|
|
89f713b951
|
移动 req.txt(本地最新版)到 BMM/ 目录
|
2026-08-27 11:22:11 +00:00 |
|
|
|
93411c15b7
|
v1.4: 同步HTML
|
2026-08-27 09:40:15 +00:00 |
|
|
|
93f8730f89
|
v1.4: 修正方形结论适用范围——最少切分原则+B≥C不切分跟随M/N+B<C才方形
|
2026-08-27 09:40:14 +00:00 |
|
|
|
ecf4dfd74a
|
v1.4: 同步HTML
|
2026-08-27 09:11:13 +00:00 |
|
|
|
bc246cd9b4
|
v1.4: 修正baseK理由(Cube粒度非搬移效率)+SingleCoreM/N方形建模推导(拉格朗日+数值验证)
|
2026-08-27 09:11:12 +00:00 |
|
|
|
5a017ade98
|
v1.3: 同步HTML
|
2026-08-27 08:47:39 +00:00 |
|
|
|
1e92d348df
|
v1.3: BaseM/N方形优先纠正长宽比跟随说法+SingleCoreM/N跟随M/N的独立理由
|
2026-08-27 08:47:38 +00:00 |
|
|
|
6690a79d9f
|
v1.3: 同步HTML
|
2026-08-27 08:28:33 +00:00 |
|
|
|
792a723bbf
|
v1.3: 补SingleCoreM/N对L2重复读的定量影响+BaseM/N长宽比跟随的三条原因
|
2026-08-27 08:28:32 +00:00 |
|
|
|
c9a0d8cf21
|
v1.2: 同步HTML
|
2026-08-27 07:57:45 +00:00 |
|
|
|
086c2c2785
|
v1.2: 结构重整——尾轮融入Step2+dValue分Bound两级+编号统一
|
2026-08-27 07:57:44 +00:00 |
|
|
|
aa778ecd0c
|
v1.2: 同步HTML
|
2026-08-27 07:49:54 +00:00 |
|
|
|
fcf6b58499
|
v1.2: 尾轮重切分访存/计算Bound两级约束+dValue放宽定量依据
|
2026-08-27 07:49:52 +00:00 |
|
|
|
ec65ac0d43
|
v1.2: 同步HTML
|
2026-08-27 07:07:25 +00:00 |
|
|
|
68dbdcd3d4
|
v1.2: 修正dValue定义(ND连续维非乘积)+重切约束推导+自包含搬移效率解释
|
2026-08-27 07:07:17 +00:00 |
|
|
|
0182224a18
|
v1.2: 同步HTML
|
2026-08-27 06:53:55 +00:00 |
|
|
|
9583588f8a
|
v1.2: 重切约束补推导链——两级阈值(min_TileSize/256B)+对齐约束+退化条件
|
2026-08-27 06:53:54 +00:00 |
|
|
|
d443f7d7e5
|
v1.2: 同步HTML
|
2026-08-27 06:38:29 +00:00 |
|
|
|
1ae5805236
|
v1.2: 尾轮重切分析(host零代价)+BaseM/N和SingleCoreM/N枚举确定过程
|
2026-08-27 06:38:27 +00:00 |
|
|
|
c5f9fe299b
|
v1.1: 同步HTML
|
2026-08-27 03:32:32 +00:00 |
|
|
|
2fd7751c05
|
v1.1: Step0补L0C双缓冲vs UnitFlag单缓冲分析+§7.2补dbL0C/unitFlag源码事实
|
2026-08-27 03:32:31 +00:00 |
|
|
|
aee5eb01fb
|
v1.1: 同步HTML
|
2026-08-27 03:09:07 +00:00 |
|
|
|
c7dc22d27b
|
v1.1: 新增§七源码对比——降核缺失+cubeBound模型+L2管理保守+经验常数
|
2026-08-27 03:09:05 +00:00 |
|
|
|
90ca685502
|
v1.0: 同步HTML
|
2026-08-27 02:13:53 +00:00 |
|
|
|
19a8e7d62a
|
