Here are my thoughts.

Bosshart LISP chip.

What was quoted in the request for reexam is probably all they have. "the on-chip RAMs which dominate chip area and transistor count." That statement is less clear than our "occupying a majority of total area of said substrate," which was unacceptable to examiner and required the patent owner to document and clarify further. Since the LISP chip area is undefined - whether active area, total area or something else, I can't see examiner allowing it. IMHO.

The other quote "80% of the transistors on the LISP chip are incorporated into memories" says absolutely nothing with respect to the surface area involved. And that 80% would necessarily be referring to active area only, not total area. The statement is meaningless and can't see the examiner allowing it either. IMHO

Uhlenhoff.

If there is actually a ring oscillator on-chip, it does not clock a CPU. And even if examiner disagreed with the definition of CPU, it is not a system clock. The patent clearly indicates a different clock is used as the system clock. Inclusion of this one is really wasting everyone's time. IMHO

Transputer.

Yes the transputer has a phase lock loop PLL. Yes TPL said that the existence of a PLL indicated the presence of a ring oscillator. Problem is TPL was referring to specific microprocessors in the claim charts that have certain operating characteristics. I do not believe the existance of a PLL necessarily always means the presence of a ring oscillator. Just more BS IMHO.

HTC pointed out that 25% of the transputer chip is memory and CPU is about 25%. In a different model the memory is doubled, so conceiveably the memory is double the CPU on that chip. And it would be obvious to increase memory to a greater %. Of course they do not mention the cost of adding that much memory, or if cheaper DRAM were used, how to get around the slowness of DRAM. Or the fact that the initial examiner didn't think it was obvious. IMHO.

Look forward to reading what others have to say.

Opty