{"id":10255,"date":"2023-12-15T17:29:38","date_gmt":"2023-12-15T08:29:38","guid":{"rendered":"https:\/\/blogs.mathworks.com\/japan-community\/?p=10255"},"modified":"2023-12-15T18:04:36","modified_gmt":"2023-12-15T09:04:36","slug":"life-in-the-fast-lane-making-matlab-even-faster-on-apple-silicon-with-apple-accelerate-jp","status":"publish","type":"post","link":"https:\/\/blogs.mathworks.com\/japan-community\/2023\/12\/15\/life-in-the-fast-lane-making-matlab-even-faster-on-apple-silicon-with-apple-accelerate-jp\/","title":{"rendered":"Apple Accelerate \u3092\u4f7f\u7528\u3057\u3066 Apple Silicon \u4e0a\u306e MATLAB \u3092\u3055\u3089\u306b\u9ad8\u901f\u5316"},"content":{"rendered":"

\u203b\u3053\u306e\u6295\u7a3f\u306f 2023 \u5e74 12 \u6708 13 \u65e5\u306b The MATLAB Blog \u3078 \u6295\u7a3f<\/a>\u3055\u308c\u305f\u3082\u306e\u306e\u6284\u8a33\u3067\u3059\u3002<\/p>\n

\n
\n
<\/div>\n
\u884c\u5217\u540c\u58eb\u306e\u4e57\u7b97\uff1a3.7 \u500d\u3001LU \u5206\u89e3\uff1a2 \u500d\u3001\u30b3\u30ec\u30b9\u30ad\u30fc\u5206\u89e3\uff1a1.7 \u500d\u3001\u305d\u3057\u3066\u7dda\u5f62\u4ee3\u6570\u304c\u7d61\u3080\u305d\u306e\u4ed6\u306e\u5834\u9762\u3067\u306e\u3055\u3089\u306a\u308b\u9ad8\u901f\u5316\u306e\u53ef\u80fd\u6027\u3002\u3053\u308c\u3089\u3059\u3079\u3066\u304c\u3001\u4eca\u3059\u3050 Apple Silicon \u30de\u30b7\u30f3\u3067\u5b9f\u73fe\u3067\u304d\u308b\u3068\u3057\u305f\u3089\u8208\u5473\u304c\u3042\u308a\u307e\u3059\u304b\uff1f\u7d9a\u304d\u3092\u304a\u8aad\u307f\u304f\u3060\u3055\u3044\u3002<\/div>\n

\u3059\u3079\u3066\u306f BLAS \u304b\u3089<\/h2>\n
<\/div>\n
\u3082\u3057\u304b\u3057\u305f\u3089\u805e\u3044\u305f\u3053\u3068\u306e\u306a\u3044\u540d\u524d\u304b\u3082\u3057\u308c\u307e\u305b\u3093\u304c\u3001BLAS<\/span> (Basic Linear Algebra <\/span>Subprograms)<\/span><\/a> \u306f\u6700\u3082\u91cd\u8981\u306a\u30e9\u30a4\u30d6\u30e9\u30ea\u3067\u3059\u30022 \u3064\u306e\u884c\u5217\u3092\u639b\u3051\u5408\u308f\u305b\u305f\u308a\u3001\u30d9\u30af\u30c8\u30eb\u306b\u30b9\u30ab\u30e9\u30fc\u3092\u639b\u3051\u308b\u3068\u3044\u3063\u305f\u51e6\u7406\u3092\u3057\u307e\u3059\u3002\u3068\u3066\u3082\u5358\u7d14\u306a\u64cd\u4f5c\u306a\u306e\u3067\u3001MATLAB \u306e\u3088\u3046\u306a\u30bd\u30d5\u30c8\u30a6\u30a7\u30a2\u304c\u305d\u308c\u3089\u3092\u5b9f\u884c\u3059\u308b\u305f\u3081\u306b\u30e9\u30a4\u30d6\u30e9\u30ea\u3092\u4f7f\u7528\u3059\u308b\u3068\u306f\u601d\u308f\u306a\u3044\u304b\u3082\u3057\u308c\u307e\u305b\u3093\u3002<\/div>\n
MATLAB \u304c BLAS \u3092\u4f7f\u7528\u3059\u308b\u7406\u7531\u306f\u5b9f\u884c\u901f\u5ea6\u3067\u3059\uff01\u4e0a\u3067\u6319\u3052\u305f\u3088\u3046\u306a\u884c\u5217\u6f14\u7b97\u3092\u5b9f\u88c5\u3059\u308b\u3053\u3068\u306f\u3001\u901f\u5ea6\u3092\u6c17\u306b\u3057\u306a\u3051\u308c\u3070\u6bd4\u8f03\u7684\u7c21\u5358\u3067\u3059\u304c\u3001\u73fe\u4ee3\u306e\u30cf\u30fc\u30c9\u30a6\u30a7\u30a2\u3092\u6700\u5927\u9650\u306b\u6d3b\u7528\u3059\u308b\u3088\u3046\u306a\u65b9\u6cd5\u3067\u5b9f\u88c5\u3059\u308b\u3053\u3068\u306f\u6c7a\u3057\u3066\u7c21\u5358\u3067\u306f\u3042\u308a\u307e\u305b\u3093\u3002\u3056\u3063\u304f\u308a\u66f8\u304b\u308c\u305f\u884c\u5217\u540c\u58eb\u306e\u4e57\u7b97\u30eb\u30fc\u30c1\u30f3\u3068\u6700\u9069\u5316\u3055\u308c\u305f\u3082\u306e\u3067\u306f\u3001\u901f\u5ea6\u5dee\u306b\u4f55\u5343\u500d\u3082\u306e\u9055\u3044\u304c\u3067\u308b\u3053\u3068\u3082\u3042\u308a\u307e\u3059\u3002<\/div>\n
\n

