- Joined
- May 30, 2018
- Messages
- 1,890 (0.88/day)
- Location
- Cusp Of Mania, FL
| Processor | Ryzen 9 3900X |
|---|---|
| Motherboard | Asus ROG Strix X370-F |
| Cooling | Dark Rock 4, 3x Corsair ML140 front intake, 1x rear exhaust |
| Memory | 2x8GB TridentZ RGB [3600Mhz CL16] |
| Video Card(s) | EVGA 3060ti FTW3 Ultra Gaming |
| Storage | 970 EVO 500GB nvme, 860 EVO 250GB SATA, Seagate Barracuda 1TB + 4TB HDDs |
| Display(s) | 27" MSI G27C4 FHD 165hz |
| Case | NZXT H710 |
| Audio Device(s) | Modi Multibit, Vali 2, Shortest Way 51+ - LSR 305's, Focal Clear, HD6xx, HE5xx, LCD-2 Classic |
| Power Supply | Corsair RM650x v2 |
| Mouse | iunno whatever cheap crap logitech *clutches Xbox 360 controller security blanket* |
| Keyboard | HyperX Alloy Pro |
| Software | Windows 10 Pro |
| Benchmark Scores | ask your mother |
Learning more about how Ryzen 2 behaves when pushed all of the time. I've managed to push 4.25 stable @ a whopping 1.375v (or to me that is, when 4.2 is actually perfectly stable at 1.3.) Went for 4.3 initially but quickly gave up on that. Mostly I've been sticking to preliminary stuff, with lots of space between. I run IBT standard and if it passes at at least somewhat reasonable temps, I make a note. Later, when my cooling is better, I can try more rigorous testing. At this point, I'm not keen on running even a couple of degrees above 80C for extended periods of time, or even doing it in repeated bursts. If those are the temps I'm stuck with, I won't be running those configurations past just seeing what is possible and grabbing some quick benches, so prolonged stability testing isn't necessary and probably does more harm than good.
A few observations...
1. Stability invariably tanks 10C before throttling territory. An overclock that peaks at 85C, no matter if it has enough voltage or not, will likely never be stable. Even one that only averages at 80 probably won't. I'm fairly sure at this point. A few times I passed 80C and almost slid by before spitting up an error and ending the test. This seems to be the hard line for this chip. Within 30 seconds of first passing 80C by even a degree or two, it falls apart. So with that in mind I'll be avoiding it entirely. Would probably advise others to do the same.
2. Droop actually matters less than temperatures. To hit my initial 4.225ghz I had LLC maxed (level 5) with almost perfectly flat voltage swing at the voltage needed for that. Just a tiny overshoot, I disregarded it until I went up up in either voltage or clock speed and the overshoot became exacerbated. Going up to the next bump of 4.25 with that sort of voltage curve and LLC took me past the above temp limit and it promptly crashed after hitting 85C a couple of times - two was enough for me to deduce that there's something to it and stop there. Even though voltage actually got significantly higher, it wasn't even close to passing. I mean, at peak I caught overshoots to 1.4! And you're telling me that's still not enough for 25mhz more? I don't fully buy that. I guess spikes like that in themselves could be to blame, but I don't know.
Now tonight, at level 4, with a switching frequency of 450 and that same 1.375 voltage set, I am drooping to 1.36ish instead of shooting to 1.4, and yet I am now passing IBT at 4.25 with a max temp of 80C. Food for thought - voltage may not be the real answer at the upper range. Not on air, anyway.
And I'm thinking it's also directly tied to clock speed. The higher up you get, the lower the max stable temp is probably gonna be. I'm starting to feel very confident that the reason I can't hit 4.3 has nothing to do with not pumping enough voltage. 1.38 or so may already be enough. But because power consumption increases with faster clocks, the temperature rise would cause me to fail any test, whether there's enough juice or not.
