aboutsummaryrefslogtreecommitdiffstats
path: root/docs/research
Commit message (Expand)AuthorAgeFilesLines
* 115: the modules replaced are freed (0 bytes a change over 4,000, was 19,284)...eunhhu29 hours1-2/+31
* 140: determinism across builds at 027, 028 and 140's state: record_all.py's 4...eunhhu31 hours1-0/+14
* 140: LIST_VALUE's copy reviewed before building (measured by spec: UAX #14 ab...eunhhu31 hours1-0/+63
* 140: aup's frictions written away and the apps measured again (no core change...eunhhu31 hours1-3/+41
* 140: the decision section before 8 and 10 points to their sections; 8's Pytho...eunhhu33 hours1-1/+3
* 140's 8, box layout, written in aup with no new opcode (aup's examples/layout...eunhhu33 hours1-1/+78
* 140's 10, RFC 6298's retransmission timer, written in aup with no new opcode ...eunhhu33 hours1-1/+59
* 028 and 140: list values' copying measured a second time (UAX #14's Text reco...eunhhu33 hours1-0/+2
* 140's 5, line breaking (UAX #14, Unicode 16.0), written in aup with no new op...eunhhu33 hours1-1/+65
* 140 measured again without the workarounds: the three specs' checks pass on t...eunhhu34 hours1-3/+26
* 140: small block fns inlined in aup, made and measured, not merged (the patch...eunhhu34 hours1-0/+16
* 140's 9, HTTP/1.1 requests a packet at a time (RFC 9112), written in aup with...eunhhu36 hours1-2/+87
* 140's DNS: the compiled path's 11% stated as measured (244 to 218 ns), which ...eunhhu36 hours1-1/+1
* 140's 1, DNS messages (RFC 1035), written in aup with no new opcode (aup's ex...eunhhu36 hours1-2/+70
* 140's inflate: the three workarounds' owners: a fn taking no bare list and a ...eunhhu36 hours1-0/+9
* 140's 3, inflate (RFC 1951), written in aup with no new opcode (aup's example...eunhhu36 hours1-2/+61
* 140 after 027: the ten probes compiled again (run.sh writing to a temporary d...eunhhu37 hours1-13/+47
* 027's 5c: money as i64 minor units shown by .Nd: NUM_TEXT's decimal form take...eunhhu37 hours1-0/+2
* Research 140, 027's stage 3 measured: a board of columns of cards of tags (au...eunhhu45 hours1-0/+60
* Research 140, 027's stage 2c measured: JSON read into a tree by recursive des...eunhhu46 hours1-0/+65
* Research 140: 2b's instruction counts: the workloads' modules the same, batch...eunhhu7 days1-0/+8
* Research 140, 027's stage 2b measured: JSON read into a tree (aup's examples/...eunhhu7 days1-0/+46
* Research 140: 2a's instruction counts: the workloads' modules the same, the e...eunhhu7 days1-0/+12
* Research 140, 027's stage 2a measured: whether bytes are JSON (aup's examples...eunhhu7 days1-0/+41
* Research 140: stage 1's instruction counts: the workloads' modules byte for b...eunhhu7 days1-1/+15
* Research 140, 027's stage 1 measured: SHA-256 written in aup (its examples/sh...eunhhu7 days1-0/+60
* Research 140: a suite of eleven specs from different fields (DNS, SHA-256, in...eunhhu7 days1-0/+130
* Batch executor: the round's fewest-left marks at the end of Batch: put after ...eunhhu9 days1-0/+16
* Research 139: what a program made at load per batch loop would take (the deci...eunhhu11 days1-2/+49
* 043 programs (research 139, in progress): a batch loop's body as the program ...eunhhu11 days1-0/+52
* Research 138: the batch executor's stage 5, design 025 made: sums into cells ...eunhhu11 days1-0/+79
* Research 137: the batch executor's stage 4 (columns read in place, comparison...eunhhu12 days1-0/+66
* Research 136: the batch executor's stage 3 (four-byte MUL, reductions with no...eunhhu12 days1-0/+65
* Research 135: batch plans' stage 2 measured (b11 = 7959885 against b10, inter...eunhhu12 days1-0/+66
* Research 134: batch plans' stage 1 measured (b10 = a6fabf0 against b9, interl...eunhhu12 days1-0/+72
* Research 132 and 133 corrected: the per-batch cost was icount's three rounds ...eunhhu12 days2-10/+14
* Research 133: register tiles measured (132's deciding experiment): past the b...eunhhu12 days2-1/+89
* Research 132 (no code): where the batch executor's cost sits (an element's pa...eunhhu12 days1-0/+166
* Research 131: the batch executor's general cuts, by the user's choice of the ...eunhhu12 days1-0/+95
* Research 130: the batch executor's frontiers (a count's guard, walks in a bat...eunhhu12 days1-0/+114
* Research 129: the batch executor's costs cut (appends at once 5.18 to 0.65 ms...eunhhu2026-10-061-0/+124
* Research 128: the batch executor's appends, recurrences, sums into cells and ...eunhhu2026-10-061-0/+147
* 024, 126, 127: the lanes are 24 buffers (stage 2), about 200 KB a thread; sta...eunhhu2026-10-052-2/+4
* Research 127: the batch executor's stage 2 (nested loops' rounds run together...eunhhu2026-10-051-0/+65
* Research 126: the batch executor's first stage (024) as built and measured: k...eunhhu2026-10-051-0/+100
* 125: forms and newton runs named; BFS's time after the graph's load is the ge...eunhhu2026-10-051-3/+5
* Research 125: the tier's first steps against C (e2e56dd: roots 1.40, knapsack...eunhhu2026-10-051-0/+80
* Research 124: 022's BFS past C only by reading ahead (15.7 ms against 9.9 wit...eunhhu2026-10-051-0/+89
* Measure 021: BFS 190 to 90 ms over HTTP against C's 9.1, short of C; Pure's f...eunhhu2026-10-051-0/+216
* The same answers on the final build and the same pure-runs on the Pi, WSL2 an...eunhhu2026-10-041-1/+5