Table of Contents

IO Ecosystem

  • IO Coin
    Connecting all the Web3 money that’s flowing into the GPU to power io.net. A blockchain-based cryptocurrency and platform designed to foster a decentralized ecosystem for applications and services. Operating on its own blockchain, IO Coin employs a hybrid consensus mechanism merging Proof of Work (PoW) and Proof of Stake (PoS) to ensure security and energy efficiency. With support for decentralized applications (dApps), including features like smart contracts and token issuance, IO Coin enables diverse functionalities. It integrates privacy-enhancing features like ZeroCoin and Ring Signatures, providing anonymity. Its versatility extends to applications in peer-to-peer transactions, remittances, and decentralized finance (DeFi).
  • IO Worker
    IO Worker is a component of the IO.NET ecosystem that enables users to rent out their computing devices like GPUs and CPUs to those needing computational power. By leasing their device's processing capabilities, users earn rewards for tasks like artificial intelligence computations or rendering. This setup fosters decentralized computing and resource sharing, fostering collaboration and mutual benefit among users on the IO.NET platform.
  • IO Explorer
    IO Explorer is a multifunctional tool within the IO.NET ecosystem designed for users to explore and navigate various aspects of the platform. It serves as a dashboard where users can monitor compute jobs, access performance metrics, and explore available resources within the IO.NET network. Additionally, IO Explorer provides insights into others' activities, offering a comprehensive view of the platform's functionalities and user engagements.
  • IO Cloud
    IO Cloud is a component of the IO.NET ecosystem that provides cloud computing services to users. It enables users to deploy and manage virtual machines, containers, and other cloud resources on-demand. IO Cloud offers scalability, flexibility, and accessibility, allowing users to harness the power of cloud computing for various applications and workloads. IO Cloud simplifies the process of managing cloud infrastructure and empowers users to focus on their core tasks without worrying about underlying infrastructure complexities.
  • BC8
    BC8 is an advanced AI-driven image generator that utilizes deep neural networks to create highly detailed and precise images based on text descriptions or seed images. It's engineered to comprehend context, style, and content, resulting in visually impressive and contextually relevant images. You can access it at bc8.ai.

