C++ requires many algorithms to be implemented in amortized constant time. I was curious about what this meant for vector insertion. To get started I first wrote an article about amortized time […]
The amortized cost of vector insert is 3
How does a mutex work? What does it cost?
Concurrent programming requires synchronization. We can’t have more than one thread accessing data at the same time; otherwise, we end up with a data race. The most common solution is to wrap […]
Is reference counting slower than GC?
“Reference counting is slower than garbage collection”, a claim often made in the discussion of memory management. I heard it again recently when discussing Leaf; somebody took issue with my arguments against […]
What is the cost of reference counting?
Reference counting is a common form of memory management. In some languages, like Python, it’s the natural way objects are tracked, and in others it’s in the standard library, such as shared_ptr […]
A real cost of low performing applications
The platform as a service model adds new importance to the performance of our software. Despite ever decreasing hardware costs, metered services make every bit count. My previous article on performance touched […]
Quickly drawing a rounded rectangle with a GL shader
Rectangles appear everywhere in user interfaces, from the backgrounds of elements to rounded border decorations. It’s helpful to be able to draw these fast. I set out to draw a rectangle with […]
What is amortized time?
Amortized time is often used when stating algorithm complexity. Instead of giving values for worst-case performance it provides an average performance. This is appropriate in many domains, but we must be careful. […]
Wait-free queueing and ultra-low latency logging
To support a low-latency trading system I needed a logging call that had minimal impact on the calling thread. It was imperative to be responsive to market events at all times. My […]
The true cost of zero cost exceptions
Exceptions are a cornerstone of programming. If we intend to use them prolifically, as in C++, we need an efficient implementation. Thus was born the zero-cost exception model. It offers zero runtime […]
Pitfalls of Performance Testing
Beware the casual performance test, it will lie to you. It is simple to write something that gives you timings, but it is difficult to make those timings meaningful. A whole host […]
Performance is irrelevant
Asking questions about performance online universally invites scorn and accusation. A large number of programmers apparently feel that the efficiency of code is nowadays insignificant. So long as the functional requirements have […]
How Polymorphism Works: Part 2: Virtual Table
In the article on “How Polymorphism Works: Part 1” we learned how to create virtual functions. The method which we chose has at one significant problem with it: it requires a lot […]
How Polymorphism Works: Part 1
Polymorphism: the core of object oriented programming. Most modern languages have some concept of interfaces, virtual functions, and classes. Though each language differs in details, and may have specialized concepts, the core […]
What is an atomic operation?
We all know that operations in a computer happen one after the other. Having multiple CPUs and instruction reordering tends to obfuscate the issue. Technically things still proceed in an orderly fashion, […]
