{"id":18174,"date":"2026-09-20T02:17:27","date_gmt":"2026-09-19T19:17:27","guid":{"rendered":"https:\/\/useful.vn\/?p=18174"},"modified":"2026-09-20T02:17:27","modified_gmt":"2026-09-19T19:17:27","slug":"apple-ra-chip-2-nm-cho-may-tinh","status":"publish","type":"post","link":"https:\/\/useful.vn\/?p=18174","title":{"rendered":"Apple ra chip 2 nm cho m\u00e1y t\u00ednh"},"content":{"rendered":"
\n

\nApple ra chip 2 nm cho m\u00e1y t\u00ednh <\/h1>\n

Apple gi\u1edbi thi\u1ec7u chip 2 nm \u0111\u1ea7u ti\u00ean c\u1ee7a h\u00e3ng l\u00e0 M6 c\u00f9ng phi\u00ean b\u1ea3n hi\u1ec7u n\u0103ng cao M5 Ultra d\u00e0nh cho d\u00f2ng Mac Studio m\u1edbi.<\/p>\n

Tr\u01b0\u1edbc th\u1ec1m s\u1ef1 ki\u1ec7n quan tr\u1ecdng v\u00e0o \u0111\u1ea7u th\u00e1ng 9, Apple g\u00e2y b\u1ea5t ng\u1edd khi gi\u1edbi thi\u1ec7u lo\u1ea1t ph\u1ea7n c\u1ee9ng m\u1edbi g\u1ed3m hai d\u00f2ng vi x\u1eed l\u00fd c\u0169ng nh\u01b0 phi\u00ean b\u1ea3n c\u1eadp nh\u1eadt cho Mac Studio v\u00e0 Mac mini.<\/p>\n

\n
\n<\/div>
\n

Chip M6 v\u00e0 M5 Ultra m\u1edbi c\u1ee7a Apple.<\/p>\n<\/figcaption><\/figure>\n

M6 – chip x\u1eed l\u00fd 2 nm \u0111\u1ea7u ti\u00ean c\u1ee7a Apple<\/strong><\/p>\n

L\u1ea7n \u0111\u1ea7u s\u1eed d\u1ee5ng ki\u1ebfn tr\u00fac 2 nm gi\u00fap Apple t\u0103ng m\u1eadt \u0111\u1ed9 b\u00f3ng b\u00e1n d\u1eabn v\u00e0 kh\u1ea3 n\u0103ng ti\u00eau th\u1ee5 n\u0103ng l\u01b0\u1ee3ng hi\u1ec7u qu\u1ea3 h\u01a1n so v\u1edbi c\u00e1c th\u1ebf h\u1ec7 tr\u01b0\u1edbc. Chip M6 s\u1edf h\u1eefu nhi\u1ec1u h\u01a1n hai l\u00f5i so v\u1edbi chip M5. \u0110\u00e1ng ch\u00fa \u00fd, \u0111\u00e2y l\u00e0 b\u1ed9 x\u1eed l\u00fd \u0111\u1ea7u ti\u00ean c\u1ee7a Apple c\u00f3 \u0111\u1ea7y \u0111\u1ee7 ba lo\u1ea1i l\u00f5i CPU g\u1ed3m hai l\u00f5i si\u00eau x\u1eed l\u00fd, b\u1ed1n l\u00f5i hi\u1ec7u n\u0103ng v\u00e0 s\u00e1u l\u00f5i ti\u1ebft ki\u1ec7m n\u0103ng l\u01b0\u1ee3ng.<\/p>\n

Trong \u0111\u00f3, l\u00f5i si\u00eau x\u1eed l\u00fd \u01b0u ti\u00ean hi\u1ec7u n\u0103ng \u0111\u01a1n lu\u1ed3ng v\u00e0 l\u00e0 lo\u1ea1i l\u00f5i m\u1ea1nh nh\u1ea5t c\u1ee7a Apple. C\u00e1c l\u00f5i hi\u1ec7u n\u0103ng \u0111\u01b0\u1ee3c t\u1ed1i \u01b0u cho t\u00e1c v\u1ee5 \u0111a lu\u1ed3ng \u0111\u00f2i h\u1ecfi cao, c\u00f2n c\u00e1c l\u00f5i ti\u1ebft ki\u1ec7m n\u0103ng l\u01b0\u1ee3ng x\u1eed l\u00fd c\u00e1c t\u00e1c v\u1ee5 n\u1ec1n. Apple nh\u1ea5n m\u1ea1nh M6 c\u00f3 “hi\u1ec7u n\u0103ng \u0111\u01a1n lu\u1ed3ng nhanh nh\u1ea5t th\u1ebf gi\u1edbi” v\u00e0 cung c\u1ea5p hi\u1ec7u n\u0103ng \u0111a lu\u1ed3ng nhanh h\u01a1n 1,2 l\u1ea7n so v\u1edbi M5.<\/p>\n

