So I think this was because I had enabled an auto-overclocking feature in the BIOS, but other people have reported high temps and voltages even with untouched BIOS settings. All the time I thought I was setting my voltage through Ryzen master while OCing, with quite low clockspeeds and voltages (voltages according to Ryzen master).
I was getting really high temps under load (Prime95 small FFT), like 85-95+C high and the sytem kept crashing. I thought the temps were inaccurate cause I had heard of that kind of thing with Ryzen. But then once I noticed from CPU-Z and the Asus utility that the voltages were hitting around 1.55V at load I realized that my problem wasn't instablity caused by an overclock with too high clockspeed and too low voltage, but rather my CPU was actually hitting TJ max and shutting the coputer down. And thinking there was a problem with my voltages set in the Ryzen master I adjusted voltage in the Ryzen master and tried again with the Small FFT, everytime the system would crash, and I thought it was an unstable OC.
The thing is that I didn't realize the motheboard would do this, especially because the CPU didn't even be set above 3000-3200Mhz without me manually setting it to that in Ryzen master. The CPU seeems to work well but I fear that the chip has been damaged and it's lifetime decreased by having it reach TJ max multiple times, and it running at voltages hitting almost 1.55V for some time after it inevitably always crashed the system.
I don't really know what to do since this isn't a problem with the chip and that's why I don't expect or even want to ask for AMD to help me. After that I have 4 options, which are:
1. Keep the CPU and hope it's lifespan hasn't been noticeably decreased because of this
2. Ask Asus for help with this and the possibility if they could exchange my chip because their motherboard has been overvolting my chip and making it overheat, and this has been with a respected aftermarket heatsink
3. Return the CPU to the retailer, which I could do under the consumer protection law return time of one month. The thing just is that I'd feel bad to know that my chip might just end up in the hands of some other consumer who might want to overclock their chip, and might end up with a chip that doesn't last as long as it should. My retailers website did say that the chip has a 1year warranty, but I think that's only the timeframe the retailer will feal with warranty requests and after that it would go according to this AMD warranty: AMD Processor in a box (PIB) 3 Year Limited Warranty.
If the CPU having the 3 year warranty is the case then maybe I could just keep the CPU because if it decides to stop working before the 3 year warranty period I'd be covered by AMD, and if it stops working after the 3 years I'd then be prepared (and actually already might have) to buy a new CPU.
What do you guys think I should do, should I even be so worried about the CPU as I am?
What BIOS version are you running on the motherboard? I have the same motherboard and I flashed the BIOS to version 0503 before installing the operating system. So far so good, Windows 10 Pro appears stable, but then again I'm NOT trying to overclock. I would make sure you are running the latest BIOS version then default BIOS settings to see if your temps are any better.
According to Ryzen Master the temperatures when Windows 10 power options are set to high performance reports around 31-34C, not bad for a stock air cooler.
CPU-Z does not show me correct voltage, but I'm on an ASRock AB350 Pro4. Similar test as yours, I see 0.648V in CPU-Z for core V under load and @86C running Prim95 16 threads on a 1700 @3.8 should be 1.35V. I wish there were tools that reported values correctly, but this is relatively new. I would assume that voaltge and temp reporting is kind of important; especially touting overclocking as thing any of their processors can do. I have heard of issues with boards that driving voltages way to high, even for Intel, when in "auto" overclock modes. This will probably need a BIOS fix, and it's probably not an AMD issue. I have experienced auto-shut down, but I'm not sure if that was over-heating or just too high a clock to be stable; experienced that at 3.9 but mostly above that and I manually to the clocks. Good luck finding your problem. Post back if you have success.
The high voltage problem was fixed by just turning off the "OC tuner" in the EUFI. Also I did notice that the Vcore in CPU-Z is wrong, but I'm certain now that the Vcore in the Asus utility is correct, and that was also alarmingly high when the OC tuner was on
I did still want to try to manually OC. I've been trying to get 3.9Ghz because I haven't even heard of anyone getting under 3.9Ghz. Unfortunately I can't even get 3.9 to be stable with Prime95 small FFT test. I've tried a whole bunch of voltages, I've had it up to 1.46V at max but still the system keeps crashing a few minutes into a test.
I suspect the problem I have is the VRMs of the board are overheating and that's causing problems. I think this is the case because of how it always seems to take the same amount of time for the system to completely crash when running P95 small FFT. I also tried seeing, or rather touching the VRM heatsink to get an idea of how hot they were getting and they were seriously burning hot to the touch.
This just makes me want to return the board even more, even though I was gonna return it anyways even if it could OC as much as I think my chip could take.
Are you running the latest BIOS firmware? 0503 I believe is the latest for the Asus Prime B350-Plus. So far I haven't had any issues with my board. The board came with the 0406 bios, but I flashed it to 0503 and haven't experienced any overheating issues. Reported temperatures from Ryzen Master appear to be normal. Here are my Ryzen7 1700 temps with the Windows 10 power option set to 'power saver' mode, all cores at 1.5Ghz and running at 31.50C
The core boost/autooverclock or what ever is called, never worked properly in amd cpus, they put too much voltages regardless of what really is needed.
I always disable them wether overclocking or not.
Temperatures are wrong only with 1700X and 1800X, leading to unnecessary high fan speeds and a louder system even with low CPU load. Your 1700 will be fine.
I'm seeing pretty much the same thing to a point. I have a 1700x at 3.7 B350-plus motherboard 406 bios idle temps are 48/52-58c and load temperatures are 61-73c and I'm totally ok with this ill stay here lulz.. I've checked this looking at bios temperatures, ryzen master temps, and hwinfo of which all temps are matching within reason. Octuner on 1.55v yes it shows 1.43 with fluctuation in bios matching hwinfo average.
The issue im having is I have tried both the 503 and the recent 513 with a 10-20c increase. Both bios versions I ran at 3.4 with octuner off.
Both 503 and 513 idle at 61c with no oc. My temps range from 73-83c under load and im seeing peaks of 91c.
Also 58c idle using 406 bios my fan speed is 850-950rpm. With 503/513 bios my fan rpm is 1300 @ 61c with a constant fan speed fluctuation/noise. (Not what I want to hear at idle lol).
The ryzen chart temperature of 58c idle is doable with the 406 bios in my case. Hard to see a performance increase when I have to under clock to keep it stable and still cant 2933 on the ram bro...
Yet I think trolling all youtube ryzen videos and other board manufacture forums will show these numbers are just all over the place from core clock readings, voltages, and temperatures along with fps. Overall I am happy with my purchase/performance I just really hope this gets dialed in and not a system parkour run.
You're temps are very high, even for air, on idle. I was running the Wraith Spire air cooler that came with my 1700, and at idle running 3.8 it was in the 30s. Chips do vary, but yours are out of the range of normal. Either you board is over-volting or your cooler is not working appropriately.
1) Have you lifted the cooler to see if there is either too little compound and/or proper contact is being made? You should clean both parts and use new compound if you do this, but it's not mandatory to reseat it. When I lifted my Spire, I had very little compound, but it was evenly spread. Probably could have gotten better temps under load with it, but I'm on water now so caesura. The saying that too much or too little is bad should really read too little is bad and to much is messy and bad if conductive. It's a pain to remove, but if you're sure that over-volting is no the issue, and it is a major issue with Ryzen and ASUS boards in general (Z270 also), then you need to look at installation.
2) Check installation of the cooler; you did not mention air or water. One of the issues has been brackets for AM4 AIO water, and possibly 3rd party air as well, coolers and specification from AMD not being correct or complete for socket/chip height off the board. This has caused delays in delivery of some product brackets, but some other manufacturers could have just guessed.