This increases visibility into your paid compute consumption. Net win for folks. Feature doesn't launch for 30 days, but I can't make any changes like this without changing ToS, which I have to pre-announce.
I've been using Colab for a few months as I dive into my personal ML education. Free at first, Pro tier now. It's very handy and I appreciate the product.
I suppose my concern is this - Was Colab using "compute units" internally for limits already? Or was it "as available"? And if so, will my compute units allocation be what it has been or will it be decreasing? To be honest, when I read the email today, I assumed the limits would be tightening. It feels like the sort of thing that happens when a company tightens up on a product. Not really a reflection on you or your product, just past experiences.
Colab has always had a notion of quota for all users. This notion was implemented in a hard to describe way, but we would throttle you as you used more compute without making it clear how or why. This change will make the quota allocation transparent and visible for paid users.
Any limits we're using here are just to ensure you don't use more GPU $ than what you've paid for.
> Any limits we're using here are just to ensure you don't use more GPU $ than what you've paid for.
So naturally, any unused balance will be refunded, right Chris? You wouldn't want to start overcharging some users while fixing the "bug" or "loophole" that undercharged others.
It's a fair critique and one we have discussed. With this change we are extending the lifetime of the units to three months to give users more flexibility. This is a net user improvement, today if you pay for a subscription any "unused" quota disappears after a month.
Just wanted to say thank you for the transparency in posting here; does this mean that for paid users, compute consumption limits will stay the same or go up?
Grand majority of paid users will see no impact to their limits. We do have some bugs where a small fraction of paid users today use more than their quota (Pro+ users burning $100 worth of an A100 when paying us $50), and those folks will see limits, yes, as we fix some loopholes.
I hesitate to post this given the anger I've seen about any changes, but I can't subsidize paid tiers, that gets expensive fast. As best I can tell from looking at prices for competitors in this space, Colab will still be the best value for money. If folks find otherwise, please do tell me where the competitors are sourcing those GPUs for cheaper.
It goes without saying we remain committed to supporting the free tier.