Overclocking Ryzen 7 1700x Question
Overclocking Ryzen 7 1700x Question
HWinfo provides the best data, but the key is to examine the average voltage during normal but extended usage. Higher voltages tend to be brief and don't significantly affect temperatures or cause lasting harm.
That's the method I've been relying on, but I only fully relied on it after performing the DVM measurements at the socket for my board.
However, I needed that confirmation first.
During a stress test where the CPU is heavily loaded, the SVI2 TFN on my 3.95Gig 1700 shows a very reasonable reading.
A VCore reading of 1.472 with a solid line is normal, but an SVI2 reading between 1.36 and 1.38 indicates it's within a safe range for the CPU.
But without any load on the CPU, the SVI2 rises to 1.419–1.425… while the VCore remains in the 300–500mV range!
I had to inspect it at the base of the socket using a DVM, where it consistently reads between 300 and 500mV during idle.
So even if the CPU claims it's getting 1.425, it's actually only receiving a small portion of that at idle—just the VRM controller's output.
The SVI2 TFN isn't a direct voltage measurement; it's an estimated value the CPU assumes.
It appears to be highly inaccurate when the processor is idle, which is what I observed.
I'm not entirely sure why this happens, but that seems to be the pattern I see.
Are you certain you're viewing Vcore, not VID or another file?
I'm confident it's VCore—it appears as a reading in the HWInfo part for the Nuvoton monitoring chip. It matches DVM readings I obtain at the VRM output (the inductor leads are clearly visible) and consistently across all operating conditions: idle, full load, and varying loads. Comparing fluctuating loads is challenging because of the inherent delay in the HWInfo data.
The VDC is shown in HWInfo under 'CPU [#0]: AMD Ryzen....' and appears separately for each core.
I think the main constraints come from the silicon quality in your 1700x CPU. Plus, your ability to handle higher voltages—perhaps between 1.40 and 1.425 during testing—might be a factor.
I managed to reach 4.0ghz at 1.398v, but it occasionally jumps to 1.404. On my H100i the temperatures stay below 60°C, which should be fine for regular use.