orl
Regular Member
Posts: 96
|
Post by orl on Feb 19, 2019 9:33:00 GMT -5
Also, you guys might want to check this out. There is someone out there who was testing degradation on his Ryzen chips with voltage variations. While this is very informal, and thus likely to be taken with salt, it brings questions to light that warrant investigation. These chips may degrade very quickly if running higher voltages despite having sufficient cooling. I would be curious to see how this develops. According to him, his chips running 1.375 and below haven't begun degrading, while above that point they show signs of degradation within a 6 month period. The issue though with this type of test is that we don't know if his power states are stable. He may be running piss PSU setups with poor stability.
https://www.reddit.com/r/overclocking/comments/amtnt4/ryzen_generation_2_safeunsafe_voltages_tested/
|
|
|
|
Post by ShrimpBrime on Feb 19, 2019 16:36:33 GMT -5
Not to sure on the validity of Cpu degradation as I've been using mine with a V-core of over 1.4v since I've gotten it. So about 5 months now. I know the boost clocks are over 1.4v stock.... So far, no issues with the processor running an overclock. It acts the same as it did once I figured out how it wants to be tuned. Always more voltage. more more more. I've been running the stock cooler also. Really hope mine doesn't degrade in just a few short months time while running maybe 0.1v over stock turbo voltage. Right now, 4.1ghz static OC at 1.405v. No issues. Been folding about 18 hours straight. PROOF:: folding.extremeoverclocking.com/team_summary.php?s=&t=234770 |
|
|
|
Post by Mr.Scott on Feb 19, 2019 16:41:08 GMT -5
Also, you guys might want to check this out. There is someone out there who was testing degradation on his Ryzen chips with voltage variations. While this is very informal, and thus likely to be taken with salt, it brings questions to light that warrant investigation. These chips may degrade very quickly if running higher voltages despite having sufficient cooling. I would be curious to see how this develops. According to him, his chips running 1.375 and below haven't begun degrading, while above that point they show signs of degradation within a 6 month period. The issue though with this type of test is that we don't know if his power states are stable. He may be running piss PSU setups with poor stability. https://www.reddit.com/r/overclocking/comments/amtnt4/ryzen_generation_2_safeunsafe_voltages_tested/ Link taken with a big grain of salt. I know who most of those guys are. They're pompous jack ass's. |
|
|
|
Post by ShrimpBrime on Feb 19, 2019 16:44:08 GMT -5
Come to think of it ORL, this would be a good new topic. I'd be happy to provide as much information as possible. I could probably replicate every screen shot that's in this thread including the IBT tests I've already run weeks ago. I know exactly what I can do with this chip and have basically 3 or 4 daily setups that I run depending on how I feel that day honestly......
|
|
orl
Regular Member
Posts: 96
|
Post by orl on Feb 19, 2019 17:36:00 GMT -5
I'd say that this could be a potential item to watch. The only reason I recommended watching this vs the usual brush to the side is that he is not the first person to report the issue. While we all know parts degrade in general, like motherboard VRMs and so forth, (if they didn't electronics would last forever) we as a group will need to watch the chips themselves to record anything of substance. My Ryzen exposure is very new currently, and I have yet to even begin trying to push for my upper limits. So my sample size is obviously useless currently. Unfortunately this CPU will likely be replaced by the next gen 8/16 or 12/24 if the VRMs can handle the big one. So to me this information is being taken two ways.
1. I might not get this chip to a point where any degradation occurs...
2. Challenge Accepted.
|
|
|
|
Post by ShrimpBrime on Feb 19, 2019 18:06:04 GMT -5
Lol. I already picked number 2. Always do... reasons:
A. Was taught that auto overclocking and software overclocking is weak.
B. Overclock manually only. I always know where my cpu frequency and voltage is truly at. It is known that actual voltage and software readouts can and will vary and generally when using automatic overclocking software.
I just don't think people are running an automatic 4.350Ghz on all 16 threads at 1.35v. From my experience with static overclocking, the cpu would require much more voltage to run all cores and threads at it's max boost state. No different than any other AMD processor made before Ryzen.
|
|
|
|
Post by Bones on Feb 19, 2019 18:13:17 GMT -5
I agree, there is certainly the possibility of degredation over time.
I'm not sure if it would be as extreme as what Clock had happen but since it's happened with everything else we know of it's plausible at least to expect it. These chips are still too new to know everything about them and that includes the first gen of Ryzen.
