Different workloads CPU

[Niekao]

Hey,

 

I was wondering, I always see people talking about different workloads, but what other categories are there than gaming?

And how is that going to affect your CPU?

 

Thanks!

[sphbecker]

26 minutes ago, Niekao said:

Hey,

 

I was wondering, I always see people talking about different workloads, but what other categories are there than gaming?

And how is that going to affect your CPU?

 

Thanks!

Off the top of my head, here are some popular ones (far from a complete list)

• Gaming (lightly threaded, uses about 4 cores, favors high clocks)
• Video encoding/streaming (heavily threaded, uses all available cores, can be offloaded to GPU in some cases)
• Photo editing (Photoshop favors Intel due to Adobe's current optimization)
• Graphic rendering (heavily threaded, uses all available cores, GPU can typically be used to accelerate)
• Mathematical modeling (workloads can very, typically high clocks are favored)
• Engineering/CAD (very similar workload to gaming, but with more emphasis on RAM and I/O)
• Code compiling (I/O and cache are king, cores and clocks help, but neither are the key factor)
• AI/Deep learning (massively parallel workload, typically given to GPUs or super high core count CPUs)
• File processing (needs dedicated PCIe lanes for each SSD, cache and RAM help, cores and clocks don't typically matter unless doing media compression)

[Slottr]

Software development, video editing, image editing / graphic design, 3D rendering, web development, streaming, folding, etc 

 

how it affects performance largely depends on what type of work it is, what software you’re using and how you’re using it.

[Niekao]

59 minutes ago, sphbecker said:

Off the top of my head, here are some popular ones (far from a complete list)

• Gaming (lightly threaded, uses about 4 cores, favors high clocks)
• Video encoding/streaming (heavily threaded, uses all available cores, can be offloaded to GPU in some cases)
• Photo editing (Photoshop favors Intel due to Adobe's current optimization)
• Graphic rendering (heavily threaded, uses all available cores, GPU can typically be used to accelerate)
• Mathematical modeling (workloads can very, typically high clocks are favored)
• Engineering/CAD (very similar workload to gaming, but with more emphasis on RAM and I/O)
• Code compiling (I/O and cache are king, cores and clocks help, but neither are the key factor)
• AI/Deep learning (massively parallel workload, typically given to GPUs or super high core count CPUs)
• File processing (needs dedicated PCIe lanes for each SSD, cache and RAM help, cores and clocks don't typically matter unless doing media compression)

Thnx