Originally Posted by jfriend00
Is something like a 9th generation i5-9600k, an original chip design itself with only 6-cores in the chip layout or is it actually a down-binned 9th generation i7-9700k that likely had some issues with 8-cores and only has 6 cores enabled?
If the i5-9600k is an original chip design, what does Intel do with an 8-core chip that has 6 good cores? Is it scrapped or are there other parts of the product line where it gets used as a different product number?
That's pretty much it.
Although it's not that simple in detail.
All Intel chips on the same process come from huge wafers. Then they use some sort of way to test for silicon purity, and then enough cores that match the same purity get put onto a chip sample, depending on what tier it's in. Then the CPU VID is stamped into all the cores (as you already know, the lower the CPU VID at a set cache frequency (NOT core frequency), the better quality the CPU usually is), and the VID should be within 5mv on all of the cores (yes you can have a slightly weak core, that may fail 'first' within a borderline voltage range, like, core 4 failing at 5 ghz at less than 1.26v) but they won't be too far off overall. So in general, all cores that get put on a package for a chip to be sold should be the same quality.
With 9900K's and 9700K's the two highest bins, what fails 9700K's bin quality may not necessarily be put into 9600K's. There's also the "non K" chips that have to get sold, as well as the "BGA" versions. On the 7700K vs 7820HK kaby lake models for example, while the 7820HK was an unlocked chip with a lower TDP (45W) that could exceed it based on Bios and EC code restrictions for laptops, the 7820HK had worse silicon quality than the 7600K and 7700 non K. I probably should be using 8920HK vs 8700K (or 8086K).
Someone on notebookreview posted the binning charts for the Kaby models, but none have been posted for CFL and CFL-S. We can assume it's not too different.