Long term knowlege will only be had with time, I say let's keep this subject in mind for you Ryzen owners at the very least.
|
|
|
|
Post by zila1 on Feb 19, 2019 18:23:20 GMT -5
My chip over volts automatically to 1.55v. And this is out of the box, totally stock. I went to a negative 0.100 offset due to my concern over it and then read that AMD Bob says it was engineered to do just that. But I still have problems accepting that. Good subject matter I think.
|
|
orl
Regular Member
Posts: 96
|
Post by orl on Feb 19, 2019 19:12:40 GMT -5
Lol. I already picked number 2. Always do... reasons: A. Was taught that auto overclocking and software overclocking is weak. B. Overclock manually only. I always know where my cpu frequency and voltage is truly at. It is known that actual voltage and software readouts can and will vary and generally when using automatic overclocking software. I just don't think people are running an automatic 4.350Ghz on all 16 threads at 1.35v. From my experience with static overclocking, the cpu would require much more voltage to run all cores and threads at it's max boost state. No different than any other AMD processor made before Ryzen. The first time I booted up my system and ran HWMON, my CPUs spiked up to 5.13Ghz on the auto clocks. I attributed to a read error, but I wont lie there was this little glimmer of hope in the back of my head that I just scored the holy grail of Ryzen. Snapping back to reality, I am trying to manage stability at 4.3Ghz right now, I am stuck and slowly moving the vCore up and have now implemented LLC. Combing through my adjustable settings to stabilize as its not responding all that much to voltage. I want to find its max then bring it into line with daily temps at a relatively high vCore. I am super curious about this factor and also obviously want to know what this chip will do. Its funny, I cannot get this chip to break 60C no matter what test I throw at it with 4100Mhz and a 1.35vCore static without LLC. However I am starting to see a real and present temp climb attempting to go any further. If it doesn't respond any more I am going to flash the Bios and take a stab at it again. Then will try bringing the memory to 2133 and see if the added stability opens the topside more. .
You overclock like me Shrimp, always did. Pretty sure we used to tag team the help seekers over at the EOCF together eight to ten years ago.
EDIT: I have basic stability now at 4.3Ghz and 1.43 vCore. Still room to climb in voltage but going to do a temp test and slightly longer stab test.
|
|
|
|
Post by ShrimpBrime on Feb 19, 2019 21:59:11 GMT -5
Well mine isn't going to do 4300Mhz stable without a bit more V-core than yours, but I'm also not using fancy liquid. However I can stabilize that with SMT turned off and run it daily. In fact I've tested as high as 43.75X multiplier with SMT turned off. SMT enabled, keep "supposed safe" voltage levels, 4.1Ghz is going to be a max for me. 4.2Ghz I add a positive offset, Run LLC to extreme and oh about 1.4500v give or take a touch. I generally won't run 4.2ghz. I did take the time to tweak the memory timings at 2666Mhz and seems good. Haven't had much troubles other than my GTX 770 giving up the video driver when I wasn't looking lol. Certainly do remember the old EOCF days. I always hope that it would go to be updated and prettied up and continued on, but Jason just gave up. I sent him a simple PM asking hows it going and if I could help with the forum.... got not an answer. But then him and I aren't exactly close buds you never see him around any more. You have thermal headroom over me. My chip stops stability at any voltage under 1.6v @ exactly 4.5ghz. That's shut down to two cores, four six...ect. The extra voltage did nothing. @ 2 cores have 20c headroom. So I hit my thermal wall at 4.5ghz. In fact, I would doubt water cooling to help. These chips are like the FX chips. COLD! GET EM COLD!! And as far as degrading my lifespan of said processor.... I know and understand the risks of overclocking. If it breaks.... oh well. The one thing I see in mainstream is using FXR and it's capabilities. Now I do have curiosity about it. But my board is rather lame for such adventure. I have to be honest, I'd like to try it. Also notice that people don't mind the thermal throttling. As the processor heats up, it clocks go down. OK. So they can run a fast Cinebench run and well, that doesn't heat soak a loop lol. I pull of the same scores with a static OC on air but never seem to suffer the throttling. Pete was telling me it'll throttle. Well, this one will not throttle unless the system is run stock and I literally load the Cpu and stop the fan with my finger. Other than that, I've ran the cpu straight up to 78c, no throttle no shut down, did NOT venture for a period of time at this temp. But near 4450Mhz 8 core 0 SMT she can get pretty warm during Zila's IBT tests I've been running it through. 5 months. What speed and Cpu bench would you like to see at 4400Mhz 0SMT? it's not an issue.  |
|
|
|
Post by ShrimpBrime on Feb 19, 2019 23:39:49 GMT -5
|
|
orl
Regular Member
Posts: 96
|
Post by orl on Feb 20, 2019 9:21:28 GMT -5