Basic Terms

  • Client
    Customers who hire GPU/CPU compute power.
  • Binary
    A binary is a file that contains executable instructions in a format that a computer can directly execute. It represents a software application in a form that the computer's processor can understand and run.
  • Worker
    Are the nodes in the cluster that execute the tasks assigned by the master node. They process data, perform computations, and contribute to the overall workload of the system.
  • Solscan
    Solscan is a Solana block explorer (blockchain explorer) that enables investors to view transactions, explore wallets, find important data, and better understand other key metrics of the Solana ecosystem.
  • Solana
    Solana is a high-performance blockchain platform designed for decentralized applications (dApps) and cryptocurrency transactions. It aims to provide fast and scalable solutions for developers, with the ability to process thousands of transactions per second.
    Solana uses a unique consensus mechanism called Proof of History (PoH) combined with Proof of Stake (PoS) to achieve high throughput and low latency. The platform also offers low transaction fees and supports smart contracts, making it suitable for a wide range of applications in finance, gaming, and decentralized finance (DeFi).
  • Aptos
    Aptos is a blockchain platform designed for high scalability, security, and efficiency in decentralized applications (DApps). It aims to provide fast transaction speeds and strong security through a novel consensus mechanism and advanced cryptography. Aptos supports smart contracts, enabling developers to build various DApps, and emphasizes a user-friendly experience and robust developer tools.
  • Computing
    Computing refers to the process of performing calculations, such as addition, multiplication, or more complex mathematical functions. This term is closely associated with computers, which are designed to perform computations rapidly and efficiently.
  • Compute Hours
    Compute hours are the measurable hours or, time that your process is loaded and executing. Compute hours are one of the two main metrics that are used to determine costs.
  • Cluster Processor
    CPU/GPU unit designed to handle parallel computing workloads within a cloud-based cluster. These processors are used for tasks that can be parallelized across multiple cores or nodes within a cluster, such as data analytics, scientific simulations, machine learning training, and other high-performance computing (HPC) workloads.
  • Connectivity Tier
    It's a speed or bandwidth of internet connectivity provided by an internet service provider (ISP) or telecommunications company. It represents the rate at which data can be transmitted over a network connection, typically measured in megabits per second (Mbps) or gigabits per second (Gbps).
  • Blockchain Prover
    A computational entity that confirms that information is accurate without revealing its underlying data. Provers create "proofs" that can be easily verified by a verifier. Traditionally provers generates proofs via Proof of Work (PoW), and some migrated to Proof of Stake (PoS), some are now generating Zero Knowledge Proofs.
  • Containerized Workload
    An application or software workload that has been packaged into a containerized format. Containers are a lightweight, portable, and self-contained unit of software that includes all the necessary dependencies, libraries, and configuration files needed to run the application.
  • DePIN
    Decentralized Physical Infrastructure Networks, leverages blockchains, IoT and the greater Web3 ecosystem to create, operate and maintain real-world physical infrastructure. These networks leverage token incentives to coordinate, reward and safeguard members of the network.
  • Node
    Node AI is a platform that connects you with GPU and artificial intelligence resources in a decentralized way. It uses blockchain technology to ensure security and transparency, enabling users to engage in a range of activities securely.
  • Decentralized Applications
    Decentralized applications (dApps), are software programs that run on a blockchain or peer-to-peer (P2P) network of computers instead of on a single computer. Rather than operating under the control of a single authority, dApps are spread across the network to be collectively controlled by its users.
  • Script FIle
    A script file is a file that contains a sequence of commands or instructions written in a scripting language. Scripting languages, such as Bash, Python, PowerShell, and JavaScript, allow you to automate tasks, execute programs, and perform various operations on a computer or within a software environment.
  • Proof-of-Work (PoW)
    The Proof-of-Work (PoW) consensus algorithm was brought to fruition with the inception of Bitcoin in 2009. It serves as the mechanism for validating transactions and generating new blocks within a blockchain. This process involves specialized devices, computers, or graphics cards performing complex calculations. In PoW, the discovery or creation of a new block is achieved through solving a cryptographic puzzle, a task known as mining. Miners invest significant computational power and energy in attempting to solve these puzzles, which forms the foundation of the term 'Proof-of-Work'.
  • Job
    Job refers to a specific task allocated to a GPU cluster for execution, such as machine learning training or data analysis. It involves parameters and instructions for efficient execution.
  • Random Access Memory (RAM)
    RAM is a type of computer memory that allows data to be accessed and read in any order, making it faster than storage devices like hard drives. It temporarily holds data and instructions that are actively being used or processed by the CPU (Central Processing Unit). RAM is volatile memory, meaning it loses its contents when the power is turned off.
  • SXM
    SXM is a high-performance connection standard that allows GPUs to be directly mounted onto a motherboard without the need for PCIe (Peripheral Component Interconnect Express) slots.
  • BIOS
    The BIOS (Basic Input/Output System) is built-in software on your computer's motherboard that starts up your computer and ensures all hardware works together properly. It also lets you change basic settings through an easy-to-navigate menu.
  • UEFI
    Unified Extensible Firmware Interface (UEFI) is modern software that starts up your computer and helps it run smoothly. It's like an upgraded version of BIOS, with a more user-friendly interface, faster startup times, and better support for large hard drives. It also provides more advanced security features to protect your system from threats.
  • WSL 2
    Windows Subsystem for Linux 2 (WSL 2) is a feature in Windows that lets you run a full Linux system on your computer without needing to set up a separate machine or use complex software. It provides a seamless way to use Linux tools and applications alongside your regular Windows programs, making it easier for developers and tech enthusiasts to work with both systems at the same time.

Device Type

  • Central Processing Unit (CPU)
    CPU stands for Central Processing Unit. It is the primary component of a computer responsible for executing instructions and performing calculations required to run software programs and operating systems.
  • Graphics Processing Unit (GPU)
    Graphics Processing Unit, is a special computer chip that helps make images and videos appear on your screen faster. It's like a supercharged engine for handling visual tasks, such as gaming, watching videos, and designing graphics. They accelerate the computational tasks involved in training and running machine learning models.

Clusters

Cluster
A group of interconnected computers or servers that work together to perform tasks or provide services. Clustering allows multiple machines to function as a single system, enabling improved performance, scalability, and reliability. Here are some key characteristics and types of clusters.

