Hacker News Reader

Comment by nextstep | Hacker News Reader

I agree that the broader point is true: each giant is using software (and hardware in some cases) to build massive infrastructure businesses (logistics, cloud, communications, advertising/search, etc.).

But I think this analysis overlooks the ways where these tech companies do sort of compete: search advertising is a duopoly between Google and Facebook; search has some (very weak) competition from Microsoft and Amazon; cloud services are competitive between Google and Amazon... I think the stronger pattern is Alphabet is in everything and in some way competing with every other giant.

View on HN

Azure is by far the next biggest cloud provider after AWS. GCP is the lame duck playing catch up.

Yea don't underestimate MS's enterprise connections. Gives them a huge leg up in cloud.

Oddly enough, the three biggest cloud vendors by revenue are MS, Amazon, and _IBM_ (see links below).

IBM Cloud revenue is a head-scratcher. I've had to use them in the past, and while they have some unique features, it's hard to imagine anyone choosing them unless they had an explicit need for those features, or were forced into using IBM due to existing business relationships. I guess corporate inertia (possibly from the SoftLayer days) is a hard thing to overcome.

https://www.forbes.com/sites/bobevans1/2018/07/11/microsoft-...

https://www.forbes.com/sites/bobevans1/2018/05/21/top-cloud-...

https://www.forbes.com/sites/bobevans1/2018/04/27/microsoft-...

Is that really what you think? Google Cloud is far and away the best service for machine learning right now. There's a reason why basically every company that has ML training needs is switching from AWS to GCP. Not only is Tensorflow, which is the only production machine learning framework that anyone uses, basically 100% integrated with GCP, the Google Compute Engine is the only way you can get access to TPUs.

At my company, we use GCP almost exclusively with preembtible TPUs as our primary compute for training models. It's not cheap, but you can't beat the performance and quality. Plus when you compare it with how much the engineers here are getting paid, it's a drop in the bucket ultimately.

How much of the average compute work load for companies and startups is machine learning?

Not sure about others, but in my field, it's probably >99%.

Sounds like GCP is a great fit for you, then!

I have worked in a number of industries and rarely needed any ML (though it may be coming). Most folks need clean data and linear regressions in my experience.

If you ignore Office 365, GCP is much closer to Azure than Azure is to AWS. Microsoft's inclusion of Office 365 into its "Cloud" revenue makes their cloud platform seem to have much larger market share than what it actually has.

Reply on news.ycombinator.com