\u7dda\u5f62\u4ee3\u6570\u306f\u73fe\u4ee3\u306e\u6280\u8853\u8a08\u7b97\u306e\u4e2d\u5fc3\u306b\u4f4d\u7f6e\u3057\u3066\u3044\u308b\u305f\u3081\u3001Intel\u3001AMD\u3001Apple \u306a\u3069\u306e\u30c1\u30c3\u30d7\u30e1\u30fc\u30ab\u30fc\u306f\u3001\u30cf\u30fc\u30c9\u30a6\u30a7\u30a2\u4e0a\u3067\u7dda\u5f62\u4ee3\u6570\u3092\u884c\u3046\u969b\u306b\u3067\u304d\u308b\u9650\u308a\u9ad8\u901f\u306b\u51e6\u7406\u3057\u305f\u3044\u3068\u8003\u3048\u3066\u3044\u307e\u3059<\/span>\u3002\u306a\u306e\u3067\u305d\u308c\u305e\u308c\u306eCPU \u306e\u7279\u6027\u3092\u5b8c\u74a7\u306b\u5f15\u304d\u51fa\u3059\u3088\u3046\u306b\u7cbe\u5de7\u306b\u30c1\u30e5\u30fc\u30cb\u30f3\u30b0\u3055\u308c\u305f BLAS \u30e9\u30a4\u30d6\u30e9\u30ea\u306e\u72ec\u81ea\u5b9f\u88c5\u3092\u6301\u3063\u3066\u3044\u307e\u3059\u3002<\/p>\n

\u3082\u3061\u308d\u3093\u3001\u30d6\u30ed\u30b0\u306e\u5e38\u9023\u8aad\u8005\u306b\u3068\u3063\u3066\u306f\u3053\u308c\u3089\u306f\u4f55\u3082\u65b0\u3057\u3044\u3053\u3068\u3067\u306f\u3042\u308a\u307e\u305b\u3093\u3002\u79c1\u306f BLAS\u3001\u305d\u3057\u3066\u305d\u306e\u5144\u5f1f\u3067\u3042\u308b LAPACK \u306b\u5c11\u3057\u57f7\u7740\u5fc3\u3092\u3082\u3063\u3066\u304a\u308a\u3001Trying out AMD\u2019s AOCL<\/a> \u3084 Apple Silicon \u306e MATLAB \u30d9\u30fc\u30bf\u7248\u306b\u95a2\u3059\u308b\u6295\u7a3f\uff08\u3053\u308c<\/a> \u3068 \u3053\u308c<\/a>\uff09\u306a\u3069\u306e\u8a18\u4e8b\u3067\u4f55\u5ea6\u3082\u53d6\u308a\u4e0a\u3052\u3089\u308c\u3066\u304d\u307e\u3057\u305f\u3002\u3053\u308c\u3089\u306e\u8a18\u4e8b\u3067\u306f\u3001\u7279\u306b Apple Silicon \u306b\u5bfe\u3059\u308b MATLAB \u306e BLAS \u30b5\u30dd\u30fc\u30c8\u304c\u6700\u8fd1\u3069\u306e\u3088\u3046\u306b\u9032\u5316\u3057\u3066\u304d\u305f\u304b\u3092\u89e3\u8aac\u3057\u307e\u3057\u305f\u3002<\/p>\n<\/div>\n