Remember ShrimpBrime , my chip has 2 less cores than yours. So I will definitely have a bit more thermal headroom than you if similarly binned due to that and the loop. Before bed last night I drove her up to 4.3 Ghz with 1.45 for stability for an hour in prime, I moved my PWM fan threshold down to 65C for giggles, it stopped climbing temps at 63C or probably 90-95% fan speed. I might be able to bring the voltage down a bit, as that voltage was out of frustration just to make it stick. Currently the only setting I have disabled is PBO and the only performance setting enabled is LLC at high. Voltages appear to be right in line with my static vcore at load. I am going to soft foot around though at these voltages as my motherboard doesn't have the power component diversity of the Taichi or the CHVII etc. Supposedly the power systems are of higher quality, but fewer VRMs etc in this board, but I cannot say for sure as no unit to compare. I will bring a thermal imager home to be sure no components on the board are cooking yet. |
|
|
|
Post by ShrimpBrime on Feb 20, 2019 16:37:44 GMT -5
Nice, I hope to see more results in the thread you made I'm hoping not to see degradation from any one anytime soon, so hopefully this particular thread goes quiet lol. |
|
|
|
Post by eidairman1 on Mar 1, 2019 20:52:18 GMT -5
Ive yet to see any AMD CPU not handle 1.5V
|
|
|
|
Post by alpi on Mar 2, 2019 10:06:37 GMT -5
C'mon, how can degrade on stock voltages ? :S 2700X's boost 1,5V+ in stock settings, can't believe that any degradation happen above 1,4V. It's some kind of legend like the 1,35 V vdimm at ddr4.  Many ppl talk about it but no one have seen yet. Anyway degradation usually happens when high voltages and high temps are present at one time. Very high voltages and high temps. At least as I've seen and heard usually not the vcore responsible for degradations but some pll or voltages around imc. |
|
|
|
Post by ShrimpBrime on Apr 25, 2019 23:00:16 GMT -5
Gotta bring up that I do not suggest using anything over 1.6v without cold (sub zero) - I admit to nothing, Just don't do it.
I'd put 1.225v at 3.7Ghz a very safe number.
Run 1.35v at 3850 would be fine sauce,
1.450v at 4.1Ghz pretty dandy.
Stretching for anything past 4.2Ghz needing more voltage than 1.450 is stretching it, the gains get small fast. Not worth it.
Cas 11 @ 800 1.25v - XD
|
|
|
|
Post by Vinster on Apr 26, 2019 6:16:09 GMT -5
I've been running over 1.4v at 4Ghz since I got my setup. must be going on near a year. now just this week I've started getting the "Failed to Start due to instability" error and having to go in the bios and just re-accept my settings and she boots in to windows.
Vin
|
|
|
|
Post by eidairman1 on Apr 30, 2019 17:16:10 GMT -5
I've been running over 1.4v at 4Ghz since I got my setup. must be going on near a year. now just this week I've started getting the "Failed to Start due to instability" error and having to go in the bios and just re-accept my settings and she boots in to windows. Vin Fresh cmos battery? |
|
|
|
Post by Macsbeach98 on Apr 30, 2019 20:38:10 GMT -5
All chips will degrade over time if given enough voltage but I consider 1.5v safe with Ryzen I have run higher than that in testing. My 1800x I was running upto 1.55v without issues but once you get to its top frequency more voltage dont help much cold does. In saying that my first 1800x I killed with 1.9v testing on LN2 about three weeks after I got it and had to RMA it. They dont need a lot of voltage I was playing with my 2700x the other week cold -190c and 1.7v core 1.3v soc was about the nice spot 1.75v didnt help at all it was stable @ 5.5Ghz spent a bit of time trying to get higher but it didnt want to play the good ones are getting through Cine 15 @ 5.7 my chips are only retail you get what you get and live with it.  Another thing I have found they dont seem to like a lot of ram voltage 1.7v and they seem to act up. |
|
|
|
Post by eidairman1 on May 3, 2019 20:36:41 GMT -5
Ive just never known a AMD CPU to not handle 1.5-1.65VCore.
Hope to eventually upgrade but Im now going to School under Post 911 GI bill with BAH to get A&P license.
|
|
|
|
Post by Vinster on May 5, 2019 22:05:00 GMT -5
I've been running over 1.4v at 4Ghz since I got my setup. must be going on near a year. now just this week I've started getting the "Failed to Start due to instability" error and having to go in the bios and just re-accept my settings and she boots in to windows. Vin Fresh cmos battery? It's possible, i've been good for a bit again... I'll try that if it comes back
Vin
|
|
orl
Regular Member
Posts: 96
|
Post by orl on May 8, 2019 17:02:15 GMT -5
Ive just never known a AMD CPU to not handle 1.5-1.65VCore. Hope to eventually upgrade but Im now going to School under Post 911 GI bill with BAH to get A&P license. Its very important to note, these are not the same chips in any way or measure compared to the last generations. Never ever assume what used to be is the same my friend.
Vinster Please keep us informed, I am currently running 1.385 for 4.1Ghz and have nothing to report as far as this subject. But my measure is quite small yet, my system settled in nicely for daily driver speeds, but finding very little time to put the top down and try higher than my max achieved. |
|
Many ppl talk about it but no one have seen yet. Anyway degradation usually happens when high voltages and high temps are present at one time. Very high voltages and high temps. At least as I've seen and heard usually not the vcore responsible for degradations but some pll or voltages around imc.
