DGPT Announces Launches Its New-Generation Artificial Intelligence Computational Power Sharing Platform

LOS ANGELES, CA, October 17, 2023 – (SeaPRwire) – DGPT, announces its new generation Artificial Intelligence computational power sharing Platform, which will bring AI benefits to everyone.

In recent years, with the popularity of deep learning and large language models such as BERT, GPT-3, and ChatGPT, the demand for high-performance computational power for AI systems has been exploding. The parameter size of these large models has reached tens or even hundreds of billions, requiring a significant number of GPUs and computational resources for training. However, the computational power provided by current data centres and cloud services is far from meeting this massive demand. On the other hand, building a professional GPU server cluster independently for training such models is highly costly and has a high entry barrier. This has led to a significant imbalance between the supply and demand of computational power in the AI field.

Solutions for contradiction between the current computational power supply and demand

To address the contradiction between the current computational power supply and demand, DGPT, a decentralized idle computational power-sharing platform, was born. DGPT aims to pool idle computational resources worldwide. Whether it’s the idle computational power of individual user devices or the idle servers of various institutions, all can provide cost-effective computational support for AI system training via the DGPT platform. DGPT has been specially designed with low entry barriers, automated access, and incentive mechanisms, allowing even non-tech professionals to quickly join. With a simple registration and client software installation via a mobile app, ordinary users can opt to provide their device’s idle computational power to the DGPT platform, participate in AI model training, and the entire process is completely transparent and invisible to the user, not affecting the device’s regular use. This opens up a new way for the general public to earn passive income.

DGPT can achieve virtual pooling and flexible scheduling of heterogeneous terminal computational resources, supporting dynamic allocation and matching of available computational power across the network based on customer AI training task requirements, ensuring rational resource allocation and effective utilization. Core technologies include efficient workload allocation algorithms, cross-platform universal data interfaces, and powerful task schedulers. These innovative designs overcome the challenges faced by traditional distributed computing platforms, achieving unprecedented flexibility and scalability.

DGPT now supports model training for various AI algorithm tasks, including computer vision, natural language processing, and reinforcement learning. The platform currently serves hundreds of clients, with a node scale reaching hundreds of thousands, making it one of the industry-leading distributed AI Infrastructure-as-a-Service (AIaaS) providers.

Business Model

DGPT adopts a dual-track profit model. On one side, AI companies and researchers with training needs can rent computational power on the DGPT platform at a cost lower than traditional cloud services, significantly reducing model training expenses. On the other side, individual users and institutions that provide idle computational power can earn rewards based on the amount of computational power they contribute, thereby obtaining additional income.

DGPT has forged partnerships with numerous AI companies, research institutions, and universities, cultivating a robust ecosystem centred around AI computational power sharing. With the AI field’s ongoing expansion, the demand for computational power will persistently rise. The platform model of DGPT harbours immense growth potential for the future.

Future Vision

As we look ahead, DGPT aims to swiftly augment the number of both individual and institutional users on its platform. Our goal is to accommodate a broader range of AI model training tasks and continually elevate user engagement levels. In the long run, DGPT envisions itself as the pivotal platform bridging computational resources globally, laying down a solid infrastructure foundation for the progression of artificial intelligence. By fostering an AI ecosystem that thrives on mutual benefits, DGPT aspires to ensure that both individual users and organizations garner tangible advantages. Our ultimate objective is to catalyze the ubiquity of AI technologies and their applications, ensuring that artificial intelligence becomes a boon for the entirety of human society.

About DGPT

DGPT was established in 2021 by a team of technical experts with profound expertise in artificial intelligence and distributed computing. Core team members are graduates of top universities and have held positions in renowned companies and research institutions in the industry. After two years of relentless efforts and breakthroughs in key technologies, DGPT finally launched this revolutionary distributed heterogeneous AI computational power scheduling and optimization platform in 2023.

Social Links

Twitter: https://twitter.com/dgpt_cloud

Telegram: https://t.me/DGPT_Cloud

YouTube: https://www.youtube.com/@DGPT_AI

Instagram: https://www.instagram.com/dgpt_ai

Facebook: https://www.facebook.com/dgptcloud

Tiktok: https://www.tiktok.com/@dgpt_ai

Media contact

Company: DGPT LLC

Contact: Media team

Email: support@dgpt.cloud

Website: https://www.dgpt.cloud/



The article is provided by a third-party content provider. SeaPRwire ( https://www.seaprwire.com/ ) makes no warranties or representations in connection therewith. Any questions, please contact cs/at/SeaPRwire.com

Sectors: Top Story, Daily News

SeaPRwire provides press release distribution services to global clients in multi-languages (Hong Kong: HKCrunch, HKBrowse, AseanTrend; Singapore: SingdaoPR, TodayInSG; AseanFun; Thailand: Bangkokok, InsightTH; Indonesia: IndoNewswire, IndonesiaFolk; Philippines: ManilaPR, PhilPR.com; Malaysia: MalaysianBuzz, KLWeek; Vietnam: TinTucFN, VietnamClipping; Arab: MENAEntry, InArabia; Taiwan: EAStory, TWNut; Germany: EuropaEiner, FirmenGate)

Disclaimer: The views, suggestions, and opinions expressed here are the sole responsibility of the experts. No  journalist was involved in the writing and production of this article.