Apple Silicon \u306e MATLAB \u3067 Apple Accelerate \u3092 BLAS \u3068\u3057\u3066\u4f7f\u7528\u3067\u304d\u308b\u3088\u3046\u306b\u306a\u308a\u307e\u3057\u305f<\/h2>\n
Apple Silicon \u5411\u3051\u306e\u6700\u521d\u306e\u6b63\u5f0f\u30ea\u30ea\u30fc\u30b9\u3067\u3042\u308b MATLAB R2023b \u306f OpenBLAS \u3092\u4f7f\u7528\u3057\u3066\u304a\u308a\u30016 \u6708\u306b\u5831\u544a<\/a>\u3057\u305f\u3088\u3046\u306b\u3001\u30d1\u30d5\u30a9\u30fc\u30de\u30f3\u30b9\u306f\u304b\u306a\u308a\u826f\u3044\u3067\u3059\u3002\u305f\u3060\u30e6\u30fc\u30b6\u30fc\u304b\u3089\u300c\u306a\u3093\u3067 Apple \u306e BLAS \u5b9f\u88c5\u304c\u542b\u307e\u308c\u3066\u3044\u308b Apple Accelerate \u3092\u4f7f\u7528\u3057\u306a\u3044\u306e\uff1f\u300d\u3068\u306e\u30b3\u30e1\u30f3\u30c8\u3092\u9802\u3044\u3066\u3044\u307e\u3057\u305f\u3002\u305d\u308c\u306a\u308a\u306e\u7406\u7531\u304c\u3042\u308a\u307e\u3057\u305f\uff08\u3053\u306e\u8a18\u4e8b\u306e\u8a72\u5f53\u30bb\u30af\u30b7\u30e7\u30f3\u3092\u53c2\u7167\uff09\u304c\u3001\u72b6\u6cc1\u306f\u5909\u308f\u308a\u5229\u7528\u3067\u304d\u308b\u30a2\u30c3\u30d7\u30c7\u30fc\u30c8\u7248\u3092\u30ea\u30ea\u30fc\u30b9\u3057\u307e\u3057\u305f\u3002\u3057\u304b\u3057\u3001\u3001\u307b\u3068\u3093\u3069\u306e\u4eba\u304c\u898b\u9003\u3057\u3066\u3044\u308b\u304b\u3082\u30fb\u30fb\u3068\u3044\u3046\u3053\u3068\u3067\u3053\u3053\u3067\u6539\u3081\u3066\u304a\u77e5\u3089\u305b\u3057\u307e\u3059\uff01<\/div>\n
\n

\u307e\u305a R2023b Update 4 \u4ee5\u964d\u304c\u5fc5\u8981\u3067\u3059\u3002\u305d\u3082\u305d\u3082 R2023b \u3088\u308a\u524d\u306e\u3082\u306e\u306f Apple Silicon \u4e0a\u3067\u30cd\u30a4\u30c6\u30a3\u30d6\u5bfe\u5fdc\u3057\u3066\u307e\u305b\u3093\u3002R2023b \u3092\u30a4\u30f3\u30b9\u30c8\u30fc\u30eb\u3057\u307e\u3057\u305f\u3089\u3001version<\/span> \u30b3\u30de\u30f3\u30c9\u3067\u3069\u306e Update \u7248\u304b\u3092\u78ba\u8a8d\u3057\u3066\u304f\u3060\u3055\u3044\u3002<\/p>\n<\/div>\n

