#GPUs

[ follow ]
gpus
Theregister
1 month ago
Data science

What Nvidia's Blackwell efficiency gains mean for DC owners

CPUs and GPUs are becoming hotter and more power-hungry, causing challenges for datacenter operators.
Nvidia's Blackwell GPUs have high power consumption, necessitating modifications like liquid cooling for efficient operation. [ more ]
Ars Technica
3 months ago
Artificial intelligence

Report: Sam Altman seeking trillions for AI chip fabrication from UAE, others

OpenAI CEO Sam Altman is seeking to raise $5 trillion to $7 trillion for AI chip manufacturing.
Altman has met with potential investors including SoftBank CEO Masayoshi Son and representatives from TSMC. [ more ]
TechCrunch
3 months ago
Artificial intelligence

Inference.ai matches AI workloads with cloud GPU compute | TechCrunch

GPUs are in high demand for AI applications, leading to supply constraints and high prices.
Inference.ai aims to provide affordable and accessible GPU compute through partnerships with data centers. [ more ]
Theregister
4 months ago
DevOps

Tens of thousands of GPUs go under-utilized in the cloud

Cloud providers are deploying a large number of GPUs and AI accelerators, but evidence suggests they are being under-utilized.
TechInsights estimates that in 2023 alone, 878,000 accelerators were responsible for $5.8 billion in revenue, but this figure could be much higher if clusters were operating closer to capacity. [ more ]
The Drum
4 months ago
Artificial intelligence

Weekly AI recap: OpenAI's GPT store, Nvidia's new GPUs

OpenAI has launched its GPT store, an online marketplace for browsing, purchasing, and selling custom versions of ChatGPT.
Nvidia has announced three new GPUs for 'local' AI systems, which will provide more options for AI developers and gamers. [ more ]
moregpus
CodeProject
6 months ago
Data science

Transfer Learning with TensorFlow on Intel Arc GPUs

Transfer learning is a deep learning method that allows the use of a pretrained model with a new dataset.
Intel Arc A-Series GPUs can run DL workloads quickly on PCs with TensorFlow and PyTorch models.
Using transfer learning with Intel Arc GPUs can significantly reduce training time compared to fully training the model. [ more ]
CodeProject
6 months ago
Data science

Transfer Learning with TensorFlow on Intel Arc GPUs

Transfer learning is a deep learning method that allows the use of a pretrained model with a new dataset.
Intel Arc A-Series GPUs can run DL workloads quickly on PCs with TensorFlow and PyTorch models.
Using transfer learning with Intel Arc GPUs can significantly reduce training time compared to fully training the model. [ more ]
[ Load more ]