This setting doesn't increase to 4.5.
This setting doesn't increase to 4.5.
hello, i recently purchased a 5800x3d and haven’t spent much time adjusting the settings until now. i ran the chip through cinebench “stock” and it didn’t exceed 4.1 with peak temperatures around 72°C. i tried undervolting the cpu using an offset of -25, which improved boost but capped at about 4.25gHz. my temps stayed at 56°C and power consumption was roughly 60-65w—something feels off. i enabled rezisable bar, PBO, and DOCP, but didn’t tweak any other settings except the negative offset. it seems like either the motherboard is limiting performance or i got a subpar chip. specs: asus tuf plus b550 (latest bios & drivers), 5800x3d 32gb, 3200ram, arctic freezer 2 (240) rm750x psu, rx 6800xt.
Use whatever OC software your board supports, possibly an older version if the bug is fixed. Adjust multiplier and voltage manually, keeping the voltage near 1.35v or lower. For non-x3d models running at 4.7-4.8ghz allcore, around 1.32 to 1.34v should work without overheating.
Overclocking on 5800x3D is disabled, multiplier and voltage can't be adjusted. There seems to be an issue. My CPU reaches 4.45GHz with the curve optimizer at -20 offset and consumes roughly 90-120W. I recommend resetting BIOS to defaults and verifying your power plan in Windows, as well as updating chipset drivers. EDIT: It's likely the newest BIOS uses AGESA version 2.0, fixing the EDC bug; it now defaults to 120 instead of 140, which might explain a power limit if EDC is indeed 120. I'm still using the older BIOS since no compelling reason to upgrade.
You can't. You can use Curve Optimizer to move the frequency / voltage curve but it has limit at -30 and there is no reason to use positive offset as the CPU by default can already hit over 90C even with a 360mm AIO at 1.25V. So going to 1.35V even if it was possible would be absolutely unrealistic. Yes, it was possible to adjust voltage due to a bug. But there is a reason why it's locked. However there is no reason that a multiplier should be locked and even with the bug that gives you control over voltage you still can't adjust multiplier. The only way to clock it higher is with BCLK which has another set of very different limits and issues.
I've never actually used bclk myself, probably because most people sharing this info haven't tried it either—maybe out of fear or just not wanting to put in effort. I've managed to do some bclk on 1155 and mostly stayed away from the CPU unless necessary. The main thing is to clock it right and make sure RAM and CPU usage aren't too high. My best CPU was a 2100, which handled 107.5 but sometimes just shut down unexpectedly. I'm not sure why it would fail and I don’t want to jump to conclusions about bclk because so many people make unsubstantiated claims. That kind of misinformation can confuse the whole overclocking community. Either way, it’s better to test it yourself if you're unsure, since others seem more eager to assume than to try.
Take a step back from your high expectations... Upgrading BIOS overclocking is straightforward on older hardware compared to newer setups where PCIe 4 becomes a challenge. Using NVMe adds another layer of risk, though data loss remains unlikely but possible. You shouldn't push these limits excessively since PCIe 4 usually hits its boundaries before meaningful improvements appear. It's not useful for everyday use versus just aiming for benchmark records. On Intel platforms, even with Haswell, you could isolate BIOS overclocking to just the DRAM and CPU, which isn't available on Ryzen systems.
So? Just drop it down to gen3 or gen2, at least that’s what I notice when the Ryzen bclkers above 104. If that’s all I have to worry about, just lower the PCIe link speed instead. Still no solid proof of SATA HDD/SSD or NVMe issues caused by bclk online, so don’t take it seriously. At least extreme RAM overclocking might cause problems in some unstable cases—especially at very high frequencies (like 5000+) and if you’re unlucky. It probably works like with RAM where you hit an unstable bclk and things go wrong. If bclk really can mess with the OS, it’ll feel similar to RAM instability. Every time I try to boost PCIe freq (pre 1155) or bclk (1155+), I raise it for better performance, though I haven’t had problems with 120 PCIe at X58 yet. I still need a working board with 1155/1150 to test bclk stability, and the Pentium struggles with high bclk speeds. The 2100 model seems unreliable or maybe I’ve lost track of it. Still, I stay bold—overclocking gives performance gains with minimal real downsides. I don’t worry about strict voltage limits; I’m okay running DDR4 at 2V or DDR3 at 2.5V if it helps my overclocks. I rarely see anyone test voltages in a risky way like this, and most people avoid it because they fear damaging the CPU. If you want real examples, check out @freeagent and @RONOTHAN— they don’t shy away from high voltages for testing. As long as I’m not playing games and just enjoy the benefits, I’m fine.