Perf tracking: cumulative impact of optimization PRs

[dyxushuai]

Goal

Track cumulative performance impact across a series of optimizations (PRs 286/287/288 and future) so we can quantify end-to-end gains.

Method

• Command: zig build bench (record iterations and terminal size)
• Use a consistent baseline commit on main for comparisons
• Record local machine details (CPU, OS) once and reuse

Tracking

• perf: cache grapheme widths in Window.print word wrap #286
• perf: add ASCII fast path in parseGround #287
• perf: SIMD scan ASCII runs in input loop #288
• One more WIP

Results (update as PRs merge)

PR	Benchmark case	Baseline ns/frame	After ns/frame	Δ%	Speedup	Notes
286	print word-cache (small/med/large/overflow)	TBD	TBD	TBD	TBD	fill in from PR 286 bench
287	parseGround ASCII fast path	TBD	TBD	TBD	TBD	fill in from PR 287 bench
288	loop ASCII SIMD scan	TBD	TBD	TBD	TBD	fill in from PR 288 bench

Cumulative

• Baseline commit: TBD
• Combined result: TBD (update after each merge)

[dyxushuai]

Hi @rockorager , sorry for your time. I opened this issue to track the cumulative benchmark impact across my perf PRs (286/287/288 and one more in progress). This should make it easier to see end‑to‑end gains
over time. I’ll update the table as new optimizations land.