lium.io
#51ComputePeer-to-peer GPU marketplace with 500+ H100 GPUs at up to 90% cheaper than traditional cloud.
Price
24h
7d Change
Market Cap
Emission
Miners
Validators
Competitor Mapping
lium.io
Decentralized
Market Cap
$55.04M
TAM
$7.30B
Model
Token-incentivized
Team
Decentralized miners
CoreWeave + Lambda Labs
Private
Market Cap
Private
TAM
$7.30B
Model
Revenue-driven
Team
Enterprise
Competes with CoreWeave ($35B valuation) and Lambda Labs. Decentralized supply means no capex, but SLAs are harder to guarantee.
Implied Valuation
If lium.io captures a share of CoreWeave + Lambda Labs's $7.30B TAM
Bear - 0.1% TAM
$7.30M
implied valuation
vs $55.04M current
Base - 1% TAM
$73.00M
implied valuation
vs $55.04M current
Bull - 5% TAM
$365.00M
implied valuation
vs $55.04M current
Scenarios illustrate potential scale relative to the traditional market. Actual outcomes depend on adoption, tokenomics, and competitive dynamics.
Thesis
Rental revenue exceeding TAO emission subsidies - critical milestone proving organic demand. 90% cost advantage and permissionless supply create a flywheel centralized providers can't match.
Team
Founded by Pierre 'Fish' of Datura AI. Also connected to Desearch (SN22) team. Fish has extensive experience in the Bittensor ecosystem and large-scale GPU deployments. Team operates under the Datura AI umbrella.
- Pierre (Fish) - Founder - Bittensor ecosystem veteran, runs Datura AI which also operates SN22 Desearch
- Giga - Co-founder (Datura AI) - Also involved in Desearch SN22
- Sally - Team member (Datura AI)
Funding
No traditional VC funding disclosed. Funded through TAO emissions (6-7% of network).
Traction
Onboarded ~500 H100 GPUs within first month of launch. Top-emitting subnet (6-7% of TAO emissions). Became one of the most staked subnets by 2025. No KYC permissionless marketplace live and operational.
Recent News
- 2026 - Rebranded from Celium to Lium
- 2026 - Continued growth as one of top-emitting subnets on Bittensor
- Late 2024 - Launched, onboarded ~500 H100 GPUs in first month
Risk
Enterprise customers require SLAs and guaranteed uptime that decentralized GPU pools struggle to deliver.