive in my opinion. You would have to be very lucky to be able to shift the curve so much with the best cores. Anyways, if you encounter crashes, just decrease the value. Cheers and good luck.

@@savvylayman its my first time undervolting a cpu and im undervolting it and testing by occt if i dont get any errors i consider it as a good undervolt . Im i right or wrong ? Cause i started from 30 and i had bsod and crashes but in negative 22 everything seems to be fine .

One thing to mention im using the integrated graphics at the moment and im having very high latency all of the sudden in latencymon due to directx graphics kernel can it be due to the undervolt ?

If making an uneducated guess, how do you then verify the system's stability? I rather choose a structured approach if you ask me.

I paid EUR 200 for them and that's a steal!

@@savvylaymani was check in another options and in my own opinión have high quality materials make It really good quality about de sound the only think is maybe the Max volumen can be better for the price honestly. But i can use this in my Xbox series x and also i can wach movies o listen music in my phone and answer calls so is pretty for me. Anothers is only for play amd thats It. All the BEST dude and regards from Mallorca Spain💪

@@SORKTAG1Thanks for sharing. The B&O Portals headphones are really well built. The build quality reminds me of my older Bowers & Wilkins PX, which imho are still stunning today. There are many things I love about B&O Portals, but truth be told, there are two downsides I dislike. 1. When PC is connected via the dongle & smartphone is connected via Bluetooth at the same time and one loses connection (e.g., dongle because I go too far), the other connection disconnects as well. This can go over and over until I go back to the PC or disconnect it manually. 2. Touch controls are not great. They are placed at a spot where I sometimes accidentally touch them when I put the headphones on or take them off. I would prefer physical buttons like B&W PX have. Cheers for reading this far. :) Greetings from Slovakia. Hope I can visit Mallorca one day. Summer vibes. :)

DRAM calculator worked alright for selected popular OC dies. The tool as such is technically a lookup table, not a calculator. I admit it had a typo in RTT_WR due to which it didn't work with my B-die right off the bat. Overclocking is alchemy to some extent. Even Hydra doesn't provide me with a 100% stable overclock with default settings. But it's close enough and I can do just final fine tuning.

Perhaps you are lucky and your drive is not affected. But who can tell for sure? I'd rather be careful. I've also heard from a colleague of mine his SX8200 Pro developed very many bad blocks, too.

so my thermals seem fine, the hottest i could get it was 56 deg, in hw info. i think that is controller temp, other 4 temps were at 44C max (nand?), and one other temp (dram?) was at 40c max. though I ran the drive in pce g2 x4 mode, because i dont have that many lanes available. speed is ok-ish, 1.5 gig/s read and write seq. conditions: in a pce expansion card below an rx590, with cheapo heatsinks zip-tied to it with thin thermal pads and some thermal paste. lifespan? time will tell. @@savvylayman

Thanks for sharing. That is very interesting. Haven't tried Silicon Power drives yet, but after what I went through with ADATA, I think I'll rather go with a premium manufacturer.

Thanks for asking. Actually, I use an old Zyxel NAS540 as a target backup device. It is rather slow and not very sophisticated, so I'm just presenting a bunch of SMB shares on RAID5 and write to these.

Yeah because backups should be necessary for the basic function of storage. 🙄

For how many years have I worked as backup engineer? Oops! 0:-)

@@allan2665 Yeah, my backup designed was bad, but unreliability of this particular SSD model is truly remarkable.

For that money, it's a steal! Congratulations. I have just finished playing Alan Wake 2 on my RX 6950 XT and it worked like a charm in 4K.

Im also running a -28 on my 7950x3d on all cores

RyzenMaster reportated stability at -30 on all cores on my Ryzen 5

Thanks for asking. WHEA *should* be logged every time Windows kernel detects a hardware error. Well, in my experience, it doesn't always do that. When your system crashes, I would choose one of the following options: 1. Decrease negative offset by 2 points (i.e., increase voltage) for the best cores. These are in most cases the ones, which struggle, because they want to boost high. 2. Decrease negative offset for all cores by 2 points. 3. Use Hydra to determine the best curve. But I would recommend to stick to classic PBO2 curve instead of going for Hybrid OC. For some reason, HOC acts strange in the last two updates or so.

@@savvylayman Thanks I'll give Hydra a look. I right now ended up without going thru a lot of testing, set the 6 cores that are the best cores to -15 and the rest to -25. Guessing basic activities like file copies and the like will mostly use those "best" cores?

@@jriker1Yes, the "best" cores are almost always preferred by the scheduler so if there is a crash, they are most likely to blame.

I agree, the utility is apparently aimed at hardcore enthusiasts. I also sometimes find it hard to decrypt what the individual settings mean. And truth be told, I had trouble with the last two versions 1.5C & 1.5D. 1.5E has just come out so I'm going to give it a try with fresh CPU diagnostics. I will most likely shoot a video update.

@@savvylayman most 7800X3D results that I’ve seen from the discord and Reddit are “negligible”