Major visions
- NOSANA’s basic offer is a non -central mathematical network based on GPU, Acknowledging the increasing demand for artificial intelligence computing.
- NOSANA provides units of consumer graphics processing to provide reasoning services upon request With cost savings up to 2.5X compared to traditional cloud service providers.
- In 2024, the Nosana 985,000 contract completed the contractWith a significant rise in job applications that occur in May after that on board 1000 new contracts.
- 29.7 million number (31 % of the distinguished symbol supplies) are steeped on Nusana. At the time of writing this report, the value of the distinctive symbols amounted to $ 92.4 million.
- Nosana’s Mainnet is to be launched in the first quarter of 2025, With GPU account account account.
Preliminary
NOSANA is a decentralized physical infrastructure network (DePin) is based on Solana. The basic project of the project is a MPU, which allows individuals and institutions to rent the inactive GPU power to the tasks of artificial intelligence. This approach addresses the increasing demand for mathematical resources in the artificial intelligence industry and allows the owners of the graphics processing unit to earn a negative income. By focusing on the inference of artificial intelligence, NOSANA aims to create a more easy and effective solution from traditional cloud services providers.
The original symbol of the project, NOS, plays a major role in the ecosystem, facilitates transactions, enabling attention to network participation, and in the end governance. Nusana’s strategic axis from CI/cd services (A graceful practice that is automated, tested and published for developers) to infer artificial intelligence, as well as its focus on benefiting from units of consumer consumer graphics processing, and unique situations within the sophisticated scene quickly for decentralized computing and artificial intelligence. With The launch of Mainnet to be held on January 14, 2025NOSANA aims to open the access and ability to afford the costs of artificial intelligence computing resources.
Website / / X (twitter)
Nosa’s strategic axis to infer artificial intelligence
In that Release In 2021, Infant At first, focus on providing decentralized automation services for program developers through CI/cd services. The project aims to create a decentralized market where developers can access account resources from a network distributed to device providers. However, after a year and a half of development, the team faced challenges with traction acquisition. As noted in Blog postMost developers in deportation hesitated to new tools, even if they offered decentralized alternatives.
This perception coincided with rapid innovation and growth in artificial intelligence after the Chatgpt version in late 2022. The demand for the mathematical resources required to operate the artificial intelligence models increased. In Q4 2023, Nosana axis Towards building a decentralized mathematical network that works at the GPU strictly to infer artificial intelligence.
Several factors pushed this axis. First, the team realized that the demand for the inference of artificial intelligence was rapidly exceeding the width of the available graphics processing units, which leads to unacceptable pricing. Second, they saw that many of the consumer graphics processing units were not used, which represents a vast unusual resource that could be harnessed to meet this request. Third, they believed that their decentralized approach could provide a more cost -effective and developmental solution than traditional cloud service providers. By focusing on artificial intelligence inference, NOSANA aims to create a market where anyone with a compatible graphics processing unit can contribute to the authorization of lethargy and gain rewards. At the same time, artificial intelligence developers and researchers can reach the resources they need in a small part of the cost. Nusana research He explains that rental of consumer graphics processing units such as RTX 4090 can provide inferences at a price of 2.5X than the normative A100 graphics processing unit while maintaining the same return on investment.
Deep diving in the conclusion of artificial intelligence on Nusana
The primary goal of the decentralized NOSANA account is to provide more conclusions for every dollar of central cloud services. Inference Indicate To use the trained artificial intelligence model to create predictions or create outputs based on new input data. For example, when ChatGPT is required to summarize an article, the Openai model processes the demand for servers in a data center, expanding resources as needed. The process ends as soon as the summary is returned, and the progress of the inference work. By focusing on inference, nosana can improve its network for the requirements specified for this type of work burden, such as low cumin and high productivity.
Nusana constantly Work About 300 job operating contracts with various GPU and AI models. RTX 4090, the famous consumer graphics processing unit with 24 GB of VRAMIt provides energy for $ 0.32 per hour to run the Llama 3.1 8B. 8-32 GB of VRAM is usually sufficient for small to medium-sized tasks. It can match a few RTX 4090s with an industrial class math A100 GPU– It is well used for the largest reasoning tasks – which cost more than $ 1.29 per hour rent.
NOSANA contract is designed, that is, programs that link graphics processing units to the decentralized NOSANA account, to be very available. the Device requirements To become a knot (recently Rename To the “hosts”), it includes the operation of any of the 19 -backed NVIDIA RTX models, at least of 4 GB RAM, and internet connection. By focusing on consumer devices, nosana provides individuals who have games or other devices with inactive graphics processing units to join the network and gain rewards. NOSANA is classified as devices into specific algorithms or models. The waiting system within each market guarantees the allocation of fair and effective resources, improves work burdens, and provides a dedicated experience for GPU service providers and consumers.
Focus on smaller open source models, such as Lama and StableIt provides more transparency and cooperative opportunity Reasonable pricesAnd if it is sometimes at the expense of performance. The closed models achieved a Performance feature 24 %According to the Stanford Institute of Managed AI (HaI). However, in the case of the NOSANA account designed for Small to medium -sized conclusionsThe smaller and open models may be the source Faster conclusion times at a reasonable performance level. This strategy improves price and efficiency.
