Computer

• I recently got a mini host for the office, thinking it would be convenient to configure an environment and have occasional access at home. I temporarily deployed internal network penetration using frp – specifying port forwarding, which requires a public server with a connection quality dependent on its bandwidth. Instead, I experimented with a fresh Zerotier virtual LAN, similar to a VPN, where I created a virtual network card locally and all machines joined it into a single virtual network.

What is ZeroTier

ZeroTier is a software-defined wide area network (SD-WAN) solution that allows users to create secure virtual networks between devices in different geographic locations. Through ZeroTier, you can easily connect multiple computers, servers, and devices into a virtual, encrypted network – as if they were on the same local network. This helps developers and IT professionals securely share data and resources across different locations without complex network setups or VPN configurations.

When installing a development system with VMware virtual machines, it’s generally recommended to allocate extra disk space. Over time, the local disk space consumed by the VM will far exceed the actual size of its files.

Scenario Description

The df -h command revealed that the current machine was using 60GB of disk space, and after deleting all snapshots and clone images, the local virtual machine still occupied significantly more than 60GB, further straining the already limited hard drive.

Domestic resources are basically all recommending Autumn Leaf’s one-click deployment package, thinking that they are open-source projects based on Python, so deployment wouldn’t be very complicated, let’s try to start from scratch.

I was messing around with AI-generated images and specifically changed my graphics card, a beginner version of the 3060 12g; the seven-year-old 960 retired gloriously.

The core pytorch cuda installation, which I previously encountered issues with when writing Python game helper scripts (I had installed it locally before), still presented problems – the cuda encryption consistently failed to activate.

one loop thread, the time taken has already been at the microsecond level, switching servers resulted in a backlog of up to 60,000 packets, to almost none.

In single-threaded loop processing data scenarios, the CPU performance depends on factors such as clock frequency, cache size, and instruction set architecture. Generally, CPUs with higher clock frequencies, larger caches, and more advanced instruction set architectures perform better in single-threaded data processing.

Single-Threaded

Performance improvements aren’t always achieved by adding threads; it’s not necessary to overcomplicate things. Refine the project workflow, identify time-consuming bottlenecks, and determine if a single thread can meet the requirements. Considering single-threaded approaches reduces complexity and minimizes potential issues.

Just as we needed to learn the techniques of searching engines back then, we also need to learn some techniques for communicating with AI, providing reasonable and sufficient constraints, and efficiently obtaining the answers we need.

If you look at it from a different angle, current AI is like a very good student with excellent memory – it has the ability to memorize everything. What we need to do is learn how to communicate with AI correctly, effectively, and precisely describe our needs to help AI generate the expected results.

WeChat Mini Program Introduction and Development Preparation

Why Mini Programs Exist

• Better Experience: Slow web loading and blank screens; native app experience is faster.
• Standards & Management: For WeChat, onboarding and management Before mini programs were released, WeChat published an SDK called JSDk to open up some of the original WeChat capabilities: WeChat Pay, coupons. However, developers used web development languages to develop logic, bypassing some WeChat regulations. Mini Programs have their own description language.

What are Mini Programs?

Mini programs are applications that can be used without needing to download and install them. They realize the dream of having applications within easy reach.

Administrative notice, office relocation from the second floor to the fifteenth floor – a standard, routine desk move.

Design Sense

Migration

Closing up shop, packing everything away, a familiar route, a new workstation – adjusting computer cabling, finding a comfortable posture to start working. (ÒωÓױ)! – Connecting the network cable, the servers frequently used by the team were inaccessible. I tried switching to wireless networking, and access was normal again.

Here’s the English translation:

“When I think of embedded systems, my mind immediately goes back to the 51-pin single-chip microcomputers and Texas Instruments we used in university labs.

LPA3399Pro is a visual host developed on the HiSilicon RK3399Pro platform, designed for portable computing hosts that require extensive visual computation. It features an integrated NPU compute unit with 3.0 TOPS of processing power and supports various algorithm models.

RV1109 is HiSilicon’s SoC for machine vision within its AI branch, featuring a dedicated NPU. RV1109 provides 1TOPS of processing power.”