Why on earth would it sting?
- I have both Intel and AMD and love them both equally.
- Unlike SOME, I do not have a narrative to espouse in making one appear better than the other.
As I have both systems I do not need to look at slides or comparisons on Puget or any other site, I can do the tests myself for my exact usage which often is not replicated by Puget etc.
I know my conclusion for my usage does not fit in with what seems your particular narrative (hence the quick to conclude "no filters for 9700K" is something surreptitious) but there are a few who just want an unbiased view point based on empirical, real usage data.
Like I asked before, are there any other tests you want me to run? You never know, the 3900X might come out on top.
I've owned all sorts of CPU's in my time, i5 and i7's of various types, Athlon and Phenom CPU's, i even had an FX9590 for a while, great heater but not much use in a Desktop, My last AMD GPU was an R9 290, which was a great GPU but really the last such thing from AMD.
Anyway, the only narrative is the raw data. You told me ST should not be extrapolated for IPC as its not accurate and distorts the narrative. IPC is how much work a CPU can do at a given clock speed, if CPU A is 100% at 4Ghz and CPU B has 90% the IPC of CPU A then CPU B needs to run at 4.4Ghz to do the same work as CPU A, this is why its important, yes if you run CPU B at 4.8Ghz then you get 110% the performance of CPU A, that doesn't make IPC any less relevant, TNA runs his 3600 at 4.4Ghz 24/7, its takes a 5Ghz 8700K just to match it, at twice the power consumption and twice the heat output, even ignoring the cost difference that distinction is not skewed and it 'alone' matters. Yes it makes one better than the other, my FX9590 beat i5's in MT workloads at 5Ghz, was it a better CPU?
You told me your main usage is photo editing, and then showed me a slide of image export comparisons when your 9700K was 2 Seconds faster in that.
To quote what i said in another thread.
You have to make a video before you export it, a 16 core will encode that video much faster than a 10 core, the fact that the 10 core will export or scratch through the video during the editing process just as well or as fast as the 16 core is neither here nor there really, unless you're just a hobbyist and your videos don't take more than 30 minutes to encode on an 8 core, in which case you wouldn't need a HEDT CPU. If however your time is money...... then you want the best encoding performance.
Another discussion where someone made the point that "Exporting" was just as quick on Intel as it was on Zen 2, you're using the same narrative... you have to do the work before you export it, in that Zen 2 is faster, if it takes you 25 minutes to render the image on Zen 2 vs 30 minutes on Intel then a 2-second difference (Or all things being equal, i suspect actually none) in exporting that image is completely irrelevant and to focus on such an argument instead of the actual performance difference during the workload you have to go through before you get to that stage is ridiculous and smacks of Ryan Shrout style distraction tactics.
