Find answers to common questions about GPU ownership, datacenter operations, and platform mechanics. Use our chatbot to answer any other questions you may have.
If you need additional support, our team is here to help.
Unlike traditional cloud providers, Silicon Network creates a decentralized marketplace where anyone can own GPU infrastructure. This model reduces costs, increases availability, provides asset liquidity through tokenization, and creates new investment opportunities while maintaining enterprise-grade reliability.
Berkeley Compute is the company developing the Silicon Network. They're building tools that empower GPU operators and owners to participate in the AI revolution through a decentralized infrastructure platform.
Silicon Network is a decentralized, blockchain-enabled AI infrastructure platform democratizes access to GPU computing resources by tokenizing GPU ownership, implementing a novel payment protocol, and creating a flexible marketplace that addresses challenges of cost, availability, and liquidity in the rapidly growing AI compute market.
Silicon Protocol exists to democratize access to AI infrastructure by creating a more equitable GPU compute marketplace. It enables anyone to own a piece of AI infrastructure through tokenized GPU ownership, allows datacenter providers to scale through crowdsourced capital, and gives AI developers access to more affordable and flexible compute resources.
Silicon Protocol is the underlying blockchain protocol that powers the Silicon Network. It enables the tokenization of physical GPU assets, facilitates payments between GPU owners and customers, and provides the technical foundation for decentralized GPU ownership and utilization.
Silicon Network ensures quality and reliability through on-chain reputation systems, performance metrics, and SLA monitoring. Datacenter providers build verifiable reputation metrics through consistent performance, while the platform's quality assurance tools provide transparency to both GPU owners and compute customers.
Silicon Network uses a dual blockchain architecture to leverage the unique strengths of each chain: Base provides efficient capital flows and liquidity for trading, while Chia offers secure, energy-efficient ownership verification and robust smart contract capabilities for managing GPU assets and rewards.
Silicon Network supports a range of AI workloads including model training (typically requiring larger GPU clusters), model fine-tuning (using medium-sized GPU clusters of 1-8 GPUs), and inference workloads (which can run efficiently on 1-2 GPUs). The network is designed to match workloads with the appropriate hardware configuration based on specific requirements.
What is the Base blockchain?
Chia is an energy-efficient blockchain founded by Bram Cohen that uses a proof of space and time consensus mechanism. In the Silicon Network, Chia is utilized for secure ownership verification of GPU assets and for trading the digital twins on a secondary market.
Silicon Protocol works by tokenizing physical GPUs as NFTs on a dual-blockchain architecture, creating a marketplace for these assets, and implementing a burn-mint equilibrium token model. GPU owners earn rewards when their GPUs are utilized, datacenter providers host and manage the hardware, and compute customers rent these resources for AI workloads—all coordinated through blockchain-based smart contracts.
Providers rent out the compute power of GPUs to developers via compute marketplaces or direct rentals. Compute marketplaces are run by brokers like Runpod, TensorDock, Hydrahost, or Akash and allow customers to rent GPUs by the hour. Direct rentals are usually longer term deals where a customer will book a certain block of time (usually in days, weeks, or months) to have exclusive access to their desired amount of GPUs.
The $SIL token is scheduled for mainnet launch on Base and Chia blockchains in Q3 2025, alongside the release of the detailed tokenomics whitepaper.
GPU owners earn rewards automatically based on their GPU's utilization on the network. When customers rent GPU computing power, a portion of those payments is distributed to GPU owners as $SIL tokens. Owners can either receive rewards directly proportional to their specific GPU's usage or opt into resource pools for smoothed, predictable earnings averaged across multiple GPUs.
Data Credits (DCs) are USD-pegged utility tokens used for GPU rentals on the Silicon Network. They provide stable, predictable pricing for compute resources. Customers or payment processors purchase $SIL tokens and burn them to mint Data Credits, which are then used to pay for GPU resources. This mechanism ensures price stability for customers regardless of $SIL token fluctuations.
$SIL is the native utility token of the Silicon Network that serves three essential purposes: as a payment mechanism for transactions within the ecosystem, for reward distribution to GPU owners and operators who contribute resources, and for governance to enable protocol decision-making by token holders.
