Stuttering and low frame rates on Ryzen 4500 with NV3070
Stuttering and low frame rates on Ryzen 4500 with NV3070
I'm experiencing low frame rates and stuttering while playing games. My system has a Ryzen 5 4500 with a b450 chipset and a 16GB DDR4 2666MHz RAM setup. I upgraded to an RTX 3070 and a 4500 series motherboard, but performance drops significantly during intense scenes—especially when aiming or using high-detail settings. On the old GTX 970, I managed around 50 FPS, but now with the RTX 3070, I hit 30–55 FPS depending on the game and settings. Despite limiting my frame rate to 60, I still see drops to around 40 FPS or lower. I suspect the CPU might be limiting the GPU performance, but I'm not sure how to confirm. I've checked my VRAM settings and haven't found any unusual values in SpecView, and the MSI afterburner and HWMonitor show normal readings. I've tried adjusting settings in NVIDIA Control Panel, but nothing helps. It's frustrating because benchmarks show the RTX 3070 should easily sustain over 60 FPS at 1080p. Could you help figure out the cause?
Also attempted capping at 60fps to align with my monitor's refresh rate and avoid tearing. Used VSync and in RTSS. No luck—fps stays the same regardless of in-game or NVIDIA settings. Edit: mistakenly double-checked the "solution" tag, lol. Thought it was a positive sign. Just new to the forum, sorry.
Nope, that has to do with power allocation, right? I'll try it when i get back home, but both msi aft and hwmonitor don't show either the cpu or gpu struggling for power. Cpu usage is nowhere near max (confusing and maybe colliding with my theory of cpu bottleneck), and gpu is shown at 95-98% usage. I'll try some synthetic benchmarks on the gpu, maybe furmark or something and compare them to ones i find online. I'm getting paranoid the gpu is f'd, since i bought it 2nd hand (i did see it installed and playing different games in the original owner's pc with no problem though).
Typically a CPU limitation is present. The R5 4500 had a unique Zen 2 design, offering only 8MB of L3 memory compared to the 3500x’s 32MB. Think of it as a R5 4600g without the Vega GPU cores turned on. Improving performance with 5600/5600x/5800x3D would be a solid step. The B450 also wouldn’t support PCIe 4.0, though the RTX 3070 doesn’t really utilize it. Stick to the 5600/5600x, as the 5500, 5600g/5700g, and 3000-series chips all perform poorly in games. I started this discussion to illustrate how diverse CPU bottlenecks can be, using the most straightforward example. Mapping any CPU from 4790k to 7950x3D would clearly place the 4500 somewhere in the middle. Warframe 4K ultra is a game that varies widely—capable of up to 1200 fps at 1080p ultra but dropping to 120 fps during intense multiplayer scenarios. It serves as a strong benchmark for overall gaming performance in rasterization.
5700g features a distinct design to fit Vega GPU cores. The main trade-off is the removal of L3 cache. It’s a solid APU, surprisingly strong on the integrated graphics side, but not ideal if you already have a dedicated GPU. The 5600/5600x comparison resembles the debate around the AM4 5800x3D versus current AM5 options. The 5600/5600x trade-off isn’t as strong as the 5800x3D, but offering half the cost without doubling gaming performance is hard to dispute. The AM5 example with 7600 versus 7800x3D shows a similar pattern—double the price without matching the performance boost. In either case, the extra $150–$200 might justify it for some, but for a gaming PC focused on value, the 5600/5600x (AM4) and 7600 (AM5) options seem compelling. If you already have a good motherboard and RAM setup, upgrading to AM5 could be worthwhile, though you might prefer a direct jump rather than chasing higher prices. The core ideas in this video translate well to the 5600 vs 7600 comparison. Most points hold, especially if you’re thinking about reselling and upgrading to AM5 instead of just swapping CPUs.