Communications of the ACM
Refine your search:
How the “Frontier” Became the Slogan of Uncontrolled AI
Artificial intelligence (AI) has been billed as the next frontier of humanity: the newly available expanse whose exploration …
From Daniel Lemire's Blog
A look back to my 2010 predictions for 2020
Back in 2010, I wrote a post Who is going to need a database engine in 2020? Let me revisit some of my 2010 statements. Apple will sell desktops with 1 TB of RAM...
From Daniel Lemire's Blog
Science and Technology links (December 14th 2019)
The computation capacity needed by artificial intelligence doubles every 3.4 months. In parallel, we are making fast progress in hardware and software: what took...
From Daniel Lemire's Blog
Are 64-bit random identifiers free from collision?
It is common in software system to map objects to unique identifiers. For example, you might map all web pages on the Internet to a unique identifier. Often, these...
From Daniel Lemire's Blog
Amazon’s new ARM servers: Graviton 2
Most servers on the Internet run on x64 processors, mostly made by Intel. Meanwhile, most smartphones run ARM processors. From a business perspective, these are...
From Daniel Lemire's Blog
Science and Technology links (December 7th 2019)
Incredibly, there is a new simpler way to solve the quadratic formula. I used to rely on the completion of the square, but this is better! There is a video report...
From Daniel Lemire's Blog
AMD Zen 2 and branch mispredictions
Intel makes some of the very best processors many can buy. For a long time, its main rival (AMD) failed to compete. However, its latest generation of processors...
From Daniel Lemire's Blog
Instructions per cycle: AMD Zen 2 versus Intel
The performance of a processor is determined by several factors. For example, processors with a higher frequency tend to do more work per unit of time. PhysicsContinue...
From Daniel Lemire's Blog
Better computational complexity does not imply better speed
A recent magazine article presents a theoretical result: Harvey and van der Hoeven have shown that you can multiply two n-bit integers using O(n log n) complexity...
From Daniel Lemire's Blog
Memory parallelism: AMD Rome versus Intel
When thinking about “parallelism”, most programmers think about having multiple processors. However, even a single core in a modern processor has plenty of parallelism...
From Daniel Lemire's Blog
Cloud computing: a story of incentives
Many businesses today run “in the cloud”. What this often means is that they have abstracted out the hardware entirely. Large corporations like Amazon, Google,Continue...
From Daniel Lemire's Blog
Science and Technology links (November 16th 2019)
We have new technology to do genetic engineering on human beings (CRISPR). In a small clinical trial, the researchers tested it on live human subjects and found...
From Daniel Lemire's Blog
Unrolling your loops can improve branch prediction
Modern processors predict branches (e.g., if-then clauses), often many cycles a ahead of time. When predictions are incorrect, the processor has to start again,...
From Daniel Lemire's Blog
Adding a (predictable) branch to existing code can increase branch mispredictions
Software is full of “branches”. They often take the form of if-then clauses in code. Modern processors try to predict the result of branches often long before evaluating...
From Daniel Lemire's Blog
Parsing numbers in C++: streams, strtod, from_chars
When programming, we often want to convert strings (e.g., “1.0e2”) into numbers (e.g., 100). In C++, we have many options. In a previous post, I reported that it...
From Daniel Lemire's Blog
Science and Technology links (October 26th 2019)
People who were the oldest in the classes in school tend to be more confident and to take more risks. At the University of Montreal, about 32% of the students are...
From Daniel Lemire's Blog
How expensive is it to parse numbers from a string in C++?
In software, we frequently have to parse numbers from strings. Numbers are typically represented in computers as 32-bit or 64-bit words whereas strings are variable...
From Daniel Lemire's Blog
Benchmarking is Hard: Processors Learn to Predict Branches
A lot of software is an intricate of branches (if–then clauses). For performance reasons, modern processors predict the results of these branches. In my previous...
From Daniel Lemire's Blog
Mispredicted branches can multiply your running times
Modern processors are superscalar, meaning that they can execute many instructions at once. For example, some processors can retire four or six instructions per...
From Daniel Lemire's Blog
Science and Technology (October 12th 2019)
In many countries, like Canada, there is relatively little private (business) research. Meanwhile, other research indicates that private research is precisely the...
From Daniel Lemire's Blog