Submissions from exafunction.com

		Loading 60 BERTs onto a single T4 (exafunction.com)
		3 points by varunkmohan on Sept 26, 2022 \| past
		Getting Rid of CPU-GPU Copies in TensorFlow (exafunction.com)
		6 points by varunkmohan on Sept 19, 2022 \| past
		Are GPUs Worth It for ML? (exafunction.com)
		131 points by varunkmohan on Aug 29, 2022 \| past \| 91 comments
		ML Deployment: When the Big Problem Is Going Big (exafunction.com)
		2 points by sebg on Aug 20, 2022 \| past
		Increase usable cloud GPU memory by up to 6.6% through disabling ECC (exafunction.com)
		22 points by vvvhuang on Aug 2, 2022 \| past \| 2 comments
		Show HN: Exafunction, efficient deep learning at scale (exafunction.com)
		8 points by varunkmohan on April 28, 2022 \| past \| 5 comments
		Show HN: Exafunction NLP – Run models like GPT-J for 6x cheaper than OpenAI (exafunction.com)
		45 points by varunkmohan on Sept 28, 2021 \| past \| 19 comments