Liquid Stake with compassSOL for an 7.50% APY from staking, MEV + fees

Enjoy the freedom of liquid staking in Solana Defi while delegating your stake to the high performance Solana Compass validator. Stake or unstake at any time here, or with a Jupiter swap.

Benefit from our high staking returns and over 2 years experience operating a Solana validator, and receive additional yield from priority fees + MEV tips

Earn 6.8% APY staking with Solana Compass

Help decentralize and secure the Solana network delegating your stake to us and earn an impressive 6.8% APY yield on your SOL, while supporting us to create new guides and tools. Learn more

Stake your SOL

  1. Click to connect your wallet
  2. Enter the amount you wish to stake
  3. Kick back and enjoy your returns
  4. Unstake from your wallet or our staking dashboard

Earn 6.8% APY staking with Solana Compass

Help decentralize and secure the Solana network delegating your stake to us and earn an impressive 6.8% APY yield on your SOL, while supporting us to create new guides and tools.

Learn more

Breakpoint 2023: IONET's Vision for the Largest AI Compute Cloud on Solana

Published on 2023-11-09

IONET reveals its plan to create a massive, cost-effective AI compute cloud by harnessing underutilized GPUs worldwide.

The notes below are AI generated and may not be 100% accurate. Watch the video to be sure!


At Breakpoint 2023, IONET unveiled an ambitious project to aggregate GPUs from various sources to create the world's largest and most cost-effective AI compute cloud. Tori Green, the COO of IONET, started the presentation by highlighting the significance of GPUs as a valuable resource central to AI advancement. She pointed out that the AI industry experiences a significant GPU shortfall due to skyrocketing demand, which impedes AI development. By leveraging underutilized GPUs from data centers, crypto mining farms, and personal devices, IONET plans to address this shortage and drive down computing costs. Angela Lee, the Head of Business Development at IONET, followed up with a live demo showcasing the ease of deploying and managing clusters on the network, marking a potential revolution for AI startups and enterprises alike.

Key Points:

The Rising Value of GPUs and AI Challenges

GPUs have become a cornerstone for advancing AI, with their use in machine and deep learning models making them a highly sought-after commodity. Tori Green argued that GPUs could be considered more valuable than traditional assets like gold or Bitcoin, given their central role in powering AI. Furthermore, she shed light on the intense computational needs of AI development, which are rapidly increasing, leading to inflated costs that can burden companies, especially startups. IONET aims to alleviate these challenges by creating a vast network of accessible and affordable GPUs to support the growing demand.

IONET's Solution and Offerings

IONET plans to harness a wide array of underutilized GPUs and integrate them into a single network, thus revolutionizing the AI compute market. Users will benefit from a scalable cloud offering immense computing power, flexibility, and speed at up to 90% cheaper costs than traditional cloud services. Suppliers or "workers" in the IONET network will also see substantial advantages, such as increased profitability and full utilization of their GPUs. The platform emphasizes ease of use and quick onboarding, with the aim to attract workers and users alike.

The Market Impact and Future Aspirations

The presentation highlighted the market's growing interest in IONET's model and how it stands apart with its unique clustering capability, which is crucial for running complex machine learning models. Tori Green's vision to place a million GPUs on Solana points to IONET's commitment to building a superior, more affordable experience for users that also holds profitability for suppliers. The numerous partnerships, incentives, and expansion programs revealed alongside the platform indicate a robust strategy to attract a wide and diverse community, looking to make a significant mark on both the AI and blockchain landscapes.

Facts + Figures

  • GPUs are considered the most valuable asset by IONET as they are essential for the development of AI technologies.
  • The fastest app to reach 100 million users was an AI app, doing so in just two months.
  • Compute requirements for AI apps are doubling roughly every 3.4 months.
  • Over the past decade, the cost of training an AI model has increased 31 times per year.
  • The CEO of NVIDIA estimated inferencing to be 80 to 90 percent of a model's total lifetime cost.
  • Demand for GPUs is around 20 to 25 exaflops, while the supply is only 10 exaflops.
  • Millions of underutilized GPUs exist across independent data centers, crypto mining farms, and consumer hands.
  • IONET aims to deploy 1 million GPUs on the Solana blockchain.
  • IONET projects significant cost savings for users and profit increases for suppliers by a factor of 10.
  • Clustering, a significant feature of IONET's offering, allows combining multiple machines to act as a potent single supercomputer.
  • Existing suppliers incentive program with Render offers up to 300,000 tokens for early participants, with more through 2024.
  • IONET ventures to supply $500,000 worth of compute power to AI startups.
  • Native tokens, iOcoin and IOSD, are slated for release next year.
  • BCA, a generative art platform, is a demonstration of IONET's capabilities in scaling AI compute power.