Like any investment, GPU NFTs carry certain risks, including: fluctuations in demand for AI compute resources, technological obsolescence as newer GPU models emerge, operational risks at datacenters, regulatory changes affecting tokenized real-world assets, and market liquidity risks. However, the Silicon Network mitigates these through transparent performance tracking, professional datacenter management, and a robust marketplace designed to maintain liquidity.
An organization is how an individual would structure payouts to a GPU or set of GPUs that are owned by that individual or group of individuals. Organization owners are the only ones with the ability to claim payments from the provider for the set of GPUs. Owners within an organization will be able to view payout data for the organization.
This can differ depending on your area of residence. We recommend consulting with a tax advisor.
Silicon Network partners with professional datacenter operators who manage enterprise-grade GPU infrastructure. These providers maintain hardware in officially tier-rated facilities with rigorous uptime guarantees, environmental controls, and security protocols. (Specific partner names are not listed in the current documentation.)
The Silicon Marketplace provides comprehensive tools to help investors evaluate GPU opportunities, including: hardware specifications comparison, historical performance data, projected return metrics, datacenter provider reputation scores, and risk assessment information. You can filter options based on your investment goals, risk tolerance, and capital availability to find the most suitable GPU assets for your portfolio.
Yes, GPU NFTs include legally binding ownership contracts as part of their metadata. These contracts establish your ownership rights to the specific physical hardware, while the hosting agreement outlines the terms of datacenter management and your rights regarding the underlying asset.
Ownership of the GPUs is first represented by whoever owns the GPU NFTs. Silicon takes security seriously to ensure the platform is safe, but should the NFTs in the Silicon platform be compromised, please raise an issue with the respective provider that is managing your GPU. There will be required proof of ownership and providers will do their best to find a resolution.
GPU NFTs represent ownership of physical hardware deployed in professional datacenters with specialized cooling, power, and security requirements. While you own the GPU, it remains hosted in the datacenter for optimal performance and security. The hosting agreement associated with your NFT outlines specific conditions under which physical possession might be claimed, typically at the end of the GPU's economic lifecycle or under specific contract termination scenarios. Contact your datacenter provider for details specific to your ownership agreement.
At the current time, you will not be able to withdraw your NFT mint pass until the GPU has been set up in the datacenter. Once the GPU NFT has been converted into a live GPU NFT, you will be able to resell the GPU NFT on the secondary market.
When you sell your GPU on the secondary market, you are selling your ownership NFT on the Chia blockchain through our marketplace for secondary trading. Initial settlement of those offers will be for wrapped USDC on the Chia blockchain in a Chia wallet of your designation. If you would like to get paid out in USDC on Base, the providers will offer a service to automatically settle into USDC on Base that will have a small fee associated with it.
The value of a GPU NFT is determined by several factors: the underlying hardware specifications, historical performance and utilization metrics, projected future earnings based on demand trends, remaining economic lifespan of the GPU, and the reputation of the datacenter provider hosting it. The Silicon Marketplace enables price discovery through open trading of these assets, with comprehensive data analytics to help investors make informed decisions.
Yes, GPU NFTs are fully transferable assets that can be sold on Silicon Network's marketplace. The secondary market provides liquidity for your investment, allowing you to exit your position when desired. Since all performance data and ownership history are transparently recorded on-chain, potential buyers can make informed decisions, creating a fair market value for your asset.
At this time your rewards will be provided as USDC on Base.
Datacenter providers are responsible for maintaining GPU hardware according to the service level agreements attached to your NFT. If a GPU fails, the provider will typically replace or repair it to maintain service. As GPUs approach obsolescence, their market value and earning potential may decrease, but you can choose to sell your NFT on the secondary market at any time. The NFT includes transparent performance history to inform potential buyers about the asset's condition and earning capacity.
Returns from GPU ownership vary based on multiple factors including the GPU model, datacenter location, market demand for compute resources, and overall network utilization. The GPU Marketplace provides historical performance data and yield metrics to help set expectations, but like any investment, returns are not guaranteed. The advantage of the Silicon Network model is the transparency of performance metrics and the option to participate in pools for more predictable earnings.
