In a significant development for the intersection of artificial intelligence and blockchain technology, io.net has announced a $30 million Series A funding round to expand its decentralized GPU compute network. The raise, led by Hack VC with participation from Multicoin Capital, 6th Man Ventures, Delphi Digital, Solana Labs, and Animoca Brands, positions io.net at the forefront of the DePIN movement as AI compute demand reaches unprecedented levels.
The Synergy
The convergence of artificial intelligence and decentralized infrastructure represents one of the most compelling narratives in the crypto space in 2024. As large language models and generative AI applications consume exponentially more compute resources, the centralized cloud providers — primarily AWS, Google Cloud, and Microsoft Azure — have struggled to keep pace with demand. GPU shortages have become a global bottleneck, with Nvidia’s H100 chips commanding premium prices and months-long wait times.
io.net addresses this gap by building what it calls the Internet of GPUs — a decentralized physical infrastructure network, or DePIN, that aggregates underutilized GPU resources from independent data centers, crypto miners, and consumer hardware into scalable virtual clusters. The network has already grown to over 25,000 GPUs and processed more than 40,000 compute hours for AI and machine learning companies.
AI Use Cases in Web3
The io.net platform enables several critical AI applications within the Web3 ecosystem. AI model training, which requires enormous computational resources, can now be distributed across a global network of GPU providers at significantly reduced costs. Inference workloads for applications like image generation and natural language processing benefit from the network’s low-latency architecture. Companies like Krea.ai are already using io.net’s IO Cloud platform to power their AI model inference pipelines.
Beyond raw compute, the DePIN model creates new economic opportunities for GPU owners worldwide. Crypto miners facing declining profitability can repurpose their hardware to serve AI workloads, earning revenue through the network’s tokenomic incentive structure. This transforms what was previously a single-use asset — mining hardware — into a flexible compute resource with multiple revenue streams.
Data Privacy Implications
The decentralized nature of io.net raises important questions about data privacy and security in distributed computing. When AI workloads are processed across geo-distributed nodes, ensuring data sovereignty and compliance with regulations like GDPR becomes more complex. io.net addresses this through its enterprise-grade security layer, which allows developers to specify cluster requirements including GPU types, host locations, and security levels.
The permissionless architecture of the network means that compute resources can be sourced from virtually anywhere, which is both a strength and a challenge. While it dramatically reduces costs — io.net claims savings of up to 90 percent compared to traditional cloud providers — it requires robust verification mechanisms to ensure that compute results are accurate and that sensitive data is not exposed to unauthorized parties.
The Innovation Frontier
io.net’s partnerships with Render Network and Filecoin signal a broader trend toward composable decentralized infrastructure. By combining GPU compute from io.net with storage from Filecoin and rendering capabilities from Render Network, developers can build entirely decentralized AI pipelines that are not dependent on any single centralized provider. This vision of composable DePIN infrastructure could fundamentally reshape how AI companies provision and consume compute resources.
The timing of this raise is notable. With Bitcoin trading at approximately $63,800 and the crypto market in full bull mode, investor appetite for infrastructure projects with real utility is strong. The participation of Solana founder Anatoly Yakovenko and Aptos founders Mo Shaikh and Avery Ching as individual investors underscores the cross-ecosystem importance of decentralized compute solutions.
Concluding Thoughts
The $30 million Series A for io.net represents more than just another crypto funding round — it is a bet on the future of AI compute infrastructure. As the demand for GPU resources continues to outstrip supply, decentralized networks like io.net offer a credible alternative to the centralized cloud monopoly. With over 25,000 GPUs already onboarded and major partnerships in place, the network is well-positioned to capture a significant share of the growing AI compute market.
Disclaimer: This article is for informational purposes only and does not constitute financial or investment advice. Always conduct your own research before engaging with any crypto project.
30M Series A with Multicoin and Delphi participating tells you the DePIN narrative had real backing. question is whether the network does anything Filecoin didnt
25k GPUs already online is nothing to sneeze at. if they can actually undercut AWS on price this could get real traction fast
the question is whether those 25k GPUs are actually being used or just sitting idle collecting token rewards. seen this movie before with Filecoin
depin_skeptic_ filecoin had massive capacity too and most of it sat idle. utilization rate is the only metric that matters for dePIN projects
depin_skeptic_ good question. filecoin had the same “massive network” narrative and most of it sat idle. need to see actual utilization metrics
@cloud_escape_ Filecoin comparison is spot on. massive network on paper but utilization was always the missing piece. need to see actual job completion rates not just node counts
25k GPUs sounds impressive but how many are A100s vs random gaming cards. the GPU quality matters more than quantity
Dmitri V. quality vs quantity is the right question. 25k GPUs sounds great in a press release but A100 count is what actually matters for enterprise contracts
checked their dashboard back when it was public. majority were consumer grade rtx cards. fine for inference, useless for training large models
@gpu_audit mostly 3090s confirms the inference-only reality. marketing says AI training cluster, reality says distributed gaming rig
gpu_audit RTX cards for inference is fine but the pitch was always AI training. training an LLM on consumer GPUs is a non-starter
h100_truther_ consumer GPUs for inference is a real use case though. not every workload needs A100s. the misleading part is io.net implying they can handle training
Multicoin and Delphi backing it is a solid signal. those two dont usually touch anything without a real revenue thesis
Multicoin and Delphi backing tells you the VC thesis is real. whether the token captures value from the compute network is a different question entirely
multicoin backing a depin project after their filecoin position makes sense strategically. they know this thesis better than most
25K GPUs sounds great until you check the dashboard and its mostly RTX 3090s. fine for inference workloads but nobody is training an LLM on consumer cards
Mihail P. exactly. the pitch deck says AI training but the actual hardware says inference at best. still a real business but the messaging is misleading
Mihail P. checked the dashboard recently? they added H100s and A100s after the funding round. the consumer card era is winding down
ran a node briefly in 2024, my 4090 got maybe 3 days of bookings a month. demand never matched that 25K supply headline. the consumer GPU crowd was subsidizing the dashboard numbers
three booking days a month matches what render farms were paying for 3090s in 2023. idle supply was never the bottleneck, matching it to actual jobs is
30M raise sounds big until you compare to CoreWeave at 7B. io.net is building the decentralized version of a rounding error
CoreWeave raised 7B because it buys its own GPUs and signs multi year contracts. io.net aggregates idle machines, the capex is other peoples money. different model, 30M is enough to keep a marketplace alive
other peoples capex is the entire point tho. coreweave has to repay that 7B, io.net rents machines gamers already own. whether the demand shows up is the real question
hack vc leading with solana labs and animoca in the round says narrative bet on depin more than a compute bet. still waiting on a public utilization dashboard post raise
^ still no dashboard months after the raise, thats the answer. good utilization numbers get published, 3 booking days a month per gpu gets you a funding announcement instead