![]() |
|
![]() |
|
![]() |
|
![]() |
|
![]() |
|
![]() |
|
![]() |
|
![]() |
|
![]() |
|
![]() |
|
![]() |
|
![]() |
|
![]() |
|
![]() |
|
![]() |
|
科技企业想要什么? GPU计算。他们什么时候想要的?现在。他们想要多少?你得到的一切。他们愿意付多少钱?
Tech companies are increasingly turning away from centralized cloud providers in favor of decentralized AI and web3 networks that can meet their needs. They’re not doing it for the decentralization – they're doing it for the significant cost savings it can yield.
科技公司越来越多地摆脱集中式云提供商,转而支持分散的AI和Web3网络,这些网络可以满足其需求。他们不是为了权力下放而进行的 - 他们是为了节省大量成本而进行的。
Traditional cloud providers like AWS and Google Cloud still dominate the market, but their market share is being nibbled away at by decentralized GPU networks (DePINs).
像AWS和Google Cloud这样的传统云提供商仍在市场上占主导地位,但是它们的市场份额被分散的GPU网络(DEXINS)拒之门外。
These DePINs are transforming how AI models are trained by delivering on-demand compute with the following five leading the charge.
这些depins正在通过交付按需计算以以下五个领导电荷来改变AI模型的训练。
1. Render Network
1。渲染网络
One of the most notable players in the space is Render Network, originally designed for decentralized GPU rendering but now expanding into AI compute. Render Network has built a mature ecosystem, connecting thousands of GPU providers with artists, developers, and data scientists. Its existing pay-as-you-go model, powered by the RENDER token, makes GPU access seamless and affordable.
该空间中最著名的玩家之一是渲染网络,该网络最初是为分散的GPU渲染而设计的,但现在扩展到AI Compute。渲染网络已经建立了一个成熟的生态系统,将成千上万的GPU提供商与艺术家,开发人员和数据科学家联系起来。由渲染令牌提供支持的现有付款方式,使GPU访问无缝且负担得起。
The network is actively expanding its capabilities to support AI inference and model training, and is betting big on AI agents – a major web3 growth area where it’s eager to provide the GPUs these autonomous bots require. Recent announcements emphasize AI resource allocation and developer-friendly APIs, indicating a deeper push into onchain AI compute.
该网络正在积极扩展其支持AI推理和模型培训的能力,并在AI代理上大放异彩 - AI代理 - 一个主要的Web3增长领域,它渴望提供这些自动机器人所需的GPU。最近的公告强调了AI资源分配和对开发人员友好的API,这表明更深入地推进了OnChain AI Compute。
2. io.net
2。io.net
One of the best known DePINS operating in the space, io.net aggregates GPU power from data centers and individual contributors across 130 countries to create an on-demand decentralized GPU cloud. io.net significantly reduces costs, offering AI compute at 90% less per TFLOP than centralized alternatives, while ensuring rapid deployment of high-performance clusters, including NVIDIA H100s, in under two minutes.
IO.NET是该空间中运营的最著名的Depins之一,它来自数据中心和130个国家 /地区的个人贡献者的GPU功率,以创建一个按需分散的GPU云。 IO.NET大大降低了成本,每TFLOP的AI计算比集中式替代方案降低了90%,同时确保在不到两分钟内快速部署了包括NVIDIA H100的高性能群集。
The platform leverages Solana’s blockchain to provide transparent transactions and features a unique Proof of Time-Lock mechanism that guarantees dedicated GPU resources. With growing enterprise adoption, expanded hardware support for models like 4090s, A100s, and H100s, and a scalable cluster rental system, io.net has become a leading choice for AI-powered organizations seeking high-performance compute.
该平台利用Solana的区块链提供透明的交易,并具有独特的时锁机制证明,可以保证专用的GPU资源。随着企业采用的不断增长,对4090,A100和H100等型号的扩展硬件支持以及可扩展的集群租赁系统,io.net已成为寻求高性能计算的AI驱动组织的领先选择。
3. Hypercycle
3。超循环
For developers seeking web3-native AI inference, Hypercycle is the DePIN they fire up. Unlike traditional decentralized GPU platforms that focus on raw compute power, Hypercycle specializes in AI model inference, allowing dapps to execute machine learning tasks without centralized bottlenecks. The protocol integrates micro-payments for inference tasks, ensuring a seamless and cost-effective billing structure. Hypercycle’s architecture is optimized for low-latency AI execution, making it a valuable resource for real-time applications.
对于寻求Web3本地AI推理的开发人员而言,HyperCrecle是他们启动的DEPIN。与关注原始计算功率的传统分散的GPU平台不同,HyperCycle专门研究AI模型推断,允许DAPP执行机器学习任务而无需集中瓶颈。该协议集成了用于推理任务的微付款,从而确保了无缝且具有成本效益的计费结构。 Hypercycle的体系结构已针对低延迟AI执行进行了优化,这使其成为实时应用程序的宝贵资源。
Hypercycle is actively forging partnerships with AI-driven dapps and continues to refine its layered AI solutions to enhance speed and efficiency. Like several of the other companies profiled here, Hypercycle is long AI agents and believes it can capture significant market share here.