v1.0: ASW_Basic独立文档——时延建模+分配策略+尾轮处理+实现方案
|
2026-08-27 02:13:52 +00:00 |
|
|
|
319787afdd
|
v0.98: 同步HTML
|
2026-08-26 15:47:41 +00:00 |
|
|
|
c98cea0181
|
v0.98: ASW第八章加时延建模+B分组vs线性映射四场景严格对比
|
2026-08-26 15:47:17 +00:00 |
|
|
|
a7444f9e93
|
v0.97: 同步HTML
|
2026-08-26 12:30:13 +00:00 |
|
|
|
e401a1ea48
|
v0.97: Step2核间分配逻辑详述——线性映射vs B分组对比+B不整除C处理
|
2026-08-26 12:30:11 +00:00 |
|
|
|
29a3e029b0
|
v0.96: 同步HTML
|
2026-08-26 12:19:29 +00:00 |
|
|
|
1326cd8ec2
|
v0.96: 恢复第八章ASW_Basic章节头+进入条件+StreamK实现方案
|
2026-08-26 12:19:27 +00:00 |
|
|
|
5b92289a54
|
v1.0: 同步HTML
|
2026-08-26 11:56:10 +00:00 |
|
|
|
6ab843077a
|
v1.0: 参考文献CANN 8.5→9.2.0-beta.1
|
2026-08-26 11:56:09 +00:00 |
|
|
|
c3c9e80353
|
v1.0: 同步HTML
|
2026-08-26 11:43:16 +00:00 |
|
|
|
9f2a1501fc
|
v1.0: §四重写——补符号表+执行模型+时延推导链
|
2026-08-26 11:43:15 +00:00 |
|
|
|
e36fc587c5
|
v1.0: 同步HTML
|
2026-08-26 11:34:51 +00:00 |
|
|
|
1b14a201e9
|
v1.0: 补摘要
|
2026-08-26 11:34:49 +00:00 |
|
|
|
ad6a92fdde
|
v1.0: 同步HTML
|
2026-08-26 11:30:26 +00:00 |
|
|
|
6f44201295
|
v1.0: 修正源码仓地址为gitcode.com/cann/ops-nn
|
2026-08-26 11:30:25 +00:00 |
|
|
|
71590ec422
|
v1.0: 同步HTML
|
2026-08-26 11:05:56 +00:00 |
|
|
|
6b03624de3
|
v1.0: MergeBatch vs IterBatch 专项对比分析
|
2026-08-26 11:05:56 +00:00 |
|
|
|
16b37ebc9e
|
v0.96: 同步HTML
|
2026-08-26 09:28:22 +00:00 |
|
|
|
667241719f
|
v0.96: MergeBatch实现方案重写——b0→kL0→kL1反推→bL1最大化
|
2026-08-26 09:28:21 +00:00 |
|
|
|
5ceaeafc97
|
v1.1: 同步HTML
|
2026-08-26 08:03:11 +00:00 |
|
|
|
994ce3648a
|
v1.1: ASW_Basic重点对比增强——BaseM/N+SingleCoreM/N+swizzle+L2管理
|
2026-08-26 08:03:10 +00:00 |
|
|
|
4ec7435ce3
|
v0.95: 同步HTML
|
2026-08-26 07:36:15 +00:00 |
|
|
|
ffdb7c6f7b
|
v0.95: 修正baseK无dValue约束,dValue约束在kL1(GM→L1)
|
2026-08-26 07:36:13 +00:00 |
|
|
|
2035a569fb
|
v0.95: 同步HTML
|
2026-08-26 07:09:38 +00:00 |
|
|
|
5d6e3197ce
|
v0.95: ASW Step0/1分层+MergeBatch补L1绑定情形分界
|
2026-08-26 07:09:35 +00:00 |
|
|
|
d5a46280df
|
v0.94: 同步HTML
|
2026-08-26 06:50:59 +00:00 |
|