Request for Factory BIOS Version for Acer Aspire A715-42G (GPU Crashes After BIOS Update)

Hi Acer Community,

I have an Acer Aspire A715-42G laptop (SNID: XXXXXXXX, Serial Number: NHQE5SP002XXXXXXXXXXXX) with an AMD Ryzen 5 5500U CPU and NVIDIA GeForce RTX 3050 GPU.

After a BIOS update was automatically installed through Windows Update, my laptop started experiencing GPU-related issues. The system now freezes or crashes during 3D rendering tasks, such as gaming or video exporting. It works fine during light tasks like browsing or watching videos.

Additional details:

When the system crashes, it freezes the screen completely.
On rare occasions, it shows a blue screen with the error: DPC_WATCHDOG_VIOLATION.
- I looked this up and it’s often caused by driver incompatibility, but I’ve already tried every available NVIDIA driver version from the official website (clean installs included), and the issue remains.
I also tested the integrated GPU (iGPU) by disabling the RTX 3050 and running games, and in this case, the system does not crash.

Here’s what I’ve ruled out:

No thermal throttling – both CPU and GPU temps are within normal limits.
RAM is healthy – no memory errors detected.
Power adapter is functioning properly.

Since these problems only started after the BIOS update, I strongly believe the issue is BIOS-related. I’d like to revert to the original factory BIOS version that the laptop came with, but I can’t find that version available on Acer’s support page.

Could anyone from Acer or the community help with the following?

What was the original BIOS version installed from the factory on this unit?
Is there a safe download link or method to get and install that version?
Is BIOS downgrading safe and supported for this model?

Any help would be greatly appreciated. I’m really hoping rolling back the BIOS will resolve these GPU issues.

[Edited the content to hide sensitive information]

Accepted answers

All comments

StevenGen

For all the bios and EC issues that these chips are preprogramed from Acer you need to take your laptop to a technician that specializes in programming these chips and has access to these programs also. To change and/or reprogram these bios or EC chips you need a special programming tool and also the software to to do all this, this is a specialist job.

If you look in the Acer Aspire A715-42G support page at its bios versions, this laptops first Acer listed bios is version 1.02 and its last is 1.10, The Insyde bios software that Acer uses for this laptops bios, does not allow a bios version to be change to an earlier version, as you need to do all the above to change the bios to an earlier version. Good luck.

If this answers your question and solved your query please "Click on Yes" or "Click on Like" if you find my answer useful👍

Puraw

Hi @JamesC1223 ,

Thanks for the detailed breakdown—it really helps pinpoint the root cause. Based on similar cases, this feels like a driver-level conflict more than a BIOS issue.

🧼 Here’s what I suggest:

Use DDU (Display Driver Uninstaller) in Safe Mode to fully remove all NVIDIA drivers, not just the RTX 3050.
After cleanup, reboot and let Windows 11 auto-install the best-suited driver version.

💡 Why it matters: The BIOS update may have adjusted how the dGPU is initialized or prioritized, but the DPC_WATCHDOG_VIOLATION error almost always points back to problematic drivers or driver timing issues.

You already verified thermals and RAM, so this next step could save you from needing a risky BIOS downgrade.

Let us know how it goes—might just be the reset your GPU needs. 👍

JamesC1223

https://community.acer.com/en/discussion/comment/1349505#Comment_1349505

Thank you for replying to my post. Unfortunately, I have already tried the DDU method, but the issue still persists.

Puraw

https://community.acer.com/en/discussion/comment/1349917#Comment_1349917

Hi JamesC1223, Just to confirm — did you run DDU in Windows Safe Mode using the Clean and Restart option, and allow Windows to auto-install the base driver afterward? Using NVIDIA’s Graphics Console may reintroduce driver timing issues if remnants weren’t fully purged.

JamesC1223

Thanks for the detailed explanation. I wanted to share some additional diagnostics I’ve done that might help narrow things down.

I noticed something strange during testing: my laptop consistently crashes when the GPU temperature reaches around 40–45°C, which is unusually low considering the normal thermal limit for the RTX 3050 is around 78–80°C. I used HWiNFO to monitor the temperatures, and in four separate tests, the system crashed as soon as the GPU hit that 40–50°C range whether the laptop was plugged in or running on battery.

At first, I thought it might be related to thermal throttling or a temperature limit setting in the BIOS, but unfortunately, the BIOS has very limited options and doesn’t allow me to change thermal settings.

I also checked the GPU wattage consumption. When the laptop is unplugged, the GPU only pulls 30–40W, but when plugged in, it jumps to 70–105W. I’m not sure if that’s normal, but it seemed worth noting.

Additionally, I looked through Windows Event Viewer and consistently found a Kernel-Power 41 (Critical) error logged after every crash. According to several sources online, this error could be tied to a faulty power adapter. However, I’ve already tested the system with a different, working adapter, and the problem still occurs. The laptop doesn’t just freeze it completely cuts off power and reboots.

All of this leads me to suspect there might be damage to the GPU chip itself, possibly triggered or exposed by the BIOS update. I understand that reprogramming EC/BIOS chips requires special tools and expertise, but since these issues only started right after the BIOS was updated via Windows Update, I can’t help but feel the update might have caused low-level conflicts with the GPU’s power or temperature management.