\n
version <\/span><\/div>\n
ans = <\/span><\/div>\n
‘23.2.0.2428915 (R2023b) Update 4’ <\/span><\/span><\/div>\n<\/div>\n
Update 3 \u4ee5\u524d\u306e\u5834\u5408\u306f Help->Check for Updates<\/span> \u304b\u3089\u30a2\u30c3\u30d7\u30c7\u30fc\u30c8\u3057\u3066\u304f\u3060\u3055\u3044\u3002<\/div>\n
\"\"<\/div>\n
\u305d\u308c\u304c\u5b8c\u4e86\u3057\u305f\u3089\u3001OpenBLAS \u304b\u3089 Apple Accelerate \u3078\u306e\u5207\u308a\u66ff\u3048\u65b9\u6cd5\u306f\u3001\u300cHow can I use the BLAS implementations included in Apple Accelerate Framework with MATLAB R2023b Update 4<\/span><\/a>\u300d\u3092\u53c2\u7167\u3057\u3066\u304f\u3060\u3055\u3044\u3002<\/div>\n

\u30d1\u30d5\u30a9\u30fc\u30de\u30f3\u30b9\u306e\u8a71<\/h2>\n
\n

\u8a00\u3046\u307e\u3067\u3082\u306a\u304f\u3001\u7dda\u5f62\u4ee3\u6570\u6f14\u7b97\u306b\u6ce8\u76ee\u3057\u306a\u3051\u308c\u3070\u30fb\u30fb\u3053\u306e\u5909\u66f4\u306b\u3088\u3063\u3066\u30d1\u30d5\u30a9\u30fc\u30de\u30f3\u30b9\u306e\u9055\u3044\u306f\u898b\u3089\u308c\u306a\u3044\u304b\u3082\u3057\u308c\u307e\u305b\u3093\u3002\u3057\u304b\u3057\u3001\u3053\u308c\u306f MATLAB \u306e\u8a71\u306a\u306e\u3067\u3001\u304d\u3063\u3068\u591a\u304f\u306e\u7dda\u5f62\u4ee3\u6570\u6f14\u7b97\u3092\u542b\u3080\u95a2\u6570\u3067\u6069\u6075\u304c\u3042\u308b\u3067\u3057\u3087\u3046\u3002<\/span><\/p>\n

\u3053\u306e\u30da\u30fc\u30b8\u306e\u30d9\u30f3\u30c1\u30de\u30fc\u30af\u3092\u81ea\u5206\u3067\u8a66\u3055\u308c\u305f\u3044\u5834\u5408\u3001\u79c1\u304c\u4f7f\u7528\u3057\u305f\u30b9\u30af\u30ea\u30d7\u30c8\u306f GitHub<\/a> \u306b\u3042\u308a\u307e\u3059\u3002<\/p>\n<\/div>\n

<\/h3>\n

\u884c\u5217\u540c\u58eb\u306e\u4e57\u7b97<\/span><\/h3>\n
\n

\u884c\u5217\u540c\u58eb\u306e\u4e57\u7b97\u3067\u6700\u3082\u5927\u304d\u3044\u901f\u5ea6\u306e\u5411\u4e0a\u304c\u307f\u3089\u308c\u307e\u3057\u305f\u3002\u3053\u308c\u306f Apple \u306b\u3088\u3063\u3066 BLAS \u304c\u304b\u306a\u308a\u6700\u9069\u5316\u3055\u308c\u3066\u3044\u308b\u3053\u3068\u3092\u793a\u3057\u307e\u3059\u3002<\/p>\n<\/div>\n\n\n\n\n\n\n
\n
Matrix Size<\/span><\/div>\n<\/td>\n
\n
OpenBLAS<\/span> time (s)<\/span><\/div>\n<\/td>\n
\n
Apple Accelerate<\/span> time (s)<\/span><\/div>\n<\/td>\n
\n
x Speed-up<\/span><\/div>\n<\/td>\n<\/tr>\n
\n
1,000<\/div>\n<\/td>\n
\n
0.<\/span>0172<\/span><\/div>\n<\/td>\n
\n
0.00<\/span>46<\/span><\/div>\n<\/td>\n
\n
3.<\/span>7<\/span>4<\/span><\/div>\n<\/td>\n<\/tr>\n
\n
5,000<\/div>\n<\/td>\n
\n
1.15<\/span>83<\/span><\/div>\n<\/td>\n
\n
0.4<\/span>171<\/span><\/div>\n<\/td>\n
\n
2.78<\/span><\/div>\n<\/td>\n<\/tr>\n
\n
10,000<\/div>\n<\/td>\n
\n
6.<\/span>8977<\/span><\/div>\n<\/td>\n
\n
3.3186<\/span><\/div>\n<\/td>\n
\n
2.08<\/span><\/div>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n

