High Core Speed, High Core Count, or Doubling Threads—each impacts performance differently.
High Core Speed, High Core Count, or Doubling Threads—each impacts performance differently.
I've noticed many discussions about CPUs, but I'm still figuring out how core speed, count, multi-threading, and CPU cache influence applications. For instance, games typically don't require many cores even with a 6-core processor. Also, Linus improving his rendering station by boosting clock speed can make a noticeable difference. Can anyone clarify this in more detail? I'm curious but want a clear explanation without relying on searches.
It's not always straightforward. You're correct that many games rely on just six threads, sometimes eight. Adding more threads or cores doesn't always boost frame rates. When you test two systems with the same CPU and higher clock speeds, the one with better performance will win. Keep in mind that comparing clocks directly across different CPUs isn't reliable because their architectures differ. Each processor handles calculations differently per clock cycle. For example, a CPU with eight cores at 3GHz might beat another with eight cores at 4GHz. Certain tasks, like rendering, benefit from more cores or threads, not just faster ones. If you're into gaming, streaming, or music, using more than six or eight threads can help. The best approach depends on your specific needs. Also check out @Skiiwee29's links for more clarity.
That R5 outperforms the i3 a lot, handles things smoother and uses less power. Probably because it has more capacity to divide tasks, unless you're multitasking heavily. But using over 8 cores becomes pointless since the game won't actually use them. Unless you're streaming or running another process in the background. That's why my i5-7200U (with both turbo and hyper-threading) is pretty much struggling to play Dota 2 and stream at the same time. I hadn't used QSV back then, though—I thought it was better than x265 on my laptop, at least. As for that Premiere video, the GPU with CUDA really speeds things up compared to just counting CPU cores. That's because GPUs are built for parallel tasks like encoding, especially with CUDA cores. It depends on the platform; I tried playing 4K on my 7200U and 940MX, and sometimes it works fine. I have some 4K movies that look great on my 768p screen, so it doesn't seem like a big issue. Maybe that's why some systems still run smoothly?