F5F Stay Refreshed Hardware Desktop Blue screen interrupt occurred during restart due to PBO2 issue.

Blue screen interrupt occurred during restart due to PBO2 issue.

Blue screen interrupt occurred during restart due to PBO2 issue.

D
DarkBoy__YT
Posting Freak
Find
898
10-25-2018, 04:33 AM
#1
I’m a Chinese DIY gamer who bought a cardless host in early January:9700x. My setup included Asgard Valkyrie v2 16*2 6000c28, ROG STRIX B850-A for gaming Wi-Fi, p44 pro, Deepcool PQ1000P, Titan A080, and a 4TB HDD. Until I got the card, I ran Expo1 with PBO2, set negative voltage to 30, and everything worked smoothly for online play. After the 9070xt came out, I got a piece of the xfx 9070xt Phoenix. Then things started to go wrong. Stress and 3DMark tests showed no issues, but playing PUGB would freeze with a black screen and a reboot after 10–20 minutes, reporting kernel power 41. Playing Apex or League of Legends sometimes caused the same black screen and long restarts. Updating BIOS didn’t help either. Eventually I returned the card, checked with PBO2 and Expo off, and updated the motherboard BIOS—still no fix. When I tried PUGB with core output and PBO2 enabled, frames appeared in single digits but no black screen. After that, I bought a Galaxy 5070Ti with driver version 70. After about 20 minutes of testing PUGB, it stopped working, so I thought the card was the culprit. When I ran Core Output tests with Expo and PBO2 on, frames appeared normally, but no black screen. Later, I got a Galaxy 5070Ti with driver 70, and after 20 minutes of PUGB testing it failed again. The error code changed each time, and Apex would flash back into play after normal gameplay. Stress tests and 3A were stable. I began suspecting a memory issue—AMD DDR5 ANTA777 reported error 3, which resolved when VDDQ voltage was increased to 1.05. But even then, two errors persisted in half an hour. Lowering the negative voltage from 30 to 15 didn’t help; it still froze and restarted. Turning off PBO2 fixed it, and games like Apex no longer flashed back. I measured sub-core voltages with Hydra and found all readings at 28V—far below my expectations. I’m confused about the problem. When I don’t install a video card, negative voltage 30 works fine; I install an AMD card and reboot with or without PBO2, and use a NVIDIA card with PBO2 enabled, which is strange. Also, when testing memory crossover at 8000Hz, the BIOS would flash on the motherboard, but my tester didn’t notice any issues. Still, I’m unsure if it’s a CPU fault, memory problem, motherboard issue, or something else. I’m also puzzled by the high power draw of my 5070Ti (90% with MSI Afterburner, +2000 MHz memory frequency), which pushes the voltage curve to 3262MHz at 1.025V. What’s going on? Is it a bad CPU, faulty RAM, motherboard fault, or just a memory issue? I’m trying to figure out whether the power supply, hardware, or software is the root cause.**
Reply
Reply
D
DarkBoy__YT
10-25-2018, 04:33 AM #1

I’m a Chinese DIY gamer who bought a cardless host in early January:9700x. My setup included Asgard Valkyrie v2 16*2 6000c28, ROG STRIX B850-A for gaming Wi-Fi, p44 pro, Deepcool PQ1000P, Titan A080, and a 4TB HDD. Until I got the card, I ran Expo1 with PBO2, set negative voltage to 30, and everything worked smoothly for online play. After the 9070xt came out, I got a piece of the xfx 9070xt Phoenix. Then things started to go wrong. Stress and 3DMark tests showed no issues, but playing PUGB would freeze with a black screen and a reboot after 10–20 minutes, reporting kernel power 41. Playing Apex or League of Legends sometimes caused the same black screen and long restarts. Updating BIOS didn’t help either. Eventually I returned the card, checked with PBO2 and Expo off, and updated the motherboard BIOS—still no fix. When I tried PUGB with core output and PBO2 enabled, frames appeared in single digits but no black screen. After that, I bought a Galaxy 5070Ti with driver version 70. After about 20 minutes of testing PUGB, it stopped working, so I thought the card was the culprit. When I ran Core Output tests with Expo and PBO2 on, frames appeared normally, but no black screen. Later, I got a Galaxy 5070Ti with driver 70, and after 20 minutes of PUGB testing it failed again. The error code changed each time, and Apex would flash back into play after normal gameplay. Stress tests and 3A were stable. I began suspecting a memory issue—AMD DDR5 ANTA777 reported error 3, which resolved when VDDQ voltage was increased to 1.05. But even then, two errors persisted in half an hour. Lowering the negative voltage from 30 to 15 didn’t help; it still froze and restarted. Turning off PBO2 fixed it, and games like Apex no longer flashed back. I measured sub-core voltages with Hydra and found all readings at 28V—far below my expectations. I’m confused about the problem. When I don’t install a video card, negative voltage 30 works fine; I install an AMD card and reboot with or without PBO2, and use a NVIDIA card with PBO2 enabled, which is strange. Also, when testing memory crossover at 8000Hz, the BIOS would flash on the motherboard, but my tester didn’t notice any issues. Still, I’m unsure if it’s a CPU fault, memory problem, motherboard issue, or something else. I’m also puzzled by the high power draw of my 5070Ti (90% with MSI Afterburner, +2000 MHz memory frequency), which pushes the voltage curve to 3262MHz at 1.025V. What’s going on? Is it a bad CPU, faulty RAM, motherboard fault, or just a memory issue? I’m trying to figure out whether the power supply, hardware, or software is the root cause.**

Reply
Reply