Top quotes

  • "We think it's GPUs. The CEO of Intel said that GPUs are the new oil."
  • "The rise of AI has been unprecedented."
  • "This transformative power comes at an enormous computing cost."
  • "We're going to put 1 million GPUs on Solana to build the world's largest AI compute cloud."
  • "Without clustering, you simply cannot be competitive in AI."
  • "We are on a mission to deploy over a million GPUs through our network."
  • "You are just 90 seconds away from deploying a 20,000 GPU cluster."
  • "Our network is also a really simple and quick process."

Questions Answered

Why are GPUs so significant to AI development?

GPUs are central to AI because they are capable of processing the complex data and algorithms necessary for machine and deep learning models. IONET identifies GPUs as an essential asset, even more valuable than traditional resources like gold or Bitcoin, due to their impact on technological progress.

What problems is IONET aiming to solve in the AI industry?

IONET is addressing the GPU shortage in the cloud and the disproportionate costs associated with AI development. Their solution intends to reduce wait times, expand choice, and drastically cut costs for both developers and small startups, many of whom spend exorbitant amounts for AI compute.

How will IONET change the current landscape of AI compute?

By bringing together underutilized GPUs into a consolidated network and offering clustering capabilities, IONET proposes to make powerful AI computing resources available at a much lower cost. This could democratize access to AI development tools and potentially spur a new wave of innovation in the field.

What unique benefits does IONET offer to users and suppliers?

Users can access massive compute power with flexibility, speed, and a fraction of the cost compared to traditional cloud services. Suppliers, on the other hand, can significantly increase their profits and utilization rates by joining IONET's network.

How will the IONET project benefit the broader market and society?

By addressing the GPU shortage and providing affordable AI compute resources, IONET has the potential to accelerate the pace of AI innovation. This democratization of AI compute power could enable startups and enterprises to solve complex problems and develop new technologies that can contribute to societal progress.

Related Content

Breakpoint 2023: Social Media on Solana

An in-depth look at the evolving landscape of social media on the Solana blockchain from the perspectives of key industry players.

Breakpoint 2023: Winners of the Hyperdrive Hackathon

The Solana Foundation announces the winners of the Hyperdrive Hackathon, highlighting innovative blockchain projects.

Breakpoint 2023: Star Atlas Session

A visionary presentation on Star Atlas's intersection of gaming and blockchain on the Solana platform.

Breakpoint 2023: Widening the Design Space of AMMs with Solana

Joe Corey discusses innovative mechanisms for AMMs leveraging Solana’s high-performance blockchain

Breakpoint 2023: Simulation of Transaction Limitation

Exploring the intricacies and future developments in the field of cryptocurrency transaction simulations and limitations

Breakpoint 2023: NFT Past & The Future

Max Zhuang, CEO of Sniper Labs, discusses the evolution of NFTs and Sniper's role in the growing market.

Breakpoint 2023: Web3 Music

A panel discussion on the future and current state of Web3 music with industry pioneers and an independent artist.

Breakpoint 2023: Creator Economy on Solana

Exploring the rising creator economy on Solana with a focus on on-chain monetization and relationships.

Breakpoint 2023: Gaming in Web3 Panel

Leaders in the Web3 gaming space discuss the challenges and opportunities within the industry.

Breakpoint 2023: School of Solana by Ackee Blockchain

Ackee Blockchain's latest initiative, the School of Solana, offers free, online blockchain developer education.

Breakpoint 2023: tBTC comes to Solana

Discussions on the integration of tBTC, a decentralized Bitcoin, into the Solana ecosystem.

Breakpoint 2023: Breaking Down the Current State of Crypto VC / Funding

Insights into the evolving strategies and considerations for VC funding in the cryptocurrency space

Breakpoint 2023: Building a Creator Community

Industry experts discuss empowerment and innovation in the NFT creator community.

Breakpoint 2023: DRiP on Solana

A discussion on the evolution and success of DRiP, a Web3 platform on Solana, and its impact on crypto engagement and creator monetization.

Breakpoint 2023: Intersection of Maps, Geospatial and Web3

Akshay Yeleswarapu discusses the integration of maps and geospatial data with Web3 technology.