Cinebench question
- Thread starter ssal
- Start date
Intel_man
VIP Member
Not everything is optimized or require multi-core processing.
You'll also get anomalies like this... where in theory, it's multi-core performance claims to match an i9-9900k, but in real world use cases, it probably runs like a turd.
https://www.tomshardware.com/news/h...tches-core-i9-9900k-in-multi-core-performance
You'll also get anomalies like this... where in theory, it's multi-core performance claims to match an i9-9900k, but in real world use cases, it probably runs like a turd.
https://www.tomshardware.com/news/h...tches-core-i9-9900k-in-multi-core-performance
Why bother testing multi core in that case, or anything at all?
Different types and styles of code, different workloads, some workloads respond differently than you'd expect on paper.
I find single core testing is valuable as you can contrast the multicore scaling off of it to get a more full picture of what performance would be under particular workloads, which is the primary reason benchmarks exist anyway. Plus most of the code in the last few decades were very single thread centric.
Different types and styles of code, different workloads, some workloads respond differently than you'd expect on paper.
I find single core testing is valuable as you can contrast the multicore scaling off of it to get a more full picture of what performance would be under particular workloads, which is the primary reason benchmarks exist anyway. Plus most of the code in the last few decades were very single thread centric.