<\/h3>\n

LU \u5206\u89e3<\/span><\/h3>\n
\n

LU \u5206\u89e3\u306f LAPACK \u306e\u51e6\u7406\u3067\u3042\u308a\u3001\u4eca\u56de\u306f LAPACK \u30e9\u30a4\u30d6\u30e9\u30ea\u3092\u5909\u66f4\u3057\u3066\u3044\u307e\u305b\u3093\u3002\u3057\u304b\u3057\u3001LAPACK \u306f BLAS \u3092\u5229\u7528\u3059\u308b\u306e\u3067\u3001BLAS \u3092\u9ad8\u901f\u5316\u3059\u308c\u3070\u3001LAPACK \u3082\u5f53\u7136\u9ad8\u901f\u306b\u306a\u308a\u307e\u3059\u3002<\/p>\n<\/div>\n\n\n\n\n\n\n
\n
Matrix Size<\/span><\/div>\n<\/td>\n
\n
OpenBLAS<\/span> time (s)<\/span><\/div>\n<\/td>\n
\n
Apple Accelerate time (s)<\/span><\/div>\n<\/td>\n
\n
x Speed-up<\/span><\/div>\n<\/td>\n<\/tr>\n
\n
1,000<\/div>\n<\/td>\n
\n
0.0124<\/span><\/div>\n<\/td>\n
\n
0.0115<\/span><\/div>\n<\/td>\n
\n
1.08<\/span><\/div>\n<\/td>\n<\/tr>\n
\n
5,000<\/div>\n<\/td>\n
\n
0.4345<\/span><\/div>\n<\/td>\n
\n
0.2556<\/span><\/div>\n<\/td>\n
\n
1.7<\/span><\/div>\n<\/td>\n<\/tr>\n
\n
10,000<\/div>\n<\/td>\n
\n
3.5928<\/span><\/div>\n<\/td>\n
\n
1.6821<\/span><\/div>\n<\/td>\n
\n
2.14<\/span><\/div>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n
Cholesky \u5206\u89e3\u3082\u8a66\u3057\u307e\u3057\u305f\u304c\u30011.28 \u500d\u304b\u3089 1.77 \u500d\u306e\u901f\u5ea6\u30a2\u30c3\u30d7\u304c\u307f\u3089\u308c\u307e\u3057\u305f\u3002<\/div>\n
<\/div>\n

\u56fa\u6709\u5024<\/span><\/h3>\n
\u56fa\u6709\u5024\u8a08\u7b97\u306e\u30c6\u30b9\u30c8\u306b\u306f\u3001eig<\/span> \u306e\u5358\u4e00\u51fa\u529b\u30d0\u30fc\u30b8\u30e7\u30f3\u3092\u4f7f\u7528\u3057\u307e\u3057\u305f\u3002e = eig(A) <\/span>\u3067\u3059\u3002\u4ed6\u306e\u30c6\u30b9\u30c8\u3068\u540c\u69d8\u306b\u3001\u30e9\u30f3\u30c0\u30e0\u306a\u884c\u5217\u3092\u8a66\u3057\u307e\u3057\u305f\u3002<\/span>\u5c0f\u3055\u3044\u884c\u5217\u3067\u306f\u3046\u307e\u304f\u3044\u304d\u3001\u901f\u5ea6\u5411\u4e0a\u304c\u898b\u3089\u308c\u307e\u3057\u305f\u304c\u3001\u5927\u304d\u306a\u884c\u5217\u3067\u306f\u3001OpenBLAS \u306e\u30aa\u30ea\u30b8\u30ca\u30eb\u3068\u6bd4\u8f03\u3057\u3066\u3001\u79c1\u306e M2 \u30de\u30b7\u30f3\u3067\u306f\u304b\u306a\u308a\u306e\u9045\u5ef6\u304c\u898b\u3089\u308c\u307e\u3057\u305f\u3002<\/div>\n
\n

