practices

If you’re in the business of creating large language models you might already think Nvidia’s pricing for its H100 super-duper