Multicoin Partner: Understand the GPU Internet io.net

Press: On March 6, 2024, Solana ecological DePIN protocol io.net announced the completion of a US$30 million Series A financing. io.net stated that the funds raised will be used to build the world’s largest decentralized GPU network And to solve the problem of AI computing shortage, Multicoin Capital participated in the investment. Multicoin Capital partner Shayon Sengupta wrote about why he invested in io.net. Golden Finance xiaozou translation.

We are pleased to announce our investment in io.net, the leading distributed computing marketplace for AI (artificial intelligence) workloads. We led its seed round and participated in its Series A financing. At present, io.net has successfully raised US$30 million from Multicoin, Hack VC, 6th Man Ventures, Modular Capital and a well-connected large angel investor to create an on-demand computing market.

My first meeting with io.net founder Ahmad Shadid was at the Austin Solana Hacker House (Solana Hacker House) in April 2023. His focus was on computing resource access for ML workloads I was immediately attracted to this aspect of democratization.

Since then, the io.net team has been executing on this concept at light speed. Today, the network has brought together tens of thousands of distributed GPUs, providing more than 57,000 computing hours to AI enterprises. We are excited to partner with them as they drive the next decade of AI renaissance.

1, Global computing shortage

The computing demand for artificial intelligence is growing at an alarming rate, and the demand cannot be met. AI workload data center revenue exceeds $100 billion in 2023, and even under the most conservative scenario, AI demand exceeds chip supply.

In a time of high interest rates and capital scarcity, new data centers capable of housing this type of hardware require significant upfront investment. The crux of the matter is that advanced chips such as NVidia A100 and H100 have production limitations. Although GPU performance continues to improve and costs steadily decrease, the physical manufacturing process is not accelerating fast enough, and shortages of raw materials, components, and production capabilities limit the pace of development.

While AI holds great promise, its ever-expanding physical footprint and demands for space, power, and cutting-edge equipment are straining budgets around the world. io.net paves the way for us to create a world where the pace of development is not limited by current supply chains.

io.net is a classic example of DePIN theory: using token incentive mechanisms to structurally reduce the cost of acquiring and retaining supply-side resources, and ultimately reducing the cost to the final consumer. The network brings together a large number of heterogeneous GPUs into a shared pool for use by AI developers and companies. Today, the network includes thousands of GPUs from data centers, mining farms, and consumer devices.

While this aggregation of resources is valuable, AI workloads will not automatically scale from centralized enterprise-grade hardware to distributed networks. In the history of cryptography, there have been several attempts to build distributed computing networks, most of which failed to produce meaningful demand-side magnitude.

Coordinating and scheduling workloads across heterogeneous hardware with different memory, bandwidth, and storage configurations is no easy task. We believe the io.net team has the most practical solution on the market today to make this hardware aggregation beneficial to end customers and cost-effective.

2, Paving the way forward for clusters

In the history of computing, software frameworks and design patterns have evolved around the hardware configurations available on the market. shaped. Most frameworks and libraries for AI development rely heavily on centralized hardware resources, but the past decade has seen significant progress in distributing workloads across geographically distributed hardware.

io.net leverages the world's underlying hardware, deploys custom networking and orchestration layers on top of it, and brings them online, creating a super-scalable GPU Internet. The network leverages Ray, Ludwig, Kubernetes, and various other open source distributed computing frameworks to allow machine learning engineering and operations teams to scale their workloads with only minor adjustments on the GPU network.

ML teams can parallelize workloads on io.net GPUs by launching clusters on demand and leverage these libraries to handle coordination, scheduling, fault tolerance, and scaling. For example, if a group of motion graphics designers contribute their home GPUs to the network, io.net can build a cluster that gives image diffusion model developers anywhere in the world access to collective computing resources.

BC8.ai is a fine-tuned variant of Stable Diffusion, trained entirely on io.net hardware, which is a good example. The io.net browser displays real-time inference and payments to network contributors.

Each inference is recorded on-chain to provide traceability. This special image generation is completed by 6 consumer-grade gaming GPU RTX 4090.

Today, there are thousands of devices on the network, spanning mining farms, underutilized data centers, and Render network consumer nodes. In addition to creating new GPU supply, io.net is able to compete on cost with traditional cloud providers by often offering cheaper resources.

They achieve cost savings by outsourcing GPU coordination and overhead to the protocol. Cloud providers, on the other hand, put a markup on infrastructure costs because they also have staff costs, hardware maintenance fees, and data center overhead. The opportunity cost of consumer clusters and mining farms is significantly lower than the cost acceptable to hyperscale entities, so there is structural arbitrage, and real-time pricing of resources on io.net is lower than the ever-increasing cloud rates.

3, Building GPUInternet

Io.net has a unique advantage, that is, maintaining light assets and integrating Reduce the marginal cost of serving any given customer to nearly zero while establishing direct relationships with both the supply and demand sides of the market. They are well-positioned to serve the thousands of new businesses that need to use GPUs to build competitive products that everyone will interact with in the future.

We are excited to partner with Ahmad and his team as they build and accelerate the advancement of artificial intelligence around the world. If you are building a compute-intensive application, you can register to access io.net's resources. If you have idle GPUs, you can also contribute them to the network to get reward points.