Mojo Academy · 课时

逐元素 SIMD 运算

一次性对完整向量进行加法和乘法

逐元素 SIMD 运算 是 CoddyKit 上的免费 Mojo Academy 课时。这是第 2 节课，共 4 节。你可以在下方免费阅读本课时的完整内容 — 然后在浏览器中使用内置代码编辑器和全天候 AI 导师进行实践。这是 Mojo Academy 学习路径的一部分，你的进度在网页和 CoddyKit 应用中同步。 Mojo Academy 课程共包含 4 节课。

本课时的部分内容尚未翻译，以英文显示。

Math on Whole Vectors

The best part of SIMD is that arithmetic works on the entire vector at once. One operation touches every lane together.

Adding Two Vectors

Add two SIMD values and each lane is added to its matching lane. This is called element-wise addition.

var a = SIMD[DType.int32, 4](1, 2, 3, 4)
var b = SIMD[DType.int32, 4](10, 20, 30, 40)
var c = a + b

Lane Lines Up with Lane

In a + b, lane 0 adds to lane 0 and lane 1 adds to lane 1. The lanes never mix, they stay aligned.

print(c[0])

Multiplying Vectors

Multiplication is element-wise too. Each lane of one vector is multiplied by the same lane of the other vector.

var p = a * b

Subtract and Divide

Subtraction and division follow the same rule. Every lane is handled independently and in parallel across the pack.

var d = b - a

Scalar Meets Vector

Add a single number to a vector and Mojo applies it to every lane. The scalar spreads across all positions at once.

var bumped = a + 100

Widths Must Match

To combine two SIMD values they need the same DType and the same width. Mismatched lane counts will not line up.

Comparisons Go Lane by Lane

Comparing two vectors gives a SIMD of Bools, one True or False per lane. Each lane reports its own result.

var mask = a > SIMD[DType.int32, 4](2)

Reducing to One Number

Sometimes you want a single answer. reduce_add sums all lanes into one scalar, perfect for totals.

var total = a.reduce_add()
print(total)

Built-in Math Functions

Standard math like sqrt also works lane-wise on SIMD, computing the result for every element in one call.

from math import sqrt
var roots = sqrt(a.cast[DType.float32]())

Same Code, Many Results

Writing a + b looks like scalar math but produces many results at once. That readability with speed is Mojo's promise.

Quick Check

You compute a + b where both are SIMD[DType.int32, 4]. How is the addition done?

Recap

Element-wise SIMD math runs +, -, *, and / on every lane at once, scalars spread to all lanes, and reduce_add folds a vector into one number. ✨

常见问题解答

「逐元素 SIMD 运算」课时是免费的吗？

是的 — 「逐元素 SIMD 运算」的完整文本可在网页上免费阅读。要进行交互式练习（内置代码编辑器和全天候 AI 导师）并解锁 Mojo Academy 课程的其余内容，请升级到 CoddyKit PRO。 Mojo Academy 课程共包含 4 节课。

「逐元素 SIMD 运算」这节课中我会学到什么？

一次性对完整向量进行加法和乘法你通过在浏览器中直接运行的动手代码来练习 Mojo Academy，全天候 AI 导师会在你学习这节课的过程中回答你的问题。

学习 Mojo Academy 需要有经验吗？

无需任何先前经验。CoddyKit 上的 Mojo Academy 课程适合初学者到高级学习者，你可以从这里开始或从头开始，按照自己的节奏学习。这是第 2 节课，共 4 节。

「逐元素 SIMD 运算」课时需要多长时间？

大多数 CoddyKit 课程大约需要 5–10 分钟。每节课都很精短且互动，所以你能稳步进步，并在网页和应用中从离开的地方继续。

我能在这节 Mojo Academy 课中编写并运行代码吗？

能。每节 Mojo Academy 课都包含内置代码编辑器，你可以在浏览器中直接编写并运行真实代码，并获得即时 AI 反馈 — 无需本地设置。

此课程中的所有课时

← 返回 Mojo Academy