  • Ray Cluster
    A cluster of machines managed by the Ray framework. Ray is an open-source framework for building and running distributed applications. It provides a simple, universal API for building distributed applications efficiently. Typically consists of multiple machines (nodes) connected together to form a distributed computing environment. These machines work together to execute tasks and manage resources efficiently.
  • Mega-Ray
    The supply offers a cutting-edge global networking infrastructure on a diverse selection of enterprise-grade GPU models, all of which meet the highest standards of security compliance. However, this comes at a premium cost.
  • Kubernetes (AKA k8s)
    Open-source platform designed to automate the deployment, scaling, and management of containerized applications. Provides a framework for automating the management of containerized workloads and services, allowing organizations to abstract away the underlying infrastructure and focus on developing and deploying their applications.

Base Image

A base image is the core starting point for creating containers or virtual machines, containing essential components and dependencies needed to run applications or systems.

  • Ray App
    Ray is an open-source distributed computing framework primarily used for scaling Python applications across clusters. Ray provides a set of libraries for building distributed applications, including machine learning training, hyperparameter tuning, reinforcement learning, and more.
  • Pytorch FSDP
    PyTorch FSDP stands for PyTorch Fully Sharded Data Parallelism. It's a distributed training technique designed to efficiently train large deep learning models across multiple GPUs or even across multiple machines. FSDP achieves this by sharding (splitting) the model parameters and activations across multiple devices, allowing for parallel computation during training.
  • Ludwig
    Ludwig is an open-source, declarative deep learning model building framework developed by Uber AI Labs. It aims to provide a simple and flexible way to train and test deep learning models without requiring extensive knowledge of machine learning or deep learning frameworks. Ludwig enables users to build and deploy deep learning models for a variety of tasks, including natural language processing (NLP), computer vision, time series forecasting, and more.
  • IO Native App
    A specialized software development kit provided by IO.NET, based on a fork of Ray, designed to streamline model development, training, and deployment within the ecosystem. It supports the parallelization of Python functions, dynamic task execution, and effortless scalability, empowering developers to build and scale their AI applications seamlessly on the network.
  • Unreal Engine 5
    Unreal Engine 5 (UE5) is a powerful and popular real-time 3D creation platform primarily used for developing video games, architectural visualizations, virtual reality (VR) experiences, and more. Machine learning algorithms for computer vision can be used to enhance augmented reality (AR) or mixed reality (MR) experiences created with Unreal Engine 5. For example, object recognition and tracking algorithms can enable more realistic interactions between virtual and real-world objects in AR applications.
  • Unity Streaming
    Unity Render Streaming is a technology that brings Unity's powerful rendering capabilities to web browsers, allowing users to experience high-quality graphics directly in their browser without additional software installations.

Cluster Type

  • General
    Best for prototyping or general end-to-end (E2E) Workloads. Virtual Machine (VM) clusters are often straightforward to set up and configure, making them suitable for prototyping. Virtual machines can be quickly provisioned and customized to match specific requirements, enabling developers to experiment with different configurations and environments.
  • Train
    For production-ready clusters for machine learning model training and fine-tuning, Train clusters with specialised machine learning orchestration tools are often preferred. This Cluster provides a scalable, reliable, and flexible infrastructure for deploying and managing containerized applications, while machine learning orchestration tools offer features tailored to the unique requirements of training and deploying machine learning models.
  • Inference
    By deploying Inference services on the cluster with efficient resource management, auto-scaling capabilities, hardware acceleration, and robust monitoring, it's possible to build a production-ready infrastructure capable of handling low-latency inference and heavy workloads at scale.
  • Inference
    Refers to the process of using a trained model to make predictions, decisions, or classifications based on new, unseen data. In other words, it's the application of a machine learning model to real-world data to derive insights or take action.
  • NV Link
    NVLink is a high-speed communication interface developed by NVIDIA for connecting GPUs (Graphics Processing Units) together. It enables direct communication between GPUs, allowing them to work together more efficiently by sharing data at extremely high speeds. NVLink is designed to enhance performance in tasks that require parallel processing, such as deep learning, scientific simulations, and high-performance computing.
  • Green GPUs
    Green computing is the practice of maximizing energy efficiency and minimizing environmental impact in the ways computer chips, systems and software are designed and used.