M\u1ed7i l\u00f5i tr\u00ean chip m\u1edbi \u0111\u1ec1u c\u00f3 b\u1ed9 t\u0103ng t\u1ed1c x\u1eed l\u00fd tr\u00ed tu\u1ec7 nh\u00e2n t\u1ea1o (Neural Accelerator), gi\u00fap n\u00e2ng 30% hi\u1ec7u n\u0103ng t\u00ednh to\u00e1n GPU t\u1ed1i \u0111a cho AI so v\u1edbi M5. GPU c\u1ee7a M6 c\u0169ng \u0111\u01b0\u1ee3c c\u1eadp nh\u1eadt ki\u1ebfn tr\u00fac l\u00f5i \u0111\u1ed5 b\u00f3ng, b\u1ed9 nh\u1edb \u0111\u1ec7m \u0111\u1ed9ng (Dynamic Caching) v\u00e0 kh\u1ea3 n\u0103ng d\u00f2 tia \u0111\u01b0\u1ee3c t\u0103ng t\u1ed1c b\u1edfi ph\u1ea7n c\u1ee9ng.<\/p>\n

M6 h\u1ed7 tr\u1ee3 b\u1ed9 nh\u1edb h\u1ee3p nh\u1ea5t 32 GB v\u1edbi b\u0103ng th\u00f4ng b\u1ed9 nh\u1edb 170 GB\/gi\u00e2y, nhanh h\u01a1n t\u1edbi 10% so v\u1edbi M5. Ngo\u00e0i ra, M6 c\u00f3 th\u1ec3 x\u1eed l\u00fd FP8 tr\u1ef1c ti\u1ebfp b\u1eb1ng ph\u1ea7n c\u1ee9ng, trong khi M5 x\u1eed l\u00fd FP8 th\u00f4ng qua ph\u1ea7n m\u1ec1m. FP8 l\u00e0 \u0111\u1ecbnh d\u1ea1ng 8 bit \u0111\u01b0\u1ee3c thi\u1ebft k\u1ebf \u0111\u1ec3 gi\u00fap c\u00e1c t\u00e1c v\u1ee5 AI nhanh h\u01a1n v\u00e0 ti\u1ebft ki\u1ec7m b\u1ed9 nh\u1edb h\u01a1n.<\/p>\n

Vi x\u1eed l\u00fd cao c\u1ea5p M5 Ultra<\/strong><\/p>\n

\n
\n<\/div>
\n

Mac Studio v\u1edbi chip M5 Ultra.<\/p>\n<\/figcaption><\/figure>\n

M5 Ultra l\u00e0 vi x\u1eed l\u00fd m\u1ea1nh nh\u1ea5t c\u1ee7a Apple hi\u1ec7n t\u1ea1i v\u00e0 ch\u1ec9 c\u00f3 tr\u00ean c\u00e1c d\u00f2ng m\u00e1y cho ng\u01b0\u1eddi d\u00f9ng chuy\u00ean nghi\u1ec7p nh\u01b0 Mac Studio. \u0110\u00e2y c\u0169ng l\u00e0 model \u0111\u00e1nh d\u1ea5u vi\u1ec7c h\u00e3ng s\u1eed d\u1ee5ng c\u00f4ng ngh\u1ec7 UltraFusion \u0111\u1ed9c quy\u1ec1n th\u1ebf h\u1ec7 m\u1edbi v\u1edbi kh\u1ea3 n\u0103ng k\u1ebft n\u1ed1i hai chip M5 Max l\u00f5i k\u00e9p, t\u1ea1o n\u00ean ki\u1ebfn tr\u00fac gh\u00e9p c\u1ee5m 4 chip \u0111\u1ea7u ti\u00ean c\u1ee7a Apple. C\u00f4ng ngh\u1ec7 n\u00e0y c\u0169ng mang l\u1ea1i b\u0103ng th\u00f4ng gi\u1eefa c\u00e1c l\u00f5i h\u01a1n 4,4 TB\/gi\u00e2y v\u00e0 m\u1eadt \u0111\u1ed9 k\u1ebft n\u1ed1i cao h\u01a1n g\u1ea5p 6 l\u1ea7n, cho ph\u00e9p c\u1ea3 b\u1ed1n l\u00f5i ho\u1ea1t \u0111\u1ed9ng nh\u01b0 m\u1ed9t b\u1ed9 x\u1eed l\u00fd th\u1ed1ng nh\u1ea5t duy nh\u1ea5t.<\/p>\n

