Why Was Intel's Westmere Considered 1st Generation and Not a 2nd Generation Part?

[BuckyJunior]

After doing some research on some of Intel's legacy CPUs, out of curiosity, I had developed an interest as to why Intel's Westmere die shrink was not considered as second generation CPUs. Sandy Bridge, at 32nm, was second generation, and the Ivy Bridge die shrink as the third generation, so why wasn't Nehalem (45nm Lynnfield/Bloomfield) kept separate from Westmere? Westmere introduced Core i3, i5 6xx, and the 980X/990X as well as architectural tweaks, so how could it not be second generation? My guess is that they used Westmere to complete the total product stack with Core i3, i5 6xx, and the final i7EE CPUs. I also found it somewhat strange that Intel decided to redesign their naming scheme when Sandy Bridge released,

 

[k1114]

In this case, generations is being used as a marketing tool, not a division of actual technological steps. Just like how 8th gen covers, 3 architectures, kaby, coffee and cannon. Hedt numbering is a whole gen behind. Sb-e is under 3rd gen, ib-e is 4th gen, etc. Westmere also didn't get a full line of cpus and was treated more along the lines of a refresh. I'm not even sure there was much of a difference in ipc or other improvements like other die shrinks. The lineup can already be said as complete. They had plenty of skus.

I don't see how sandy was a change in naming. If you have a 1st gen but have no 1 used in the name, how would you name 2nd gen keeping the same naming scheme? Add a 2 in front as sandy did.

 

[COLGeek]

Go info here: https://en.wikipedia.org/wiki/List_of_Intel_CPU_microarchitectures