In 2024, Nusana’s contract seized 985,000 jobs. The total running function includes, waiting menu, complete, and stopping jobs. The jobs may not be completed for various reasons, including the expiration of the job, or the node is disrupted, or the contract that rejects a job while operating.
In April, Nusana Stage 2 launched From their test network, add more than 1,000 new contracts to the network. The growth of the node coincided with an increase of 1,667 % in job requests in the following month and a noticeable increase in requests each month for the rest of 2024. Stage 3 It started in September and entered the node, a new pricing system to treat encryption fluctuations, and improvements to tools such as Explorerand Node LeaderboardAnd Nusana Kleal.
Partnerships and the growth of the ecosystem
NOSANA has created partnerships to expand the decentralized GPU in various fields, including artificial intelligence development, cloud infrastructure, and decentralized applications:
- One matrix: The Nosana’s GPU is used to support the creation of the Acting Avatar.
- picnicIt provides cloud infrastructure at the level of the institution, including NVIDIA A5000 Settings, to enhance the GPU market in Nosana.
- SOGNI.AIThe GPU enhances NOSANA for intense tasks in accordance with a stable proliferation model.
- my dear: GPU market is used in NOSANA to verify the validity of the cost -effective machine learning.
- theoryUsing the GPU market in NOSANA to test and improve artificial intelligence agents with advanced models settings.
- Display: Integrated with NOSANA in January after approval an offerEnably submit the contract to implement Nosana and Nosana jobs to spread jobs on the display network.
- OCA: GPU is used in NOSANA to support artificial intelligence agents, reduce costs, improve performance, overcome restrictions, and enable more operational flexibility.
- Alpha nervous intelligence artificial intelligence: It enhances the GPU NOSANA network to spread, train and host artificial intelligence models.
- HDSP research collection: NOSANA’s GPU is used to accelerate AI and ML academic research.
NOSANA partnerships shows its GPU’s ingenuity in dealing with various mathematical tasks, which range from stable proliferation to the symbolic symbolism. Cooperation with protocols such as Piknik and the Render network can enhance GPU supplies and the availability of node as Nosana evaluates its account market.
Distinguished code and NOS code
No., and Spl Code with a The maximum supply of the distinctive symbol of 100 millionIt is in the heart of the Nusana network. The symbol settles network transactions fees, pays for inferring tasks, and supports network safety through residence. Once the main network is launched in the first quarter of 2025, Nos Token holders will have the rights to vote on proposals and influence the future direction of the project.
The node operators should share NOS as a guarantee. The attention mechanism stimulates this sincere behavior, as malignant actors risk the loss of their exciting symbols through the reduction. As of January 2, 2025, 29.7 million NOS has been subjected to self With more than 14100 wallets, equivalent to $ 92.4 million. From the circulating offer, 31 % of the distinctive symbols.
The initial distribution of NOS, which happened January 17, 2022The symbols are allocated to the various stakeholders, including the team, the company, supporters and society. It is worth noting that 62 % of the symbols were allocated to the informed (i.e. the team, the company and the supporters), taking into account the various timelines. The allocation of the team rises in written 48 months, while the company’s allocation is more than 36 months. Supporters have finished nine months of launching the distinctive symbol. These timetables are compatible with the incentives of the team and the first supporters with the success of the project in the long run.
Stokeing is an important component of the Nosana network, as the knot operators stimulate to provide reliable and high -quality mathematical resources. Through Stoke NOS, the node operators refer to their commitment to the network and become eligible to receive rewards for their contributions.
The team implemented a Dynamic dairy bonus system To ensure the long -term network sustainability. Initially, Staking APY presented a relatively high program to attract the first adopters and pave the network. However, APY will gradually reduce the network maturity and join more nodes. This adjustment is necessary to prevent excessive inflation and maintain a healthy balance between supply and demand for NOS icon.
The team plans to present Rewards based on transactions In addition to exciting rewards. With the growth of the network and more inference functions are treated from artificial intelligence, some transactions will be distributed to Stakers. This mechanism links the use of the network and Stakeer, which motivates the participants to contribute to the growth and adoption of the platform.
Summary of the closure
The transformation of NOSANA towards the conclusion of artificial intelligence, supported by a centralized network of consumer graphics processing units, puts it as a potential competitor in the growing market for artificial intelligence. NOSANA partnerships display the ingenuity of GPU and its ability to deal with different job requests. the The next Mainnet launch in January 2025 He will explain the full capabilities of the network account.
While the decentralized account market displays high barriers to entry and competition from players such as Akashand i.netand He goesAnd MakeNOSANA will rely on its ability to capture specialized specialized markets. This includes providing food to the ecosystems of specific developers and small to medium inference tasks that seek to obtain costly effective solutions.
Moreover, by exploring decentralization, Nosana can address a decisive pain point in the artificial intelligence pipeline – confidence and verification – which provides transparency and reliability enhanced to its users. By focusing on neglected areas of improvement, such as decentralized reasoning, NOSANA has the ability to put a unique position in the account market and contribute usely to the advanced ecosystem.
Let’s know what I loved in the report, or what may be missing, or shares any other notes Fill this short model. All responses are subject to us privacy policy and Conditions of service.
