I live on the edge of rural backwoods Florida. There is definitely a market for selling old cheap shit. If I sell it for $20, I come out on top.
Its built! Its running! I love it! I had no idea what I was missing out on. I knew I was on old hardware, but I had no idea just how terrible the experience was until now. I'm the most floored by boot time. I currently have it OCed to 3.8GHz @ 1.34V and max temps I've found during stress testing is 75C. I've only stress tested for an hour, but no issues at all so far. I might start slowly going higher on the multiplier. I originally set it to 3.9GHz @1.35V, because I saw someone do it online, and it passed an hour running AIDA64, but I wasn't aware of how unreliable that test is, because I ran Geekbench to compare gains against stock speeds and my system crashed. So now I'm sticking with prime95. I have to do voltage in offsets though, which I really dislike. I knew it was a consequence of choosing a B350 board, but its more obnoxious than I thought it would be. I do have two questions: 1) There is a Vcore voltage and a VCore SOC voltage. What is the difference? I don't know what SOC means. 2) What should be the max temp I don't want to go over? My understanding is Ryzen can't get as hot as Intel CPUs, but I keep finding conflicting information. I read that after 3.9GHz, it will require bigger jumps in voltage, which will increase the heat faster.
As a rule of thumb, don't go over 1.4v unless you want to shorten the life of the CPU. Most everyone agrees that 1.4 is safe but others say keep it under 1.375. Best thing to do is start at 1.4 volts and see what your most stable multiplier is. AMD actually has a tech video recommending this. Honestly, I can't imagine an air cooler will handle 1.4v but you never know. Part of the silicon lottery isn't just top speed. Some chips can run at better speeds at much less voltage too. So maybe you'll get lucky and hit 3.9 at a lower voltage. 3.8 was my only stable overclock on air. 3.85 seemed stable but the system would shut down when it got too hot. Also, temperature readings seem to vary a lot depending on the software you use to read it. I still don't know whats the most accurate software. You may also notice that the CPU and motherboard will report different CPU temps. Best thing to do is go with the CPU reading. I think 90 is okay for the CPU but my motherboard always shutdown around 83. Something was off there. The voltage and heat curve after 3.7 is extreme. I believe Vcore SOC voltage is specific to Ryzen. It's the voltage for the other "systems on the chip", like the memory controller. I'm not sure of the benefits of raising it.
Yeah I had no intention of going over 1.4v because it wouldn't be worth it for a gain of 100MHz. I'm currently running prime95 and I'm going on 15 minutes at 3.89GHz and 1.39v. Hasn't crashed yet, which is good as I've been stepping up the voltage in +0.05 increments and this is the first time it didn't crash almost instantly. Max temp so far is 82C. Crossing my fingers. Even if it's stable, I may step it back down to 3.8GHz because then I can lower the heat and voltage significantly. I just want to know what my limit is. I'm already having better luck than I ever had on my Phenom II. Mine was never stable over 3.4GHz, which was well below what most people could achieve with that same CPU. I'm much closer to the average gains this.
NVM. It crashed just shy of 20 minutes into the test. I'm not going any further. I'm dropping back down to 3.8GHz and 1.34v, run a much longer stress test tomorrow to be sure it's stable, then I'll maybe slowly back off the voltage little by little if I'm feeling ambitious.
It might have been temps that shut you down. Mine always shut down around the same temp. I suspect there are temp spikes that happen too fast to register on the screen but will still hit the protection limit in the CPU. 3.8GHz is really good for air cooling. The 1800X can hit 3.9GHz on air but it uses less voltage to do it.
Dell has a 1080ti for $634. If I could get $500 for my 1070 I'd be tempted. http://www.dell.com/en-us/work/shop...acd=12309198375458460&VEN3=111603834286848715
So my old tower will not boot. I push the power switch and nothing happens at all. I'm assuming its either the power supply or the switch has gone bad. I don't need it, but my curiosity has me wondering what the issue is. Is there a way to diagnose what is causing the problem without having to swap out parts by trial and error?
Two most likely culprits: bad power supply or bad RAM Solution unfortunately will involve trial and error to figure out the culprit.
If the CPU fan wont spin it's a PSU, motherboard, or power switch. If you can find where the power switch connects on the board you can pull the cable and short the prongs with a piece of metal or jumper.
To test a PSU just unplug it from the mains, use something that conducts like a paper clip to short between the switch wire (green) and a ground line (black) and plug it back into the wall, if it works it will power on and the fan will spin. I use this technique all the time to power things up outside of PC cases or if I need to use 2 PSUs with a single system.
I have a jumper that does this and for some reason it didn't work on my last power supply which was about 11 years old. The PSU did work so I'm not sure how it was different.
In simple terms any switch works by connecting it to ground so as long as the switch line and ground were being found it should have worked. It could be that the ground line you were using on the ATX connector was broken, there are plenty to choose from on an ATX PSU so try shorting switch (green wire) to any other.
Laptop battery needs changing It's not able to fully charge anymore and lasts 3 hours less than it used to. Never bought a replacement before so no idea what is good or not. How can you tell what is a good or dodgy one? My laat one could hold around four hours charge so I'd like the same if possible.