FX 8320 power usage sensor details?
FX 8320 power usage sensor details?
hello guys recently i overclocked my FX 8320 to 4.7 ghz and had to turn off every svm, power saving and other settings that limit manual control. After this, the hw info, aida hwmonitor and similar apps no longer show cpu wattage consumption. So my main question is what i did to stop the cpu from displaying its power usage in real time?
Check out this link: https://www.techspot.com/downloads/4645-...drive.html
thanks but could u tell me how is this supposed to help me i mean i used the app before but it didn't do much
Reading values in this format are typical for the old APP developed by AMD specifically for the FX series of their CPUs, ensuring high accuracy.
What motherboard are you running?
On the Gigabyte GA990FXA board, the only adjustment I make is turning on HPC to keep the CPU (FX6300) at its configured overclock (4.6Ghz) during demanding tasks like Cinebench or Prime95, or even just gaming. When HPC is off, an overclocked FX CPU typically reduces clocks and voltage under heavy load to stay within its average TDP limits. This behavior makes sense if your motherboard’s VRM can’t supply enough power without overheating, so disabling it depends on the VRM’s capabilities and cooling performance.
I keep APM and Cool N Quiet enabled. With them active, the CPU lowers clocks and voltage when not under stress, which is what I prefer. If you need to see high clocks even when idle, turn them off. However, maintaining consistently high voltage for overclocking stability can accelerate CPU wear, so consider this if preserving these components is important.
SVM handles CPU virtualization. If it’s enabled, HWInfo may not be able to read the CPU telemetry required to estimate power use. That’s likely your issue. Unless you need virtual machine support for your work, you should leave it disabled.
I never used AMD’s Overdrive tool and assumed it was outdated or unsupported. I doubt it will function in Windows 10 or 11, especially considering their strict requirements for low-level hardware access. I usually perform overclocking directly through the BIOS.
For FX cpus, AMD Overdrive or CoreTemp remains the most reliable method for measuring temperatures, an essential aspect of any overclocking process. The case temperature for all FX models is around 63°C, yet without thermal strips on the cores, it's impossible to obtain a precise reading. Overdrive and CoreTemp rely on intricate calculations involving voltages, loads, socket temperatures, and other factors to determine a Thermal Margin. What matters about TM readings isn't the exact figure, but rather the thermal headroom relative to the case temperature—expressed as a value of zero. A TM rating of 40 indicates substantial headroom, while lower numbers suggest limited room for error. Single-digit ratings mean you're running hot, and zero or negative values signal either no thermal margin or temperatures exceeding the case limit (such as an Intel at 100°C when it should be 105°C). Most claims about specific temperatures on FX CPUs are inaccurate; software cannot truly capture the actual CPU temperature due to these limitations. Socket temperature does not reflect CPU or core temperature, which is why many tools give misleading results—like showing 8°C idle or 18°C during gaming.
OD and CoreTemp share the same thermal margin code, developed jointly by the creators. It remains the most reliable method, though it doesn’t guarantee consistent interpretation of the data.
There’s no temperature telemetry available from an FX device. Any figures provided aren’t actual temperatures; bios simply assume those addresses should correspond to a thermal strip inside the CPU. This isn’t the case.
With Intel, software displays core temperatures in °C and requires some mental adjustment to understand what they mean. It took several generations to train users away from the misconception that 70°C is dangerous, and now many accept that their 13900k running at 90°C is acceptable.
Thermal margin readings also differ—they aren’t shown as a fixed °C value but as a measure of available thermal headroom. Interpreting this requires some context; a reading in the 40s is acceptable, 20s is fine, while values below 1-9 indicate overheating concerns. It’s similar to assessing an Intel processor at 32°C, 60°C, or 91-99°C.
I personally use HWInfo for my Ryzen, and it provides useful insights for FX users—it’s practical and relatable. However, it lacks thermal margin data, which is crucial for running an FX 8320 at 4.7GHz, especially since stable booting may require bypassing certain instructions to avoid Tcase issues.
CoreTemp fulfills its basic purpose without extra features; it simply reports temperature and isn’t much more than that. Its main advantage over other tools is the thermal margin feature, which must be activated in the settings menu. Without it, it only shows °C values. It’s not as detailed as HWInfo, but for many users, simplicity is preferable to unnecessary complexity.
One advantage of HWInfo is its background monitoring capability—it keeps temperature visible in the Start Bar alongside other apps, ensuring temperatures remain visible without manual checks. In contrast, HWInfo only displays readings when you actively open the app, which can be inconvenient.
Both software and hardware face constraints. The FX offers limited thermal precision per core, while software is restricted by these limitations. This makes OD and CT the most practical options, especially since we’re dealing with an AMD FX processor where Intel’s temperature standards don’t apply.
But without exploring the meaning of what you think 'telemetry' means, the BIOS and OS tools must be pulling data from the CPU to determine what's seen as "temperature". It's not correct, it hasn't been claimed that way, and especially when CPU usage is low, the reported numbers can be completely misleading. At full load it might still be useful for controlling fans, but it remains inaccurate.
All the thermal margin relies on mathematical tricks using the same flawed information the CPU provides. It's useful, though, and the temperature reading from HWINfo is also helpful.
But then... what does exceeding "Tcase" actually mean? Is there a specific standard for that? Is it something AMD has set up? AMD has defined a Tjmax and shares it publicly.