Development

• Streaming of content from the rich text editor to other infrastructure tools like full-text search, warehouses, etc.
• Retrieving of content to clients: mobile, web, desktop.
• Storing of content in some kind of storage (SQL database in my case)
• Analyzing of content, which includes point 1, but also analyzing it from the perspective of our application

Update 12 of the Big Data Tools plugin for IntelliJ IDEA Ultimate, PyCharm Professional Edition, and DataGrip has been released. You can install it from the JetBrains Plugin Repository or from inside your IDE. The plugin allows you to edit Zeppelin notebooks, upload files to cloud filesystems, and monitor Hadoop and Spark clusters.

In this release, we've added experimental Python support and global search inside Zeppelin notebooks. We’ve also addressed a variety of bugs. Let's talk about the details.

This year, each of us seeks a special way to pass the time. I am writing a book, for example. A book about one thing I love dearly: the API. (You may read who am I and what expertise got in APIs in my LinkedIn profile.)

I've just finished the first large section dedicated to the API design. You may read it online, or download either pdf or epub version, or take a look at the source code on Github.

The book is distributed for free under a CC-BY-NC license. Enjoy!

This is the second part of the article covering the Sliding Window Pattern and its implementation in Go, the first part can be found here.

Let's have a look at the following problem: we have an array of words, and we want to check whether a concatenation of these words is present in the given string. The length of all words is the same, and the concatenation must include all the words without any overlapping. Would it be possible to solve the problem with linear time complexity?

Let's start with string catdogcat and target words cat and dog.

How can we handle this problem?

Designing a circuit that calculates integer cube root.

In this article I will make a circuit on Verilog HDL on FPGA which will calculate cube root from integer value.

"Come, let us make bricks, and burn them thoroughly."
– legendary builders

You may have noticed by 2020 that data is eating the world. And whenever any reasonable amount of data needs processing, a complicated multi-stage data processing pipeline will be involved.

At Bumble — the parent company operating Badoo and Bumble apps — we apply hundreds of data transforming steps while processing our data sources: a high volume of user-generated events, production databases and external systems. This all adds up to quite a complex system! And just as with any other engineering system, unless carefully maintained, pipelines tend to turn into a house of cards — failing daily, requiring manual data fixes and constant monitoring.

For this reason, I want to share certain good engineering practises with you, ones that make it possible to build scalable data processing pipelines from composable steps. While some engineers understand such rules intuitively, I had to learn them by doing, making mistakes, fixing, sweating and fixing things again…

So behold! I bring you my favourite Rules for Data Processing Pipeline Builders.

The FT4232H is USB 2.0 High speed to UART IC converter. The FT4232H has four UART ports and one USB port.

By connecting EEPROM memory to this chip, you can set specific operating modes or change the manufacturer's data.

Let's look at the example and configure FT4232H directly on a system running GNU/Linux. We will do this using the ftdi_eeprom.

Let's consider the following problem: we have an array of integers and we need to find out the length of the smallest subarray the sum of which is no less than the target number. If we don't have such a subarray we shall return -1.

We can start with a naive approach and consider every possible subarray in the input:

Every time when the essential question arises, whether to upgrade the cards in the server room or not, I look through similar articles and watch such videos.

Channel with the aforementioned video is very underestimated, but the author does not deal with ML. In general, when analyzing comparisons of accelerators for ML, several things usually catch your eye:

• The authors usually take into account only the "adequacy" for the market of new cards in the United States;
• The ratings are far from the people and are made on very standard networks (which is probably good overall) without details;
• The popular mantra to train more and more gigantic models makes its own adjustments to the comparison;

The answer to the question "which card is better?" is not rocket science: Cards of the 20* series didn't get much popularity, while the 1080 Ti from Avito (Russian craigslist) still are very attractive (and, oddly enough, don't get cheaper, probably for this reason).

All this is fine and dandy and the standard benchmarks are unlikely to lie too much, but recently I learned about the existence of Multi-Instance-GPU technology for A100 video cards and native support for TF32 for Ampere devices and I got the idea to share my experience of the real testing cards on the Ampere architecture (3090 and A100). In this short note, I will try to answer the questions:

• Is the upgrade to Ampere worth it? (spoiler for the impatient — yes);
• Are the A100 worth the money (spoiler — in general — no);
• Are there any cases when the A100 is still interesting (spoiler — yes);
• Is MIG technology useful (spoiler — yes, but for inference and for very specific cases for training);