Follow the CAPEX: Separating the Clowns from the Clouds
platformonomics.com
platformonomics.com
Had a good, long laugh at that one. Jesus, is there anyone on this planet that doesn't regard Oracle as the corporate equivalent of a slimy, violent mob boss?
> You may not get fired for buying IBM, as the old saying goes, but it is increasingly likely your employer will go out of business if you’re in an industry where technology matters.
* "Amazon, Google, and Microsoft each spent more on CAPEX in 2017 than Oracle has in its entire history."
* "That red line you may mistake for the x-axis is Oracle’s CAPEX spending"
* "Maybe they are reconciled to sitting at the children’s table of cloud, but the problem for both IBM and Oracle is cloud is eating their existing businesses. It has eaten the server business and now starting to feast in earnest on software infrastructure, including the database, which is the profitable heart of these companies."
* "Both companies have acquired a number of SaaS applications, which will bolster their sense of self-worth and belonging in the cloud, but there is little to no platform leverage associated with these apps (and platform leverage = profits!!!)."
* "Watson which is in serious contention to be the biggest “overpromise and underdeliver” in tech industry history, now blockchain as they try to save humanity from our looming existential tomato provenance crisis"
* "And their (IBM’s) customer problem is who their customers are at this point: the disrupted. You may not get fired for buying IBM, as the old saying goes, but it is increasingly likely your employer will go out of business if you’re in an industry where technology matters."
They could piss off us nerds and survive, but they’ve now lost a lot of credibility among C-level execs too, at this point. And to cap it off, they’re trying to force customers to make big migrations to cloud products that are simply not as good as the existing on-prem versions, just so they can toot inflated cloud numbers to Wall Street analysts.
And again, they didn’t have to: just sell both cloud and on-prem and let your customers choose.
You have one team trying to save the company (however fruitless it might be) by creating a cloud and being forward thinking and you have another team that’s trying to squeeze their existing customers for every drop. The former can only be hampered by the latter.
A tip for anyone running Oracle: I have a friend at a previous employer who put the screws to Oracle by running some non-prod on a massive box and just creates instances on the fly with docker containers for developers and automated testing. Of course I’d ask an attorney before doing that, YMMV.
As if AWS was a mere experiment, and any day now the big boys would come in and show how it’s done.
Microsoft has become the de facto number two cloud provider, and Azure is now AWS's main competitor. That is in great part because IT managers (and behind them, CIOs and procurement) are familiar with Microsoft products.
Keeping the AD sync stuff under the Azure umbrella was a genius move really. Microsoft understands big enterprises along with their politics and slow processes.
Surely Digital Ocean, Linode, Vultr etc don't have the resources to invest in these massive facilities.
Yet, they have customers, and provide options.
How are the big three (AWS, GCP, MS) able to sell on such a scale?
I don't know the whole answer, but part of it is that they give free credits to new users, to get them using their cloud. It's smart. Once the free credits are used up, the business just starts paying, because switching to a different cloud provider is usually non-trivial.
I think they also focus on a few large customers. SNAP, Netflix, and others can't really use Digital Ocean (can they? Does anyone know enough to do some back of envelope math?)
E.g. I tried finding the Azure equivalent of a $5/m nano/micro instance the other day and it doesn’t seem to exist. I think their minimum is $25/m just for the compute portion. That gets you 1 vcpu/1 gig. I honestly have no idea why anyone uses Azure.
hopefully you can see what I see, but $0.008 x 730 hours is about $5.5
- Arbitrary private network topologies
- Managed databases
- Other managed services for the "undifferentiated heavy lifting" of infrastructure like source control, CI/CD, monitoring, log aggregation, alerting, load balancing, encryption, directory services, etc.
As such, they're in a different business - providing single servers for hobbyist and small-business websites - rather than seriously competing with on-prem for big-co infrastructure needs. They're making some attempts to catch up but are far behind.
Since this discussion is "old", I might have to craft an Ask HN
Are you implying that they're particularly high? Reference please :)
I wouldn't be surprised if full list price for "brand name" vendors like IBM, HP, and maybe even Dell (presumably Oracle, too) are inflated, but in every case I've heard of, for even modest volumes, the discounts tend to equalize the prices (well, maybe not IBM) and can occasionally be competitive with SuperMicro.
I'd aregue that the only company that really makes (or at least ultra-customizes/optimizes) their own, to the point of being outside of any "class" of hardware, is Google. I suspect they're the only ones whose capex might be deceptively low because of the cost optimization.