Laderman's 1975 paper about the 3x3 case in 23 multiplications is nice: https://www.ams.org/journals/bull/1976-82-01/S0002-9904-1976-13988-2/S0002-9904-1976-13988-2.pdf
It explains how he reached it by solving some huge algebraic system by hand. A lot of fancy algebra went into the later improvements. I haven't looked at Bläser's result but it sounds like 19 is a lower bound rather than an upper one? Anyway https://en.wikipedia.org/wiki/Computational_complexity_of_matrix_multiplication goes into this, and it's informative.
solrize
npub14x…kc6zp
2026-01-02 09:09:08 UTC
in reply to nevent1q…f4zv
Author Public Key
npub14xdtfxpgz4hlnv8ww87apk28rq4pvcndnczvnsldjrjd4u2g6n9qekc6zpSeen on
wss://relay.momostr.pinkPublished at
2026-01-02 09:09:08 UTCEvent JSON
{
"id": "0c7687fb419e8591990777a528626eb56f451bd96f2acf668b853470e69b689f",
"pubkey": "a99ab49828156ff9b0ee71fdd0d947182a16626d9e04c9c3ed90e4daf148d4ca",
"created_at": 1767344948,
"kind": 1,
"tags": [
[
"e",
"704741eddfd914e5d987f3004e7515960b476615338b001c615066ad4521c0f4",
"",
"root",
"9a6a1a8eefd0b53d7e0c966ab35bd904151246f03b1be98fa0d2d0eeb4940320"
],
[
"p",
"f1dbe59d5195d61cb0a6f8b5908954f06b36003219b1f60748caaf18bd433890"
],
[
"e",
"f4ef5178016de76d992da87aa856b75208989b55fd2b40ffd485ec89c0e86d76",
"",
"reply",
"9a6a1a8eefd0b53d7e0c966ab35bd904151246f03b1be98fa0d2d0eeb4940320"
],
[
"p",
"9a6a1a8eefd0b53d7e0c966ab35bd904151246f03b1be98fa0d2d0eeb4940320"
],
[
"proxy",
"https://mathstodon.xyz/@solrize/115824718540236964",
"web"
],
[
"p",
"8b43e134ed66eccdb6ceb8acbeb01a61b39ad84bb6e0db19723e9cb1de69ee4b"
],
[
"proxy",
"https://mathstodon.xyz/users/solrize/statuses/115824718540236964",
"activitypub"
],
[
"L",
"pink.momostr"
],
[
"l",
"pink.momostr.activitypub:https://mathstodon.xyz/users/solrize/statuses/115824718540236964",
"pink.momostr"
],
[
"-"
]
],
"content": "Laderman's 1975 paper about the 3x3 case in 23 multiplications is nice: https://www.ams.org/journals/bull/1976-82-01/S0002-9904-1976-13988-2/S0002-9904-1976-13988-2.pdf\n\nIt explains how he reached it by solving some huge algebraic system by hand. A lot of fancy algebra went into the later improvements. I haven't looked at Bläser's result but it sounds like 19 is a lower bound rather than an upper one? Anyway https://en.wikipedia.org/wiki/Computational_complexity_of_matrix_multiplication goes into this, and it's informative.",
"sig": "a66cffdeba6c4ad44dee416da54b769fb3b2a697796d453941bea4cb1c9d8a51754f7668ade31c072f0aedc2353e088271c09ca28819c571e2c7d7403f443821"
}