GamePP Frequently Asked Questions - Professional Hardware Monitoring Software FAQ Knowledge Base

The raw power of this card is beastly, but the drivers are a total nightmare. Walking through Kyoto, I'd get these random freezes that made the game feel like a slideshow. The Sapphire RX 9070 XT 16G was struggling with high-frequency lighting calculations, causing the shader compilation queue to pile up in the background, with frame times swinging wildly between 15-40ms. I jokingly tried maxing out every single setting, but that just made the drops worse. I decided to do a full purge using DDU, installed the latest official Beta driver, and wiped 8.2GB of old shader cache. The performance monitor showed the frame time variance settling into a smooth 12-16ms. I did have some brief screen flickering right after the Beta install, but a fresh install of the DirectX runtime sorted it out. Core temps stayed between 65-72℃, and fan speeds stabilized at 1400-1600RPM. Last updated onMarch 14, 2026 8:29 PM.

This cooler's limit was seriously testing my patience. In massive charge scenes, I'd experience a brutal dive from 60 FPS to 15 FPS in about three seconds. After an hour of gaming, the AK500 ARGB just hit total thermal saturation, with core temps pinned at 95℃ and clocks crashing from 4.8 GHz to 3.0 GHz. It was honestly pathetic. I tried limiting the CPU TDP to 65W in software, but the calculation speed slowed down so much that I lost 20 FPS overall—that's not a solution, that's just handicapping the PC. I ended up rigging a 120mm intake fan at the bottom of the case to feed cold air directly into the fins. In Cinebench, my multi-core score climbed from 14,000 back up to 16,800, with peaks held at 82-86℃. I actually bumped the RAM gold fingers while installing the fan, which caused a boot failure, but a quick reseat fixed it. CPU now runs at 78-84℃. Exported all the stability data via system analysis tools. Last updated onFebruary 18, 2026 9:41 PM.

The frame rate would suddenly dive to 20 FPS, turning the game into a slideshow. It was honestly ridiculous. The PCIe bus on the Onda 9D4-DVH was having a meltdown trying to handle concurrent requests from the NVMe drive and the GPU, leading to 15-25ms of contention latency. I wasted time updating every single Windows driver, but the bandwidth utilization was still jumping all over the place. I eventually went into the BIOS, changed the PCIe link mode from Auto to a forced Gen 3, and disabled all the useless power-saving ports. AIDA64 bandwidth tests showed read speeds jumping from 2100MB/s to over 3200MB/s. The hitching during scene transitions completely vanished. I did notice a slower boot time (about 8 seconds longer) after the change, but a manual microcode update sorted that out. Core temps are steady at 45-52℃ and the fans are humming along at 1400-1600RPM. Last updated onFebruary 24, 2026 12:56 PM.

It's honestly ridiculous that a 16GB card can be completely choked out by this game; by the third hour, the whole thing turns into a slideshow. The VRAM on my Sapphire RX 7800 XT slowly crawled from 9GB up to 15.5GB, a textbook memory leak that eventually just killed the app. Restarting the game only bought me another 30 minutes, and that cycle of frustration almost made me throw my keyboard. I ran a memory analyzer and found a ton of redundant texture data that wasn't being released, so I set up a script to force-clear the DirectX cache every hour. In Resource Monitor, the VRAM usage finally stabilized into a valley between 13-15GB instead of just climbing in a straight line. I actually messed up and deleted some pre-compiled shaders during the process, which added two minutes to my next load time—lesson learned. Now, the GPU stays between 62-68℃ and power draw is around 190-220W. The logs confirm the leak is suppressed, and fans are chilling at 1400-1600 RPM. Last updated onMarch 22, 2026 12:25 PM.

Running 6000MHz RAM with this game's optimization felt like a lottery—one minute it's buttery, the next it's a slideshow. In the dense areas of the Shadow Realm, the memory controller hit response peaks of 15-30ms, causing the frame times to jump all over the place. I tried lowering the resolution to ease the load, but while the FPS went up, the stuttering frequency didn't budge—complete waste of time. I headed into the BIOS, bumped tRFC from 480 to 560, and locked the voltage at 1.38V for absolute stability. RTSS showed my 1% lows jump from 35 to 52 FPS, and the screen tearing practically vanished. I did crash the game a few times during save loads when I tried to push the timings too low, but backing off fixed it. Temps hovered between 52-58℃. Exported the frame time logs and the results are clear. Last updated onMarch 12, 2026 1:50 PM.

Back to Top