Additional Software

  • CUDA
    The NVIDIA CUDA Toolkit provides a development environment for creating high-performance, GPU-accelerated applications. With it, you can develop, optimize, and deploy your applications on GPU-accelerated embedded systems, desktop workstations, enterprise data centers, cloud-based platforms, and supercomputers. The toolkit includes GPU-accelerated libraries, debugging and optimization tools, a C/C++ compiler, and a runtime library.
  • Docker
    Docker is a platform that allows developers to develop, ship, and run applications in containers. Containers are lightweight, portable, and self-sufficient units that contain everything needed to run an application, including the code, runtime, system tools, libraries, and settings. Docker provides a way to package and distribute applications along with their dependencies, making it easier to deploy and manage software across different environments.
  • Fabric Manager
    It's a software tool developed by NVIDIA that manages the hardware resources and interconnects in NVIDIA GPUs, particularly those using NVLink and SXM architectures. It is essential for ensuring that the high-speed interconnects between GPUs are functioning correctly, which is crucial for applications requiring intense computational power and fast data transfer between GPUs.
  • Terminal
    A terminal is a text-based interface in a computer system used for entering commands and interacting with the operating system or applications. It provides a way to navigate the file system, run programs, manage processes, and perform various tasks using command-line instructions. Terminals are commonly used in Unix-based systems like Linux and macOS, where users can access a terminal window to enter commands directly.
  • NVIDIA driver
    A NVIDIA driver is a software component that allows your computer's operating system to communicate and interact with NVIDIA graphics processing units (GPUs). It acts as a bridge between the hardware and the operating system, facilitating the proper functioning and optimization of NVIDIA GPUs for tasks like graphics rendering, gaming, AI processing, and more.
  • Hiveon OS
    Hiveon OS is an operating system specifically designed for cryptocurrency mining. It is optimized to maximize mining efficiency and profitability by providing features such as easy setup, remote monitoring and management, mining software integration, and performance optimization for various mining rigs.
  • Rosetta 2
    Rosetta 2 is a special software for Apple computers with M1 chips that lets them run apps designed for older Intel-based Macs. It works behind the scenes to translate the app's instructions so they can work on the new hardware, allowing you to use your favorite apps even if they haven't been updated for the new chips.

Security Compliance

  • E2E Encrypted
    It’s a method of secure communication where the data is encrypted on the sender's device, remains encrypted while it's transmitted over a network, and is only decrypted on the recipient's device.
  • SOC2/HIPAA
    SOC 2 and HIPAA are compliance frameworks that address different aspects of data security and privacy. SOC 2 focuses on assessing the controls implemented by service organizations to protect customer data, while HIPAA sets standards for protecting sensitive personal information.

Monitoring Services

  • Ray.io
    Ray is an open-source unified compute framework that makes it easy to scale AI and Python workloads — from reinforcement learning to deep learning to tuning, and model serving.
  • IO Version Control
    IO Version Control refers to a specific version or release of components within the IO.NET platform, including IO Cloud, IO Worker, or IO SDK. Each version includes updates, bug fixes, and enhancements aimed at improving performance, security, and overall user experience.
  • IO Monitor
    IO Monitor is a tool within the IO.NET ecosystem that enables users to oversee the performance, status, and metrics of their computing resources. This includes monitoring real-time data on GPU utilization, computing efficiency, and possibly financial aspects related to usage and earnings from contributing computing power to the network.

Supplier

  • FileCoin
    Is designed specifically for decentralized storage. It's a decentralized storage network that enables users to store and retrieve data in a decentralized manner. Users who have excess storage capacity can become storage providers on the Filecoin network. They can offer their storage space to store files for others. (Kind of what we do with GPUs).
  • Render Network
    The Render Network is a blockchain and crypto-enabled platform where users can contribute their unused GPU power to assist in rendering motion graphics and visual effects for projects. In exchange for their contributions, users receive Render tokens (RNDR), the native utility token of the network.
  • IO Network
    IO Network is a sophisticated networking backend that employs a secured mesh VPN to enable ultra-low latency communication among the IO.NET Miner nodes, also known as "workers."