ChillFish8

[0,29]    .   D============eeeER   .    .    .    .    .    .    .    .    .    .    .  .   vaddps      ymm0, ymm0, ymm1
[0,31]    .    D==============eeeER.    .    .    .    .    .    .    .    .    .    .  .   vaddps      ymm0, ymm0, ymm1
[0,33]    .    D=================eeeER  .    .    .    .    .    .    .    .    .    .  .   vaddps      ymm0, ymm0, ymm1
[0,35]    .    D====================eeeER    .    .    .    .    .    .    .    .    .  .   vaddps      ymm0, ymm0, ymm1
[0,38]    .    .D======================eeeER .    .    .    .    .    .    .    .    .  .   vaddps      ymm0, ymm0, ymm1
[0,41]    .    .D=========================eeeER   .    .    .    .    .    .    .    .  .   vaddps      ymm0, ymm0, ymm1
[0,44]    .    . D===========================eeeER.    .    .    .    .    .    .    .  .   vaddps      ymm0, ymm0, ymm1

ChillFish8

Iterations:        100
Instructions:      12600
Total Cycles:      10225
Total uOps:        22600

Dispatch Width:    6
uOps Per Cycle:    2.21
IPC:               1.23
Block RThroughput: 37.7

ChillFish8

.section .text,"xr",one_only,blogpost::avx2_dot::avx2_dot_product
        .globl  blogpost::avx2_dot::avx2_dot_product
        .p2align        4, 0x90
blogpost::avx2_dot::avx2_dot_product:
        .cv_func_id 6
.seh_proc _ZN8blogpost8avx2_dot16avx2_dot_product17hffb9005f074b96fbE
        sub rsp, 104
        .seh_stackalloc 104
        .seh_endprologue
        mov qword ptr [rsp + 40], rdx

ChillFish8

Iterations:        100
Instructions:      12600
Total Cycles:      10225
Total uOps:        22600

Dispatch Width:    6
uOps Per Cycle:    2.21
IPC:               1.23
Block RThroughput: 37.7

ChillFish8

Iterations:        100
Instructions:      12600
Total Cycles:      10225
Total uOps:        22600

Dispatch Width:    6
uOps Per Cycle:    2.21
IPC:               1.23
Block RThroughput: 37.7

ChillFish8

use core::arch::x86_64::*;

#[target_feature(enable = "avx2")]
unsafe fn avx2_dot_product(vector_a: &[f32], vector_b: &[f32]) -> f32 {
    assert_eq!(vector_a.len(), vector_b.len(), "Vectors must be equal in length");

    let dims = vector_a.len();
    
    // We want to ensure we're going in steps of 8 elements
    // and then handle the remainder separately.

ChillFish8

fn simple_dot_product(vector_a: &[f32], vector_b: &[f32]) -> f32 {
    assert_eq!(vector_a.len(), vector_b.len(), "Vectors must be equal in length");
    
    let dims = vector_a.len();
    let mut total = 0.0;

    // Quick disclaimer: I'm going to be using while loops here for consistency across
    // the code samples.
    let mut i = 0;
    while i < dims {

ChillFish8

    import {Jumper} from "svelte-loading-spinners";
    import {DISCORD_LOGIN_URI} from "$lib/api/config";
    import {onMount} from "svelte";
    import {ALLOWED_ORIGIN, flags} from "../../lib/api/config.js";
    import {extractUserData, requestUserAccess} from "../../lib/api/auth.js";
    import {accessToken, currentUser} from "../../lib/stores/auth.js";
    import {goto} from "$app/navigation";
    import Primary from "../../lib/compontents/buttons/Primary.svelte";
    import Secondary from "../../lib/compontents/buttons/Secondary.svelte";
    import {writable} from "svelte/store";

ChillFish8

use std::io;
use std::path::{Path, PathBuf};
use std::sync::Arc;
use std::time::{Duration, Instant};
use ahash::{HashMap, HashMapExt, HashSet};

use parking_lot::Mutex;
use tokio::task::JoinHandle;

use crate::backends::ReadBuffer;

ChillFish8

import asyncio
import os
import time
import cv2
import numpy
import pygame
import shutil
from threading import Thread
from queue import Queue