1 Why python and C++ Pros and Cons of each language Respective usecases 2 C ++getting usable Language “simplifications” Making bad code harder to write 3 Performant C++ New performance related features Templates Avoiding virtuality when possible 4 Migrating from Python 2 to python 3 Tour of python 3 changes How to support both versions How to migrate 5 Conclusion

1 Introduction 2 Measuring vectorization 3 Vectorization Prerequisite 4 Vectorizing techniques in C++ Autovectorization Inline assembly Intrinsics Compiler extensions Libraries 5 What to expect ?

1 Introduction 2 Measuring vectorization 3 Vectorization Prerequisite 4 Vectorizing techniques in C++ Autovectorization Inline assembly Intrinsics Compiler extensions Libraries 5 What to expect ?

Intro Threading Safety Solutions 1 Introduction and expectations Scope : what is parallelism ? Finding parallelism Expectations 2 Threading, theory and practice Processes and threads Python threading C ++threading and async 3 Thread-safety issues Data races Thread safety 4 Thread-safety solutions Avoid the problem Replicate Atomics Locking

Intro Threading Safety Solutions 1 Introduction and expectations Scope : what is parallelism ? Finding parallelism Expectations 2 Threading, theory and practice Processes and threads Python threading C ++threading and async 3 Thread-safety issues Data races Thread safety 4 Thread-safety solutions Avoid the problem Replicate Atomics Locking

中国科学院高能所计算中心：数据技术上机 Data Technologies – CERN School of Computing 2019

1. IO system • What do you know already? • IOPS, bandwidth, latency & blocksize • media and their characteristics • cache & IO optimisation strategies • how to debug IO problems 2. Redundancy Technology • Parity for RAID technology 3. Cloud Storage Technology • Scalability, Hashing, Indexing,, Deduplication

略去余式R[f],由定理5.1.2知,它如果是插值型求积公式,则至少有n次代数精度

在数学分析中,我们学习过微积分基 本定理 Newton-Leibniz-公式: f(x)dx=fx)=fb)-f(a)5.0.1) 其中,F(x)是被积函数f(x)的原函数。 随着学习的不断深化,发现Newton- Leibniz公式有很大的局限性

前面我们根据区间[ab]上给出的节点做 插值多项式Ln(x)近似表示f(x)。一般总 以为Ln(x)的次数越高,逼近f(x)的精度 越好,但实际并非如此,次数越高,计 算量越大,也不一定收敛。因此高次插 值一般要慎用,实际上较多采用分段低 次插值