I appreciate any insight this has been a frustrating issue, and I’d really like to resolve it without replacing the hardware if possible.

JamesC1223

Just a quick update with some new diagnostics and fixes I’ve tried:

When the laptop crashes, it sometimes gives a BSOD. The two most common errors I’ve encountered are:

DPC_WATCHDOG_VIOLATION
VIDEO_TDR_FAILURE (nvlddmkm.sys)

I analyzed the crash dumps using WinDbg, and they pointed to nvlddmkm.sys, which is the NVIDIA driver. I updated to the latest NVIDIA driver, which fixed the DPC_WATCHDOG_VIOLATION, but the TDR error still occurred.

After researching, I found that the TdrLevel in the registry was set to 0 (disabled). I changed it to 1 and added the following keys:

TdrDelay = 20
TdrDdiDelay = 20

This completely fixed the TDR-related BSOD. So both BSODs are now resolved.

However, the laptop still crashes and restarts suddenly without a BSOD, and Event Viewer always logs a Kernel-Power 41 (Critical) error. I ruled out thermal and power issues after doing a full log capture using HWiNFO. Here’s a summary of what I found:

1. Driver(SSD):

Temperature: Stable at 40–41°C.
Remaining Life: 98%.
Failure/Warning Flags: No drive is healthy.

2. GPU:

Temp: Peaks at 58.5°C, Hot Spot: 66.5°C both safe.
GPU Clock: Running at ~2100 MHz both plugged and unplugged, which seems abnormal as it should throttle on battery.
Performance Limit - Power: Yes, triggered during load this is expected.
Thermal Limit: No never hit.
GPU Load: Reaches 97% during crash test, with normal power draw up to ~60W.
Memory Usage: GPU VRAM usage only around 16.9% well within safe limits.

3. Memory (System & GPU):

Virtual Memory Load: Max 72%
Physical Memory Load: Max 62%
No signs of memory pressure or leakage.

4. Battery:

Initially discharging (-9W), then switched to charging (+25W).
Battery % increased from 75.2% to 76.4% normal behavior during power state switch.

5. Other Observations:

At the crash window (around 9:23:27), GPU load and clock speed suddenly spike (as a game or app starts).
Framerate drops briefly then recovers consistent with normal GPU ramp-up.
No abnormal drive or memory activity logged during the crash.

Conclusion:

System is operating within normal thermal, power, and memory limits.
GPU is behaving normally under load but does not downclock on battery, which might be a BIOS-related power management issue.
Both BSODs are resolved, but the random crash/restart persists.
Given that the problem started right after the BIOS update, I still believe the root cause is BIOS level possibly affecting GPU clock scaling or power control.

If anyone can confirm whether the GPU clock remaining at full speed while unplugged is normal or not, that would really help. Still hoping for a way to safely revert or reflash the original BIOS.

JamesC1223

I’ve done a deeper dive into the latest logs and found several critical performance bottlenecks especially related to power delivery and GPU throttling.

Critical Anomalies Identified:

1. GPU Power Throttling

Performance Limit - Power: Always "Yes", even at low GPU utilization (0–9%).
GPU Power Draw (ASIC Power) rarely exceeds 11W (peak: 11W at 10 :34:20), which is way below the RTX 3050’s typical 35–60W TGP.
Suspected Cause: BIOS/Firmware-imposed power limits or underpowered 135W adapter (especially during combined CPU + GPU load).

2. CPU Power Spikes + Throttling

CPU Package Power hits 54.85W at 10:34:30 far above the R5 5500U’s 25W TDP.
CPU EDC Limit: Hits 78.4% (electrical current throttling).
CPU PPT Limit: Peaks at 32.8% (power limit engagement).
Results in unstable clock speeds (400–4049 MHz) and CPU Tdie spikes to 55.6°C (not dangerous, but erratic behavior).

3. APU GFX Sensor Bug

APU GFX Temp is frozen at 87.0°C across all samples highly unlikely under light load.
Most likely a sensor bug or reporting issue, since discrete GPU temps are only 38–47°C.

4. High Memory Usage

Physical Memory Load >70% consistently; peak at 87% at 10:34:30.
With only 8GB RAM, ~5.3–5.8GB is used, leaving just 1.4–2.2GB free.
Could lead to paging and slowdowns during heavy multitasking or gaming.

Performance-Limiting Factors:

GPU Underutilization

RTX 3050 core clock fluctuates between 400–1800 MHz (idle speeds).
GPU Utilization stays under 10%, even during supposed load.
Power throttling is the primary limit, not temperature (no thermal throttling detected).

CPU Efficiency Issues

High average core clocks (up to 3100 MHz) but low effective clocks (123–1652 MHz) shows inefficient thread scheduling.
C6 Package Residency (idle state) above 80%, even during CPU load spikes. This could suggest background tasks or poor core management.

Summary:

No thermal throttling observed (CPU/GPU < 63°C).
But severe GPU power throttling and inconsistent CPU power delivery are impacting performance.
Fixing the GPU power limit and ensuring sufficient power delivery should be the first priority.