Skip to content

Instantly share code, notes, and snippets.

@cbalint13
cbalint13 / old-vs-new-winograd-2_4.log
Created July 24, 2019 20:23
TVM_Winograd_OLD_vs_NEW_2_4_6
# Licensed to the Apache Software Foundation (ASF) under one
# or more contributor license agreements. See the NOTICE file
# distributed with this work for additional information
# regarding copyright ownership. The ASF licenses this file
# to you under the Apache License, Version 2.0 (the
# "License"); you may not use this file except in compliance
# with the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
// attr [get_valid_counts_phase_one.v0] storage_scope = "global"
allocate get_valid_counts_phase_one.v0[int32 * 15360]
// attr [get_valid_counts_phase_one.v1] storage_scope = "global"
allocate get_valid_counts_phase_one.v1[int32 * 15360]
// attr [get_valid_counts_phase_two.v1] storage_scope = "global"
allocate get_valid_counts_phase_two.v1[int32 * 496]
// attr [get_valid_counts_phase_two.v0] storage_scope = "global"
allocate get_valid_counts_phase_two.v0[int32 * 15360]
// attr [get_valid_counts_phase_three] storage_scope = "global"
allocate get_valid_counts_phase_three[int32 * 496]
@wiggin15
wiggin15 / a.c
Created July 29, 2018 12:08
vm_region_recurse_64 / proc_regionfilename
#include <errno.h>
#include <stdlib.h>
#include <stdio.h>
#include <libproc.h>
#include <mach/mach.h>
#include <CoreFoundation/CoreFoundation.h>
int main(void)
{

bcache

bcache 实测对于小文件随机写场景下效果好,而且减少了小写的延迟。

目前 bcache 这类方案主要的问题是过于复杂,难以实现和维护。

Intro

引入块层的缓存对于优化磁盘的读写性能来说效果显著。块级别的缓存有多种实

@baymaxium
baymaxium / content.md
Created October 18, 2017 09:41
聊聊磁盘I/O那些事

原文:高效开发运维

  

背景  

计算机硬件性能在过去十年间的发展普遍遵循摩尔定律,通用计算机的 CPU 主频早已超过 3GHz,内存也进入了普及 DDR4 的时代。然而传统硬盘虽然在存储容量上增长迅速,但是在读写性能上并无明显提升,同时 SSD 硬盘价格高昂,不能在短时间内完全替代传统硬盘。传统磁盘的 I/O 读写速度成为了计算机系统性能提高的瓶颈,制约了计算机整体性能的发展。

硬盘性能的制约因素是什么?如何根据磁盘 I/O 特性来进行系统设计?

@zhanwenchen
zhanwenchen / Install NVIDIA Driver and CUDA.md
Last active March 13, 2024 23:42 — forked from wangruohui/Install NVIDIA Driver and CUDA.md
Install NVIDIA CUDA 9.0 on Ubuntu 16.04.4 LTS
@jannson
jannson / gf2.py
Created October 24, 2016 07:22 — forked from fcicq/gf2.py
Finite Field / Galois Field (GF(2^n)) in python 2.7+
# GPLv2+, by fcicq
# from http://www.hpl.hp.com/techreports/98/HPL-98-135.pdf
IRRED_COEFF = [
4,3,1,5,3,1,4,3,1,7,3,2,5,4,3,5,3,2,7,4,2,4,3,1,10,9,3,9,4,2,7,6,2,10,9,
6,4,3,1,5,4,3,4,3,1,7,2,1,5,3,2,7,4,2,6,3,2,5,3,2,15,3,2,11,3,2,9,8,7,7,
2,1,5,3,2,9,3,1,7,3,1,9,8,3,9,4,2,8,5,3,15,14,10,10,5,2,9,6,2,9,3,2,9,5,
2,11,10,1,7,3,2,11,2,1,9,7,4,4,3,1,8,3,1,7,4,1,7,2,1,13,11,6,5,3,2,7,3,2,
8,7,5,12,3,2,13,10,6,5,3,2,5,3,2,9,5,2,9,7,2,13,4,3,4,3,1,11,6,4,18,9,6,
19,18,13,11,3,2,15,9,6,4,3,1,16,5,2,15,14,6,8,5,2,15,11,2,11,6,2,7,5,3,8,
3,1,19,16,9,11,9,6,15,7,6,13,4,3,14,13,3,13,6,3,9,5,2,19,13,6,19,10,3,11,