I discovered a highly parallel gaming scenario requiring over eight CPU cores powered by Zen 3 architecture.
I discovered a highly parallel gaming scenario requiring over eight CPU cores powered by Zen 3 architecture.
I've been experimenting with different games to assess how much performance I can achieve on my 5900X. Usually, the overall consumption stays under 33% according to HWInfo64, which aligns with the idea that 4c/8t CPUs are still sufficient for many titles. In more demanding games, usage rarely exceeds 50%, supporting the notion that "6c/12t is enough for modern gaming." Even the Singularity Escalation benchmark typically lands around 55% and occasionally reaches mid-60s. This suggests a 5800X would likely perform similarly. If such a benchmark tops out at 67%, it would clearly be adequate for gamers. It's important to note that no single core remains idle in these tests—each core contributes usage across all scenarios. However, the 5600X provides essentially no extra headroom beyond what’s needed, and the 5800X consistently stays within safe limits. In my experience, a game should never require more than this threshold. Except... when I loaded Metro Exodus Enhanced Edition from an NVMe drive, my 5900X maintained over 50% usage for a short period, reaching about 70.8%. That’s roughly 8.5 cores in action (including SMT). With an 8c/16t CPU, that would have pushed the limit to 100%. On such a setup, the load would have taken longer. I didn’t have other processes running, so this reflects only the game and Windows utilizing over 8 cores of Zen 3. While this isn’t unusual for games, it’s surprising to see such high demand on my machine. To be honest, this situation doesn’t occur in normal gameplay. (That exceeds my comfort zone at 40% usage.) But loading times noticeably improved compared to my previous CPU—a 9600K would have hit its limit during that load. I often paused to grab a drink or something while waiting, since the wait was lengthy. I’m not sure if I’d detect a difference with a 5800X without timing it, but this feels like a modern benchmark pushing beyond expectations.
I've been experimenting with pushing the 12 cores to their limits, but the game I discovered was quite unique. Minecraft really pushed my CPU hard, especially during unusual play sessions. When I tried spawning in a massive fireball with a radius of 1000, it caused a spike reaching 97% utilization across the entire CPU. While this isn't typical gameplay, it can definitely strain your processor.
Has remained consistent since Ryzen, especially as Vulkan and DX12 gain traction in numerous games.
I'm curious about what would happen if you upgraded from your 2060 model to a 3080/3090. With more frames, CPU usage tends to rise, especially when the Nvidia scheduler increases load. I've noticed in Cyberpunk it often hits around 60-75% of my cores during spikes, pushing heavily into high frame rates while using lots of hardware. I managed it with specific settings—10850K, 5.1Ghz, 4266C17, 2080Ti at high detail, 1080p DLSS. It's interesting how lowering draw call details affects performance, just like you're wondering.
Cyberpunk 2077 can handle up to 12c or 24t smoothly. Right now it's not the norm, but I think games will begin leveraging more cores soon. Much of this depends on console sales since that’s the closest reference point we have. With so many different computer setups, it’s hard to build a game for one specific PC spec. You can still target a particular console hardware and then adapt it for a PC with similar power. This is the first generation where consoles offer more than seven slow threads, so developers using this new speed will likely bring it over to PC projects as well.
It's accurate. The 2060 Super isn't particularly weak, though the GPU is currently holding the CPU back. It's running at 98%+ across most modern games, especially when I'm playing at 1440p. Once GPU costs stabilize and upgrades become feasible, I plan to conduct further tests to explore the limits.