3. Unfortunately this all suggests to me that I will never quite hit 4.3 with this cooling setup. Here's hoping fixing my case flow situation does the trick. I should have my new fans tomorrow. Looking at replacing all of the other fans, too. A 120mm Fractal Venturi non-pwm for the rear, a 140mm non-pwm for the top, and a 140mm pwm for the cooler. Whether or not this is enough... ...well I have my doubts. It all comes down to how much only having one intake fan is holding me back. I'm not sure how much further I wanna take this just for entertainment. I'm at the point where I'm seriously thinking about swapping CPU coolers again... ...and that's not good! I swear it isn't!
If I'm not talkin' nonsense here, a throttle point of 95 doesn't make a whole lot of sense to me. That's a whole range of nearly 15C that this CPU has no business of operating in. Absolutely no benefit to ever even going there. I mean, nothing would be lost if it throttled at above 80C instead of 95C, you know? To me it just suggests that it might actually be kind of risky to ever even approach 95C. Or at least, knowing what I now know, I'll never allow it. Instability doesn't always mean degradation, but there's always instability from temperature degradation, so... yeah. Live and learn from my foolishness.
A few observations...
1. Stability invariably tanks 10C before throttling territory. An overclock that peaks at 85C, no matter if it has enough voltage or not, will likely never be stable. Even one that only averages at 80 probably won't. I'm fairly sure at this point. A few times I passed 80C and almost slid by before spitting up an error and ending the test. This seems to be the hard line for this chip. Within 30 seconds of first passing 80C by even a degree or two, it falls apart. So with that in mind I'll be avoiding it entirely. Would probably advise others to do the same.
2. Droop actually matters less than temperatures. To hit my initial 4.225ghz I had LLC maxed (level 5) with almost perfectly flat voltage swing at the voltage needed for that. Just a tiny overshoot, I disregarded it until I went up up in either voltage or clock speed and the overshoot became exacerbated. Going up to the next bump of 4.25 with that sort of voltage curve and LLC took me past the above temp limit and it promptly crashed after hitting 85C a couple of times - two was enough for me to deduce that there's something to it and stop there. Even though voltage actually got significantly higher, it wasn't even close to passing. I mean, at peak I caught overshoots to 1.4! And you're telling me that's still not enough for 25mhz more? I don't fully buy that. I guess spikes like that in themselves could be to blame, but I don't know.
Now tonight, at level 4, with a switching frequency of 450 and that same 1.375 voltage set, I am drooping to 1.36ish instead of shooting to 1.4, and yet I am now passing IBT at 4.25 with a max temp of 80C. Food for thought - voltage may not be the real answer at the upper range. Not on air, anyway.
And I'm thinking it's also directly tied to clock speed. The higher up you get, the lower the max stable temp is probably gonna be. I'm starting to feel very confident that the reason I can't hit 4.3 has nothing to do with not pumping enough voltage. 1.38 or so may already be enough. But because power consumption increases with faster clocks, the temperature rise would cause me to fail any test, whether there's enough juice or not.
3. Unfortunately this all suggests to me that I will never quite hit 4.3 with this cooling setup. Here's hoping fixing my case flow situation does the trick. I should have my new fans tomorrow. Looking at replacing all of the other fans, too. A 120mm Fractal Venturi non-pwm for the rear, a 140mm non-pwm for the top, and a 140mm pwm for the cooler. Whether or not this is enough... ...well I have my doubts. It all comes down to how much only having one intake fan is holding me back. I'm not sure how much further I wanna take this just for entertainment. I'm at the point where I'm seriously thinking about swapping CPU coolers again... ...and that's not good! I swear it isn't!
If I'm not talkin' nonsense here, a throttle point of 95 doesn't make a whole lot of sense to me. That's a whole range of nearly 15C that this CPU has no business of operating in. Absolutely no benefit to ever even going there. I mean, nothing would be lost if it throttled at above 80C instead of 95C, you know? To me it just suggests that it might actually be kind of risky to ever even approach 95C. Or at least, knowing what I now know, I'll never allow it. Instability doesn't always mean degradation, but there's always instability from temperature degradation, so... yeah. Live and learn from my foolishness.
Last edited:
Other than that all peaches, honey n chocolates. 
