--- In [email protected], n...@... wrote: > I wouldn't be so sure. Like PICs, the ATmega644P is an 8-bit device. Max. > clock speed is 20 MHz. Some high end PICs can run @ 40 MHz. > > The TT3+ used a mid-range PIC (16F series), which probably wouldn't be up > to the task of decoding, & at the assembly level the mid & high range PICs > aren't code-compatible, so porting code from the mid-range probably > wouldn't be any easier between the high-end PICs & the Atmel. > > Bob NO6B
Hi Bob, PIC and AVR are significantly different architectures. They can't be compared on a clock cycle basis. PIC's execute 1 instruction per 2 clock cycles. AVR's execute 1 instruction per 1 clock cycle for the majority of instructions. The 16 series PIC's has a single 8-bit register (accumulator) and you have to constantly shuffle data back and forth into that register, wasting clock cycles. AVR architecture has 32 8-bit registers. There's a significant difference in how much work you can perform for a given clock rate between PIC's and AVR's, with AVR's having a significant advantage there. So, in terms of instruction throughput, a 20MHz AVR = 40MHz PIC. Then the additional performance advantage due to the AVR's 32 registers depends on the actual code or algorithm in use. -Ryan
