In LTT’s recent video they simulated the performance of the 9700K by disabling HT on the 9900K. And as I predicted, it does not perform worse in multi-threaded workloads than the 8700K despite the lower thread count. Should actually show even more improvement over the 8700K with overclocking as the large die size with each core doing less work should be easier to cool.
Just goes to show we shouldn’t judge hardware before we see the benchmarks.
If anyone has a 9700K, I’d love to see Cinebench scores of it OC’d.
(Yes I know it costs more and Intel is the devil)
It’s funny because with each generation, I keep thinking about when to upgrade. The 9700k seems like a great performer (read the AnandTech review).
I just still haven’t felt the need to pull the trigger yet though. I’m on Devil’s Canyon i7-4790k, and I overclock that to a stable 4.6Ghz 24-7 with a Noctua cooler, and it performs adequately…
1 Like
Depends what else you’re running.
Unless you’re rocking a top flight GPU and run in low res, anything sandy bridge or more recent is probably “good enough” for most stuff (read: GPU accelerated graphics, games, general web/office stuff, etc.).
But, there are instances where newer processors will be much faster, particularly at things like h.265, possibly encryption, etc.
The big driver lately is if you are doing things like virtualisation which need more cores. If you aren’t doing that then 4-6 cores is currently “enough” (and as an aside this is why AMD is so hot right now; 4ghz is “fast enough” for most people easily, but more cores = you can cut the CPU up into more pieces for better responsiveness under load with multitasking or VM use).
Pretty sure even at same clock/core count, even skylake is about 25% faster than something like haswell, and even faster vs. sandy or ivy bridge.