Chip cao c\u1ea5p nh\u1ea5t c\u1ee7a Apple c\u00f3 nhi\u1ec1u t\u00f9y ch\u1ecdn v\u1ec1 CPU, GPU, trong \u0111\u00f3 CPU t\u1ed1i \u0111a 36 l\u00f5i v\u1edbi 12 l\u00f5i si\u00eau m\u1ea1nh v\u00e0 24 l\u00f5i hi\u1ec7u n\u0103ng cao. C\u00f4ng ty cho bi\u1ebft \u0111i\u1ec1u n\u00e0y gi\u00fap t\u0103ng hi\u1ec7u n\u0103ng \u0111\u01a1n lu\u1ed3ng 1,25 l\u1ea7n v\u00e0 hi\u1ec7u n\u0103ng \u0111a lu\u1ed3ng 1,3 l\u1ea7n so v\u1edbi M3 Ultra.<\/p>\n

V\u1ec1 kh\u1ea3 n\u0103ng \u0111\u1ed3 h\u1ecda, M5 Ultra cung c\u1ea5p GPU th\u1ebf h\u1ec7 ti\u1ebfp theo v\u1edbi t\u1ed1i \u0111a 80 l\u00f5i v\u00e0 m\u1ed7i l\u00f5i \u0111\u1ec1u c\u00f3 b\u1ed9 t\u0103ng t\u1ed1c x\u1eed l\u00fd tr\u00ed tu\u1ec7 nh\u00e2n t\u1ea1o (Neural Accelerator). Chip m\u1edbi cung c\u1ea5p hi\u1ec7u n\u0103ng x\u1eed l\u00fd \u0111\u1ed3 h\u1ecda cho tr\u00ed tu\u1ec7 nh\u00e2n t\u1ea1o (AI) cao h\u01a1n 4,5 l\u1ea7n so v\u1edbi M3 Ultra.<\/p>\n

Tu\u1ea5n H\u01b0ng<\/strong><\/p>\n<\/article>\n

\n

https:\/\/vnexpress.net\/apple-ra-chip-2-nm-cho-may-tinh-5113362.html<\/p>\n","protected":false},"excerpt":{"rendered":"

Apple ra chip 2 nm cho m\u00e1y t\u00ednh Apple gi\u1edbi thi\u1ec7u chip 2 nm \u0111\u1ea7u ti\u00ean c\u1ee7a h\u00e3ng l\u00e0 M6 c\u00f9ng phi\u00ean b\u1ea3n hi\u1ec7u n\u0103ng cao M5 Ultra d\u00e0nh cho d\u00f2ng Mac Studio m\u1edbi. Tr\u01b0\u1edbc th\u1ec1m s\u1ef1 ki\u1ec7n quan tr\u1ecdng v\u00e0o \u0111\u1ea7u th\u00e1ng 9, Apple g\u00e2y b\u1ea5t ng\u1edd khi gi\u1edbi thi\u1ec7u lo\u1ea1t ph\u1ea7n c\u1ee9ng m\u1edbi<\/p>\n","protected":false},"author":1,"featured_media":18175,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[3],"tags":[],"class_list":["post-18174","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-cong-nghe"],"_links":{"self":[{"href":"https:\/\/useful.vn\/index.php?rest_route=\/wp\/v2\/posts\/18174","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/useful.vn\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/useful.vn\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/useful.vn\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/useful.vn\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=18174"}],"version-history":[{"count":0,"href":"https:\/\/useful.vn\/index.php?rest_route=\/wp\/v2\/posts\/18174\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/useful.vn\/index.php?rest_route=\/wp\/v2\/media\/18175"}],"wp:attachment":[{"href":"https:\/\/useful.vn\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=18174"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/useful.vn\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=18174"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/useful.vn\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=18174"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}