Confessions of a Code Addict

Confessions of a Code Addict focuses on computer science topics, including programming language mechanics, optimization techniques, modern applications of data structures and algorithms, insights into AI, and practical guides to optimizing code performance. It targets a broad audience, from beginners to experienced engineers, aiming to enhance their understanding and skills.

Computer Science Fundamentals Programming Language Internals Code Optimization Techniques Data Structures and Algorithms Artificial Intelligence Database Internals Performance Engineering Python Internals Text Classification Techniques

The hottest Substack posts of Confessions of a Code Addict

And their main takeaways

Live Session: CPUs, GPUs & LPUs + Ask Me Anything

336 implied HN points • 08 Mar 24
🕹 Technology Hardware Software Events
  1. The upcoming live session will focus on CPUs, GPUs, and LPUs, exploring what makes them non-deterministic and discussing Groq's LPU design and architecture.
  2. The live session will include an Ask Me Anything (AMA) segment, where attendees can freely ask questions related to the discussed topics or anything else they're curious about.
  3. The event is scheduled for March 17th from 16:30 UTC to 18:00 UTC and will be limited to 100 participants, offering a mix of informative content and interactive Q&A.

Workshop on Performance Optimizations in Code: The One Billion Row Challenge

577 implied HN points • 15 Jan 24
🕹 Technology Programming Workshop Optimization Performance Coding
  1. Code efficiency at scale is crucial - data structures and algorithms matter, but execution cost is also important.
  2. Participating in challenges like the 1 Billion Row Challenge can enhance performance engineering skills.
  3. The workshop covers optimization techniques like flamegraphs, I/O strategies, system calls, SIMD instructions, and more.

Live Session on Performance Optimization Using 1BRC as a Case Study

360 implied HN points • 02 Feb 24
🕹 Technology Coding Performance optimization Live session Software Development Programming Languages
  1. The live session focuses on learning to analyze and reason about code performance through iterative optimization using 1BRC as a case study.
  2. Attendees will explore various topics including performance profiling with flamegraphs, I/O strategies, and leveraging SIMD instructions.
  3. Prerequisites include a few years of coding experience in languages like C, C++, Java, or others, with a specific focus on Java during the session.

CPython Dynamic Dispatch Internals: What Happens When You do a + b?

293 HN points • 06 Dec 23
🕹 Technology Programming Python
  1. Each type in Python implements functions for the operators it supports and populates a function pointer table in its header.
  2. The CPython Virtual Machine calls a function in the abstract object interface based on the operator being executed.
  3. The abstract object interface performs function pointer table lookup in the object headers to call the right function for dynamic dispatch.
Get a weekly roundup of the best Substack posts, by hacker news affinity:

What Every Developer Should Know About GPU Computing

465 HN points • 18 Oct 23
🕹 Technology Programming Architecture
  1. GPUs are designed for high throughput and massive parallelism, while CPUs focus on executing sequential instructions quickly.
  2. GPU architecture includes streaming multiprocessors with cores, various memory layers, and dynamic resource partitioning for efficient execution.
  3. Executing code on GPUs involves launching grids of thread blocks, with each block consisting of threads that work in parallel to optimize performance.

1024 Reasons to Celebrate: Journey and Future of 'Confessions of a Code Addict'

481 implied HN points • 07 Oct 23
🕹 Technology Coding Community Subscription Tech industry
  1. The growth of subscribers for 'Confessions of a Code Addict' has been significant, with spikes after posting articles.
  2. The author's journey started after a job loss, and the community support keeps them dedicated to creating content.
  3. Future plans for the community include introducing paid subscriptions with various perks and additional initiatives.

Fast Computation of Fibonacci Numbers: Part II

288 implied HN points • 12 Nov 23
🕹 Technology Mathematics Programming Algorithms Education
  1. A new method to compute Fibonacci numbers using a closed-form expression without having to resort to floating point arithmetic.
  2. Representation of irrational numbers using two parts can be done in code allowing for precise computation of Fibonacci numbers.
  3. Understanding rings and implementing arithmetic operations within it can help in computing Fibonacci numbers without any loss of precision.

A Linear Algebra Trick for Computing Fibonacci Numbers Fast

158 HN points • 05 Nov 23
🔬 Science Mathematics Computing Algorithms Linear Algebra Complexity
  1. A linear algebra technique can be applied to compute Fibonacci numbers quickly with a logarithmic time complexity.
  2. Efficient algorithms like repeated squaring can compute powers of matrices in logarithmic time, improving performance for Fibonacci number calculations.
  3. A closed form expression using the golden ratio offers a direct method to compute Fibonacci numbers, showing different approaches with varied performance.

A Deep Dive into the Underlying Architecture of Groq's LPU

4 HN points • 01 Mar 24
🕹 Technology Hardware Compiler Distributed Systems Scalability Reliability
  1. Groq's LPU showcases an innovative design departing from traditional architectures, focusing on deterministic execution for enhanced performance.
  2. The TSP architecture achieves determinism through a simplified hardware design, enabling precise scheduling by compilers for predictable performance.
  3. Groq's approach to creating a distributed multi-TSP system eliminates non-determinism typical in networked systems, with the compiler efficiently managing data movement.

Why and How Does Python Use Bloom Filters in String Processing?

46 HN points • 14 Sep 23
🕹 Technology Programming Data Structures Implementation Python
  1. Python uses Bloom filters in its string data structure to speed up certain string processing functions like strip and splitlines.
  2. The unique Bloom filter implementation in CPython uses an unsigned long type to represent the bit vector, making storing and querying items more efficient.
  3. CPython determines the position in the bit vector for adding and querying characters by using the lower n-bits of the character, avoiding costly hash computations.

CPython Object System Internals: Understanding the Role of PyObject

4 HN points • 23 Nov 23
🕹 Technology Programming
  1. Programming languages use polymorphism and inheritance to create unified interfaces between modules.
  2. CPython mimics inheritance and polymorphism in C by embedding a parent struct in the child struct definitions, ensuring a unified memory layout.
  3. In CPython, the PyObject struct acts as the parent struct, containing fields for object reference count and type-related data, enabling type-specific method calls via function pointer tables in the object header.

Understanding Immortal Objects in Python 3.12: A Deep Dive into Python Internals

6 HN points • 25 Aug 23
🕹 Technology Memory management
  1. Python 3.12 introduced immortal objects to improve performance by avoiding constant reference count updates for immutable objects like None, True, and False.
  2. Immortalization in Python aimed to address performance issues like CPU cache invalidation, data races, and impact on Copy-on-Write applications.
  3. The implementation of immortalization in Python included marking objects as immortal, incorporating checks in Py_INCREF and Py_DECREF functions, and dealing with some performance overhead.

Bloom Filters and Beyond: An Illustrated Introduction and Implementation

5 HN points • 05 Sep 23
🕹 Technology Data Structures Programming Algorithms Python Applications
  1. Bloom filters are efficient data structures for quick searches in large datasets and minimize memory usage, with a probabilistic approach to determining membership
  2. Bloom filters use hash functions and bit vectors to store data item membership information while conserving memory by not storing actual items
  3. Counting Bloom Filters are an extension that allow item deletion but come with weaknesses such as handling hash collisions and counter overflow, providing an advanced data handling tool