当前位置:网站首页>Common hardware delays
Common hardware delays
2022-08-05 02:18:00 【FatherOfCodingMan】
Latency Numbers Every Programmer Should Know · GitHub
Latency Comparison Numbers (~2012)
----------------------------------
Level 1 Cache Reference 0.5 NS
Branch prediction error Branch Mispredict 5 NS
Tier Class Cache Reference 7 NS 14X L1 Cache
Mutex Lokk/UNLOCK 25 NS
Main Memory Reference 100NS 20X L2 Cache, 200x L1 Cache
Compression Compress 1K Bytes with Zipp 3,000 NS 3 us
Send 1K Bytes Over 1 GBPS Network 10,000 NS 10 US
Read 4K Randomly FROMLY SSD* 150,000 NS 150 US ~ 1G MSD* 150 US ~ 1G MSD* 150 US ~ 1G MSD/sec SSD
Read 1 MB sequentially from memory 250,000 ns 250 us
Round trip within same datacenter 500,000 ns 500 us
Read 1 MB sequentially from SSD* 1,000,000 ns 1 ms ~1GB/sec SSD, 4X memory
Disk seek 10,000,000 ns 10,000 us 10 ms 20x datacenter roundtrip
Read 1 MB sequentially from disk 20,000,000 ns 20,000 us 20 ms 80x memory, 20X SSD
Send packet CA->Netherlands->CA 150,000,000 ns 150,000 us 150 ms
Notes
-----
1 ns = 10^-9 seconds
1 us = 10^-6 seconds = 1,000 ns
1 ms = 10^-3 seconds = 1,000us = 1,000,000 ns
Yearly data are shown below
Numbers Every Programmer Should Know By YearNumbers Every Programmer Should Know By YearNumbers Every Programmer Should Know By Year

Level 1 cache 1ns
Branch prediction failed 3ns
Second level cache 4ns
Mutex lock/unlock 17ns
Main memory 100ns
Zipp Compression 1k 2,000 ns
Public network sending 2k 44ns (the network has improved a lot, I don't know if there is any problem with the data, if the sending and receiving are added up)
Memory reads 1m data continuously for 3,000ns
SSD random read 16,000ns
边栏推荐
- hypervisor相关的知识点
- Greenplum数据库故障分析——版本升级后gpstart -a为何返回失败
- C language basics -- pointers
- CPDA|运营人如何从负基础学会数据分析(SQL)
- Pisanix v0.2.0 发布|新增动态读写分离支持
- Understand the recommendation system in one article: Recall 06: Two-tower model - model structure, training method, the recall model is a late fusion feature, and the sorting model is an early fusion
- PHP技能评测
- HOG feature study notes
- 常见的硬件延迟
- 继承关系下构造方法的访问特点
猜你喜欢

MySQL learning

Using OpenVINO to implement the flying paddle version of the PGNet inference program

The 2022 EdgeX China Challenge will be grandly opened on August 3

Understand the recommendation system in one article: Recall 06: Two-tower model - model structure, training method, the recall model is a late fusion feature, and the sorting model is an early fusion

J9数字货币论:web3的创作者经济是什么?

Xunrui cms website cannot be displayed normally after relocation and server change

RAID disk array

source program in assembly language

.Net C# Console Create a window using Win32 API

Hypervisor related knowledge points
随机推荐
释放技术创新引擎,英特尔携手生态合作伙伴推动智慧零售蓬勃发展
Tree search (bintree)
Fragment visibility judgment
Domain Driven Design - MDD
在这个超连接的世界里,你的数据安全吗
关于#sql shell#的问题,如何解决?
HOG feature study notes
Log an error encountered when compiling google gn "I could not find a ".gn" file ..."
行业案例|世界 500 强险企如何建设指标驱动的经营分析系统
高数_复习_第1章:函数、极限、连续
leetcode-对称二叉树
力扣-二叉树的前序遍历、中序遍历、后序遍历
【日常训练】1403. 非递增顺序的最小子序列
Utilities CMS website construction process
Dotnet 6 Why does the network request not follow the change of the system network proxy and dynamically switch the proxy?
协作D2D局部模型聚合的半分散联合学习
C语言基础知识 -- 指针
".NET IoT from scratch" series
Live preview | 30 minutes started quickly!Look at credible distributed AI chain oar architectural design