Apple Goes on Cloud Computing Hiring Spree

[AdonisSMU]

I remember time when Apple was hiring auto specialists like crazy... Where are those? Also, remember all the noise about Apple building server farms? What happened to those?

I dont know. Tell us.

 

[lilkwarrior]

Eh, poaching employees from other companies hasn't really helped them much. Also, they still need to pay Amazon for using AWS.

They use Google Cloud Platform (GCP) as well, & probably Azure as well. It's perfectly sensible to diversity your use of Cloud to multiple vendors for the sake of responsible business strategy.

 

[Romeo_Nightfall]

Ahm, ok. So Apple has absolutely no vision where they want to go or what they want to do with "the cloud", so they hire a bunch of people that have worked on different projects that had something to do with technologies that either work in "the cloud" or that are sometimes used to create "a cloud".

"The cloud" can be an unbelievable amount of completely different things. For most people, it's just "someone else's computer". This is what AWS is at its core.

Then there are all those streaming and file sharing services and applications running as subscription services somewhere else.

Azure can also do what AWS offers, and it hosts a bunch of application services as well. But on top of that, Microsoft is migrating a lot of compute (and AI) tasks away from the client side into Azure -- and those services can also be "consumed" by software developers that write software using Microsoft development tools. That kind of "cloud" is an entirely different beast - and this is where it gets really interesting and where everybody else is playing catch-up with Microsoft. The vision is to turn Azure into "the computer of the world", and their entire product strategy has been evolving around this vision for years already.

So... What's Apple trying to achieve here? Is there a vision involved or are they just panicking because that's another party they've missed?

well, do you expect Apple, where secrecy is king to lay out their plans to you?
But I can tell you what happens anyways. Apple has missed that game, but has now the commitment to at least catch up because the internal people aren’t satisfied.
There could be no vision, because they are too far behind and talent is limited.

for me the make it or break it Moment for Apple was always when there comes a company with a cloud based OS which works on any device.
seems finally m€ is working on it, maybe amazon gets the curve with Alexa, but doesn’t seem so.
Google is just from another world, who knows BUT Apple has nothing but a fragmented old school software base on different devices.

let’s see if Cook eats that soup himself or his successors has to sort it out like at m€

 

[chucker23n1]

But I can tell you what happens anyways. Apple has missed that game,

What game? Apple isn't in the cloud provider market any more than they were in the server hosting market before. They also aren't in the vacuum cleaner market; did they "miss that game" as well?

for me the make it or break it Moment for Apple was always when there comes a company with a cloud based OS which works on any device.

Why on earth would Apple do that when a huge chunk of their revenues is from selling hardware?

macOS in the cloud for continuous integration and continuous deployment is also possible. Not every business can afford expensive Mac hardware for these tasks.

Cloud Xcode development? Some folks have mentioned this as a way to bring Xcode to the iPad, perhaps the back ends for all their pro efforts? (No more Apple hardware required)

The only way they'll do stuff like this is with a fairly pricey subscription. Otherwise, they'll shrug and ask you to simply buy a Mac (or iPad, as the case may be).

 

[Abazigal]

for me the make it or break it Moment for Apple was always when there comes a company with a cloud based OS which works on any device.

A 1-size-fits-all OS that is somehow expected to scale from the smallest smartphone to the largest display sounds like a UI nightmare. Apple developing separate OSes for their smartphones, tablets, TV, watch and Macs might entail a lot more work, but it clearly results in the most optimised experience for each form factor.

 

[EmotionalSnow]

From what I've read, Apple's backend infrastructure could definitely use some work. Especially iCloud, which has been duct-taped together from parts of AppleID, MobileMe, .Mac and iTools over the past 20 years. Kind of like Microsoft Windows, Apple's cloud services have become a cobbled-together hodgepodge that generally works on the surface... but underneath it's a band-aid job. Like the water tables underneath New York and San Francisco, it's...

Okay, enough with the analogies. The infrastructure needs work (and ideally, a lot of rework) for both Apple's sake and its customers.

I think this also applies to Apple's ordering system. When I ordererd my 16" MBP, the card limit was too low and that's why it failed but the problem is that even after calling the bank and having them increase the card limit, Apple's system should retry the charge twice a day - but it did not. Therefore I had to call Apple Support twice to get it working.

And let's not forget that the ordering system has multiple bad translations and inconsitencies. I would love it if they treated everything as good as iPhone because everything else seems to be a second class citizen ...

 

[himanshumodi]

Ahm, ok. So Apple has absolutely no vision where they want to go or what they want to do with "the cloud", so they hire a bunch of people that have worked on different projects that had something to do with technologies that either work in "the cloud" or that are sometimes used to create "a cloud".

"The cloud" can be an unbelievable amount of completely different things. For most people, it's just "someone else's computer". This is what AWS is at its core.

Then there are all those streaming and file sharing services and applications running as subscription services somewhere else.

Azure can also do what AWS offers, and it hosts a bunch of application services as well. But on top of that, Microsoft is migrating a lot of compute (and AI) tasks away from the client side into Azure -- and those services can also be "consumed" by software developers that write software using Microsoft development tools. That kind of "cloud" is an entirely different beast - and this is where it gets really interesting and where everybody else is playing catch-up with Microsoft. The vision is to turn Azure into "the computer of the world", and their entire product strategy has been evolving around this vision for years already.

So... What's Apple trying to achieve here? Is there a vision involved or are they just panicking because that's another party they've missed?