The provider has custody of the wallet through the Silicon platform to transact into and out of that wallet. So when you see that you have been paid by a provider, the balance you see if in this shared custodied wallet. You can then withdraw those tokens into a non-custodial wallet of your choosing.
Currently, no. Depositing your GPU NFT in the Silicon platform is required to claim rewards. We're developing a more non-custodial experience for future implementation, so stay tuned for updates.
Currently, no. You won't have direct access to the private keys for importing into a non-custodial wallet. However, you can still issue transactions through your account on app.silicon.net.
Earnings from your GPU are automatically distributed to your connected wallet based on your GPU's utilization in the network. The Silicon Network payment system tracks utilization metrics on-chain and distributes newly minted $SIL tokens directly to NFT holders according to either individual GPU performance or pool-based earnings if you've opted into a resource pool. These earnings can be viewed in real-time through the owner dashboard and withdrawn at any time.
After depositing your GPU NFT, the private keys are securely managed by the Silicon platform. Transaction signing access is shared between you and the provider, ensuring joint custody of the asset.
To claim rewards, you must deposit your NFT into the Silicon platform. Once deposited, your NFTs are stored in a custodied wallet where transaction signing access is shared between you (the GPU owner) and the provider.
Depositing your GPU NFT into the Silicon application allows the platform to verify ownership and automatically distribute earnings to you based on your GPU's utilization. The application provides a secure interface to monitor your GPU's performance, track earnings in real-time, and participate in optional pooling for more predictable revenue. This integration ensures seamless reward distribution without requiring any technical management of the physical hardware.
Yes, you can resell a mint pass prior to the GPU being deployed in a datacenter. Listing a mint pass will require you to list it on a Silicon Marketplace or any other Base Ethereum marketplace.
Silicon Network partners with professional datacenter operators who manage enterprise-grade GPU infrastructure. These providers maintain hardware in officially tier-rated facilities with rigorous uptime guarantees, environmental controls, and security protocols. (Specific partner names are not listed in the current documentation.)
In time, anyone can become a provider on the network, as Silicon Network becomes increasingly decentralized. Silicon will continue to play a role in verifying providers.
Silicon Network provides datacenter operators with a comprehensive management platform that includes tools for hardware lifecycle management, GPU DevOps, financial operations, and quality assurance. These tools streamline operations, reduce complexity, and improve efficiency while maintaining the transparency required for decentralized ownership.
Datacenter providers should operate tier-rated facilities with redundant power, cooling, and network connectivity. They need the technical expertise to manage enterprise-grade GPU infrastructure and the ability to integrate with Silicon Network's management platform and marketplace.
Datacenter providers interested in joining the Silicon Network can partner with Berkeley Compute to integrate with the tokenization platform scheduled to launch in Q2 2025. The platform will provide tools for onboarding new GPUs, managing owners, and integrating with compute marketplaces.
To become a community cloud provider, you'll need to have a compatible GPU and internet connection. When the community inference cloud launches in Q3 2025, you'll be able to download specialized software that enables secure workload processing while maintaining local control of your hardware, allowing you to monetize idle computing capacity.
Datacenter providers receive a portion of the rewards generated when GPUs in their facilities are utilized. The payment system automatically allocates newly minted $SIL tokens to datacenter providers based on utilization metrics, creating a sustainable revenue stream that scales with demand.
We currently support NVDIA RTX 4090s, 5090s, A6000, A5000, 4000 Ada, H100s, and H200s through professionally managed data centers. You can get more details about active GPUs on our network at https://app.silicon.net/. Our Community Cloud will allow anyone to contribute their at-home compute power from a larger variety of chipsets. If you have other AI rigs you'd like to add to the network as a Provider, please reach out to us from the Provider page contact form.
Community cloud providers are individual GPU owners who contribute computing power from personal or small-scale GPU deployments. These participants—typically home users, small businesses, or academic institutions—connect their existing GPUs to the Silicon Network to support inference workloads while earning passive income. Community cloud providers will be rewarded $SIL tokens on the Chia blockchain.