\u3053\u306e\u3053\u3068\u306b\u3064\u3044\u3066 M1 \u3092\u6301\u3063\u3066\u3044\u308b\u540c\u50da\u3068\u8b70\u8ad6\u3057\u307e\u3057\u305f\u304c\u3001\u5f7c\u306e\u74b0\u5883\u3067\u306f\u9045\u5ef6\u306f\u898b\u3089\u308c\u305a\u3001eig \u306b Apple Accelerate \u3092\u4f7f\u7528\u3059\u308b\u65b9\u304c\u5e38\u306b\u901f\u3044\u3002\u3053\u308c\u306f M2 \u56fa\u6709\u306e\u554f\u984c\u3067\u3042\u308b\u3068\u8003\u3048\u3089\u308c\u308b\u306e\u3067\u3001\u4f55\u304c\u8d77\u3053\u3063\u3066\u3044\u308b\u306e\u304b\u3092\u3088\u308a\u8a73\u3057\u304f\u8abf\u67fb\u3059\u308b\u4e88\u5b9a\u3067\u3059\u3002<\/p>\n<\/div>\n\n\n\n\n\n\n
\n
Matrix Size<\/span><\/div>\n<\/td>\n
\n
OpenBLAS<\/span> time (s)<\/span><\/div>\n<\/td>\n
\n
Apple Accelerate time (s)<\/span><\/div>\n<\/td>\n
\n
x Speed-up<\/span><\/div>\n<\/td>\n<\/tr>\n
\n
1,000<\/div>\n<\/td>\n
\n
0.4245<\/span><\/div>\n<\/td>\n
\n
0.24<\/span>07<\/span><\/div>\n<\/td>\n
\n
1<\/span>.<\/span>76<\/span><\/div>\n<\/td>\n<\/tr>\n
\n
5,000<\/div>\n<\/td>\n
\n
22<\/span>.86<\/span>54<\/span><\/div>\n<\/td>\n
\n
24.74<\/span><\/div>\n<\/td>\n
\n
0.92<\/span><\/div>\n<\/td>\n<\/tr>\n
\n
10,000<\/div>\n<\/td>\n
\n
145.4<\/span>076<\/span><\/div>\n<\/td>\n
\n
201.57<\/span><\/div>\n<\/td>\n
\n
0.72<\/span><\/div>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n
<\/div>\n

\u30c6\u30b9\u30c8\u74b0\u5883\u306e\u8a73\u7d30<\/h2>\n
\n

\u4eca\u56de\u96fb\u6e90\u306b\u63a5\u7d9a\u3055\u308c\u305f\u72b6\u614b\u306e M2 \u30d9\u30fc\u30b9\u306e MacBook Pro \u3092\u4f7f\u7528\u3057\u307e\u3057\u305f\u3002File Exchange\u3067\u5165\u624b\u53ef\u80fd\u306a cpuinfo \u30b3\u30de\u30f3\u30c9<\/a>\u306e\u51fa\u529b\u306f\u4ee5\u4e0b\u306e\u901a\u308a\u3067\u3059\u3002<\/p>\n<\/div>\n

\n
cpuinfo <\/span><\/div>\n
ans = <\/span><\/div>\n
struct\u202f<\/span>with fields: <\/span><\/div>\n
<\/div>\n
CPUName: ‘Apple M2 Pro’ <\/span><\/span><\/div>\n
Clock: ‘N\/A’ <\/span><\/span><\/div>\n
Cache: 65536 <\/span><\/div>\n
TotalMemory: 1.7180e+10 <\/span><\/div>\n
NumCPUs: 1 <\/span><\/div>\n
TotalCores: 10 <\/span><\/div>\n
OSType: ‘macOS’ <\/span><\/span><\/div>\n
OSVersion: ‘13.3.1’ <\/span><\/span><\/div>\n<\/div>\n
eig \u306b\u95a2\u3059\u308b\u554f\u984c\u306f\u3044\u305a\u308c\u89e3\u6c7a\u3055\u308c\u308b\u3053\u3068\u3092\u9858\u3063\u3066\u3044\u307e\u3059\u304c\u3001Apple Accelerate \u306b\u5207\u308a\u66ff\u3048\u308b\u3053\u3068\u306f\u975e\u5e38\u306b\u6709\u76ca\u306a\u3088\u3046\u3067\u3059\u3002Apple Silicon Mac \u3092\u304a\u6301\u3061\u306e\u5834\u5408\u306f\u3001\u8a66\u3057\u3066\u307f\u3066\u3001\u3069\u306e\u3088\u3046\u306a\u7d50\u679c\u304c\u51fa\u305f\u304b\u6559\u3048\u3066\u304f\u3060\u3055\u3044\u3002<\/div>\n<\/div>\n