This blog post discusses the parallel signal sending optimization that recently got merged into the master branch (scheduled to be included in Erlang/OTP 25). The optimization improves signal sending throughput when several processes send signals to a single process simultaneously on multicore machines. At the moment, the optimization is only active when one configures the receiving process with the {message_queue_data, off_heap} setting. The following figure gives an idea of what type of scalability improvement the optimization can give in extreme scenarios (number of Erlang processes sending signals on the x-axis and throughput on the y-axis):

A shared Erlang Term Storage (ETS) table is often an excellent place to store data that is updated and read from multiple Erlang processes frequently. ETS provides key-value stores to Erlang processes. When the write_concurrency option is activated, ETS tables use fine-grained locking internally. Therefore, a scenario where multiple processes insert and remove different items in an ETS table should scale well with the number of utilized cores. However, in practice the scalability for such scenarios is not yet perfect. This blog post will explore how the decentralized_counters option brings us one step closer to perfect scalability.

Finally Erlang/OTP 24 is here! A release that for me has been about 10 years in the making. As is tradition by now, this blog post will go through the additions to Erlang/OTP that I am most excited about!

Message passing has always been central to Erlang, and while reasonably well-documented we’ve avoided going into too much detail to give us more freedom when implementing it. There’s nothing preventing us from describing it in a blog post though, so let’s have a closer look!

As long as Erlang has existed, there has always been the need and the ambition to make it faster. This blog post is a history lesson that outlines the major Erlang implementations and attempts to improve the performance of Erlang.

This post continues our adventures in the JIT, digging a bit deeper into the implementation details.

Now that we’ve had a look at BEAM and the interpreter we’re going to explore one of the most exciting additions in OTP 24: the just-in-time compiler, or “JIT” for short.

In my previous post we had a look at BEAM, and now that we’re more familiar with it it’s time for us to look at the reference implementation: the interpreter.

This post is a brief primer on BEAM, the virtual machine that executes user code in the Erlang Runtime System (ERTS). It’s intended to help those new to BEAM follow an upcoming series of posts about the JIT in OTP 24, leaving implementation details for later.

The scalability of ETS tables of type ordered_set with the write_concurrency option is substantially better in Erlang/OTP 22 than earlier releases. In some extreme cases, you can expect more than 100 times better throughput in Erlang/OTP 22 compared to Erlang/OTP 21. The cause of this improvement is a new data structure called the contention adapting search tree (CA tree for short). This blog post will give you insights into how the CA tree works and show you benchmark results comparing the performance of ETS ordered_set tables in OTP 21 and OTP 22.

OTP 23 has just been released (May 13:th 2020). It has been a long process with three release candidates in February, March and April before the final release. We are very thankful for the feedback we have got regarding the release candidates, which has revealed some bugs and flaws that our internal testing did not find.

This blog post will go through three different uses of persistent_term that I have used since its release and explain a bit why they work so well with persistent_term.

OTP 22 has just been released. It has been a long process with three release candidates before the final release. We decided this year to try to get one month more testing of the major release and I think that the extra time has paid off. We’ve received many bug reports from the community about large and small bugs that our internal tests did not find.

When working with the implementation of the new scalable ordered_set we came across a strangeness with the guarantees when iterating over a table while inserting elements in parallel.

Erlang/OTP 22 will bring many performance improvements to the table, but most of them have a broad impact and don’t affect the way you write efficient code. In this post I’d like to highlight a few things that used to be surprisingly slow but no longer need to be avoided.

Erlang/OTP 22 will be an important release for the ssl application. We are working on several new features and improvements such as support for TLS 1.3, some of those are already on the master branch. This blog post presents the new ssl debug logging built on the new logger API.

This blog post looks back on the development of the SSA-based intermediate representation from the beginning of this year to the end of August when the branch was merged.

This blog post continues the exploration of the new SSA-based intermediate representation through multiple examples. Make sure to read the Introduction to SSA if you missed it.

This blog post is an introduction to the new SSA-based intermediate representation that has recently been merged to the master branch in the Erlang/OTP repository. It uses the same example as in the previous blog post, first looking at the generated SSA code, and then at some optimizations.

Back after the summer holidays, this blog will now change tracks and start a series of blog posts about Static Single Assignment (SSA). This first installment will set the scene for the posts that follow by looking at the traps and pitfalls one can fall into when trying to optimize BEAM assembly code.

This blog post is a brief history lesson about the Erlang compiler for the BEAM machine. To provide some context, there will first be a quick look at the abstract machines for Erlang.

The BEAM interpreter in erts has been completely re-written in OTP 21. Most of the instructions have remained the same, but the perl scripts used to generate the C code have a new implementation. This blog post will look at some of the optimizations that were possible because of those changes.

This blog post wraps up the exploration of Core Erlang started in the previous two blog posts. The remaining default Core Erlang passes are described, followed by a look at how Core Erlang is represented internally in the compiler.

This blog post continues the exploration of Core Erlang by looking at some optimizations done by the sys_core_fold compiler pass. The Core Erlang language was introduced in the previous blog post.

This blog post is the first about the Core Erlang format. In this blog post, we introduce the Core Erlang format through examples that compare Erlang code to the corresponding Core Erlang code.

OTP-21 Release Candidate 1 has just been released. I thought that I would go through the changes that I am the most excited about. Most likely this will mostly mean features in erts and the core libraries as those are the changes that I am the most familiar with.

The memory instrumentation module was rewritten for Erlang/OTP 21 to make it easier to use. In this post I’ll describe the rationale behind the new features and how to make use of them.

In this blog post, we will explore the compiler passes that make up the compiler’s front end.

This is the first of a series of blog posts about the compiler. There will be blog posts about how the compiler works now, how it might work in the future, and some historical notes to explain why some things are what they are. In this blog post I will talk about one of the most useful options for exploring the compiler, namely the time option.

Erlang/OTP 21 will introduce a completely new IO polling implementation. This new implementation comes with a new set of tuneable parameters that can be used to get the most out of your system. This blog post describes the parameters and attempts to describe what they should be used for.