Hypercycle正在积极建立与AI驱动DAPP的合作伙伴关系,并继续完善其分层AI解决方案以提高速度和效率。像其他几家公司一样,Hypercycle是很长的AI代理商,并认为它可以在这里占据很大的市场份额。
4. Akash Network
4。Akash网络
Akash Network serves as a decentralized cloud marketplace catering to a blend of AI and web3 developers. Its auction-based pricing model ensures that developers get access to computing resources at competitive rates, while its decentralized matching system allows for instant provisioning of GPU instances. Akash has steadily expanded its support for AI training and inference workloads, onboarding new infrastructure providers to meet the growing demands of decentralized applications and machine learning research.
Akash网络是一个分散的云市场,可满足AI和Web3开发人员的融合。其基于拍卖的定价模型可确保开发人员以有竞争力的速度访问计算资源,而其去中心化匹配系统可以立即提供GPU实例。阿卡什(Akash)稳步扩大了对AI培训和推理工作量的支持,攻入新的基础设施提供商,以满足分散应用程序和机器学习研究的不断增长的需求。
Its commitment to creating a truly open and developer-friendly cloud infrastructure has made Akash an attractive alternative to traditional centralized providers. Ongoing integrations with AI frameworks have extended Akash’s capabilities beyond general-purpose workloads into specialized GPU-driven tasks for machine learning. Its 2025 roadmap reveals where Akash is headed next with plenty of new products poised to be rolled out.
它致力于创建真正开放且开发的云基础架构,使Akash成为传统集中提供商的有吸引力的替代品。与AI框架进行的持续集成已将Akash的功能扩展到通用工作负载之外,以用于机器学习的专用GPU驱动任务。它的2025年路线图揭示了阿卡什(Akash)接下来的前进,许多新产品准备推出。
5. Gensyn
5。聚会
Another rising star in the decentralized AI compute space is Gensyn, a blockchain-powered solution designed to facilitate large-scale ML workloads. Unlike other decentralized GPU providers, Gensyn operates on a Proof-of-Compute mechanism that ensures verifiable contributions to AI training tasks. By tokenizing AI compute incentives, Gensyn makes high-performance AI training accessible to organizations of all sizes, democratizing machine learning infrastructure.
分散的AI计算空间中的另一个后起之图是Gensyn,这是一种旨在促进大规模ML工作负载的区块链供电的解决方案。与其他分散的GPU提供商不同,Gensyn采用了验证机制,可确保对AI培训任务的可验证贡献。通过对AI的计算激励措施,Gensyn使各种规模的组织可以访问高性能的AI培训,从而使机器学习基础设施民主化。
The company has gained significant industry recognition, securing funding from venture capital giant a16z, which has accelerated its development. Recent pilot projects have demonstrated its potential for decentralized AI model training, and ongoing infrastructure upgrades are enhancing its ability to support large-scale compute operations. Gensyn doesn’t currently have a token but one is on the way, making this a DePIN to watch.
该公司获得了重要的行业认可,并从风险投资巨头A16Z获得资金,这加速了其发展。最近的试点项目已经证明了其进行分散的AI模型培训的潜力,并且正在进行的基础设施升级正在增强其支持大规模计算运营的能力。 Gensyn目前没有令牌,但正在途中,这使它成为值得关注的depin。
Decentralized Compute Is Just Getting Started
分散的计算刚刚开始
With AI computational requirements escalating, decentralized platforms offer a compelling alternative to traditional cloud services. Whether it’s Render Network’s expansion into AI workloads, io.net’s rapid deployment of high-performance clusters, Hypercycle’s real-time AI inference, Akash Network’s permissionless cloud compute, or Gensyn’s blockchain-based AI training framework, decentralized GPU compute is allowing AI innovation to flourish.
随着AI计算要求不断升级,分散平台为传统云服务提供了令人信服的替代方案。无论是渲染网络的扩展到AI工作负载,IO.NET快速部署了高性能群集,HyperCycle的实时AI推理,Akash Network的无许可云计算,还是Gensyn的基于区块链的AI AI训练框架,允许AI Innrovation允许AI Innrovation允许AI Innovation允许繁荣。
免责声明:info@kdj.com
所提供的信息并非交易建议。根据本文提供的信息进行的任何投资,kdj.com不承担任何责任。加密货币具有高波动性,强烈建议您深入研究后,谨慎投资!
如您认为本网站上使用的内容侵犯了您的版权,请立即联系我们(info@kdj.com),我们将及时删除。
-
- 咒语(OM),疼痛(痛苦),阴影(阴影)今天趋向于山寨币:这就是为什么
- 2025-02-21 07:10:22
- 经过几天的大幅下降,加密货币市场显示出恢复的迹象,过去24小时内总市值增长了580亿美元。
-
-
- 十大加密硬币要投资 - 一个简短的清单
- 2025-02-21 07:05:23
- 驾驶加密货币市场可能具有挑战性,尤其是在市场上不可预测的波动中。这使加密爱好者和买家询问
-
- XRP价格预测:2025年及以后的XRP下一步是什么?
- 2025-02-21 07:05:23
- XRP价格预测再次加热,每个人都想知道XRP是否可以最终领导下一个加密货币市场公牛。
-
-
-
-
-
- SUI(SUI)形成逆向头和肩膀的模式,目标$ 3.60弹跳
- 2025-02-21 07:05:23
- SUI(SUI)正在显示潜在突破的指标。今天,市场分析师阿里·马丁内斯(Ali Martinez