You can typically see the L1/L2 SRAM cache regions just by looking for continuous patterns, then try to divide it into the number of cores or core groups you're looking for.
This is only applicable to die photographs of the silicon layer.
For CT scans, which use X-rays, the visible structure is very likely due to the metallic (copper) interconnect, clock- and power-delivery layers, which are generally quite different from the underlying transistors.
These layers are larger-scale, and copper also absorbs X-rays much more strongly than silicon.
Comments
You can typically see the L1/L2 SRAM cache regions just by looking for continuous patterns, then try to divide it into the number of cores or core groups you're looking for.
This is only applicable to die photographs of the silicon layer.
For CT scans, which use X-rays, the visible structure is very likely due to the metallic (copper) interconnect, clock- and power-delivery layers, which are generally quite different from the underlying transistors.
These layers are larger-scale, and copper also absorbs X-rays much more strongly than silicon.
Ah, I see. Thank you for the correction.