Tether launches AI framework for smartphones and consumer GPUs
Tether launches a framework for AI training on smartphones and consumer graphics processors. The company assures that the results not only significantly reduce hardware requirements but also the training process itself.
Highlights
- Tether launches AI framework for smartphones and consumer GPUs
- BitNet reduces memory needs by up to 77.8% for training
- Framework enables on-device and federated AI model training
Nvidia chips are not the only option anymore
The framework, part of Tether’s QVAC platform, allows fine-tuning large language models on consumer hardware, including smartphones and graphics processors, expanding support beyond the dominant Nvidia GPUs typically used for AI training.
The platform supports cross-platform training and inference on various chips, including AMD, Intel, Apple Silicon, and mobile GPUs from Qualcomm and Apple. The system uses Microsoft’s BitNet architecture and LoRA techniques to reduce memory and computational requirements.
Thanks to the 1-bit BitNet model architecture, the platform can reduce video memory requirements by up to 77.8% compared to similar 16-bit models, enabling larger models to run on resource-constrained devices. Tether’s engineers have fine-tuned models with up to 1 billion parameters on smartphones in under two hours, with smaller models taking just a few minutes, while supporting models up to 13 billion parameters on mobile devices.
Mobile GPUs can process BitNet models several times faster than CPUs. Potential use cases include on-device training and federated learning, where models update across distributed devices without sending data to centralized servers, potentially reducing reliance on cloud infrastructure.
A significant shift in the AI industry
The launch marks an important step toward decentralizing the AI industry, which currently relies heavily on cloud providers and expensive GPU clusters. Training models directly on user devices opens the door to more private and autonomous AI applications, keeping data on-device—a key advantage amid growing data protection regulations.
Additionally, reducing dependence on Nvidia hardware could reshape the competitive landscape, strengthening alternative chip makers and encouraging the development of more energy-efficient solutions. If widely adopted, this technology may accelerate mass AI deployment in consumer products and create new business models centered on edge computing and distributed training networks.
As we reported, Tether QVAC hits mobile with LLAMA 3.2, Paolo Ardoino notes
Ваша пробная версия Premium закончилась
-
Afghanistan
-
Albania
-
Algeria
-
Angola
-
Argentina
-
Armenia
-
Australia
-
Austria
-
Azerbaijan
-
Bahamas
-
Bahrain
-
Bangladesh
-
Belarus
-
Belgium
-
Bolivia
-
Botswana
-
Brazil
-
Brunei
-
Bulgaria
-
Cambodia
-
Cameroon
-
Canada
-
Chile
-
China
-
Colombia
-
Congo
-
Costa Rica
-
Cote d'Ivoire
-
Croatia
-
Cuba
-
Cyprus
-
Czechia
-
DR Congo
-
Denmark
-
Dominican Republic
-
Ecuador
-
Egypt
-
El Salvador
-
Estonia
-
Eswatini
-
Ethiopia
-
Finland
-
France
-
Georgia
-
Germany
-
Ghana
-
Greece
-
Haiti
-
Hong Kong
-
Hungary
-
India
-
Indonesia
-
Iran, Islamic republic
-
Iraq
-
Ireland
-
Israel
-
Italy
-
Jamaica
-
Japan
-
Jordan
-
Kazakhstan
-
Kenya
-
Korea
-
Kuwait
-
Kyrgyzstan
-
Laos
-
Latvia
-
Lebanon
-
Lesotho
-
Libya
-
Lithuania
-
Luxembourg
-
Madagascar
-
Malaysia
-
Malta
-
Mauritius
-
Mexico
-
Moldova
-
Mongolia
-
Montenegro
-
Morocco
-
Mozambique
-
Myanmar
-
Namibia
-
Nepal
-
Netherlands
-
New Zealand
-
Nigeria
-
North Macedonia
-
Norway
-
Oman
-
Pakistan
-
Palestine
-
Panama
-
Papua New Guinea
-
Paraguay
-
Peru
-
Philippines
-
Poland
-
Portugal
-
Puerto Rico
-
Qatar
-
Reunion
-
Romania
-
Rwanda
-
Saudi Arabia
-
Serbia
-
Singapore
-
Slovakia
-
Slovenia
-
Somalia
-
South Africa
-
Spain
-
Sri Lanka
-
Sweden
-
Switzerland
-
Syria
-
Taiwan
-
Tajikistan
-
Tanzania
-
Thailand
-
Trinidad and Tobago
-
Tunisia
-
Turkey
-
UAE
-
Uganda
-
Ukraine
-
United Kingdom
-
United States
-
Uruguay
-
Uzbekistan
-
Venezuela
-
Vietnam
-
Yemen
-
Zambia
-
Zimbabwe
- Forex
- Crypto