Whatever the article says, is based solely on job openings. Not as if Apple has published a vision document or anything. My first instinct is that it's just highly unlikely that they will build something to compete against AWS/Azure/GCP for enterprises. That's just not where Apple is. In all likelihood, these people will build inhouse capabilities for Apple's own consumption.

 

[ned1]

I have stopped using Apple’s Cloud based services for storage of documents having had such a poor technical experience where documents were just not syncing correctly across all my devices, an issue that took Apple over a year to resolve, the issue being passed between six senior advisors and finally passed to a Shane Barton Executive Liaison, who had little if any technical understanding, who refused to review the work the previous advisors had undertaken in preference to starting the investigation all over again - A further three months were wasted, no apology from Apple and no route to be able to lodge a complaint on the way Executive Liaison had dealt with the case.

Apple really need to hire people who are technically proficient and understand the software they are dealing with…..So Apple, hire away and we might eventually get some software that works seamlessly!

 

[KanosWRX]

There is no way Apple would want to get into the Enterprise Cloud business. They would have to invest a whole lot more then 10B and hire tens of thousands of more people. What they are doing is realizing they need to modernize their own infrastructure probably to run more like Google as they invented the whole idea of containers and Kubernetes. With all their new services they need to modernize and optimize and who better then to look at how Google and GCP does it.

 

[chucker23n1]

There is no way Apple would want to get into the Enterprise Cloud business. They would have to invest a whole lot more then 10B and hire tens of thousands of more people. What they are doing is realizing they need to modernize their own infrastructure probably to run more like Google as they invented the whole idea of containers and Kubernetes. With all their new services they need to modernize and optimize and who better then to look at how Google and GCP does it.

Kubernetes is a Google project, but containers in their modern form appeared a little early at Docker, Inc.

I'm also not sure why everyone seems to be an expert on Apple's infrastructure not being modern. Maybe? But how do we know that?

 

[UKgaryb]

I've been approached by Apple recently, turned down an interview as i'm happy where I am... but they are looking to hire a lot of SRE/Kubernetes people or that's what they are asking for when I spoke to them.

Kubernetes is a Google project, but containers in their modern form appeared a little early at Docker, Inc.

I'm also not sure why everyone seems to be an expert on Apple's infrastructure not being modern. Maybe? But how do we know that?

Kubernetes is a child of a few projects from Google including Borg & Omega but K8's isn't a Google project anymore it's owned by the CNCF (Cloud Native Computing Foundation).

Docker used LXC but branched off to it's own libcontainer ... LXC used cgroups and linux namespaces, cgroups was invented by Google (then called Process Containers)

 

[ct2k7]

Right now, there's a drive for Apple Services SREs. They've approached me, but I had to turn down.
[automerge]1589287865[/automerge]

I've been approached by Apple recently, turned down an interview as i'm happy where I am... but they are looking to hire a lot of SRE/Kubernetes people or that's what they are asking for when I spoke to them.



Kubernetes is a child of a few projects from Google including Borg & Omega but K8's isn't a Google project anymore it's owned by the CNCF (Cloud Native Computing Foundation).

Docker used LXC but branched off to it's own libcontainer ... LXC used cgroups and linux namespaces, cgroups was invented by Google (then called Process Containers)

Out of interest, was this for an Apple Services SRE role?

 

[UKgaryb]

Right now, there's a drive for Apple Services SREs. They've approached me, but I had to turn down.
[automerge]1589287865[/automerge]


Out of interest, was this for an Apple Services SRE role?

Yeah they asked ridiculous questions like 'how many clusters do you run in production'
Was yours the same?

 

[ct2k7]

Yeah they asked ridiculous questions like 'how many clusters do you run in production'
Was yours the same?

I can't interview with them as I'm working with one of their partners, with an anti-poaching agreement

 

[name99]

Maybe we will see new developer-oriented iCloud services designed specifically for iOS, iPadOS, and macOS apps.

macOS in the cloud for continuous integration and continuous deployment is also possible. Not every business can afford expensive Mac hardware for these tasks.

Yeah, this guy gets it ^^^

As the saying goes, well duh! OF COURSE Apple was eventually going to do this, and OF COURSE part of the eventual expanded offerings will be developer support so that devs don't need to be using AWS or Azure as one more foreign thing they have to learn and pay for.

Likewise personal elastic computing (eg compiling, video editing, Mathematica) obviously makes sense. The issues that matter are
- are there enough users with enough bandwidth (especially upstream bandwidth) to make this a win? This is something Apple doesn't have much control over; all they can do is monitor mainstream internet speeds and mainstream compute loads.
- control of the back-end, which may extend to control of the CPU. You don't want to give random code the ability to interact with other simultaneously running tasks... Which may mean Apple CPUs (can't much trust Intel these days...) and/or only allowed particular validated code to be running on behalf of users?

All this stuff is coming, everything just takes time. Like the ARM servers. People like me (and ARM) were mocked for suggesting 2020 but, oh look, 2020 and there are now multiple credible ARM server SoCs, including a large number installed in AWS.

Forget all the comments whining about some bug in iCloud three years ago. That's the past, this effort is targeting the future.

 

[Karma*Police]

As usual, Tim Cook (and Eddy Cue) is late to the party. I wrote to him 7 years ago imploring him to think big and invest more in cloud technology, satellites and to expand iAds. Now, Amazon is raking in billions in 2 of those categories while Apple, once the undisputed leader in digital media, plays second fiddle to Spotify and is wasting billions on not so great video content and lags half a dozen other companies in that space, including AT&T... and if you’re behind AT&T in anything, you know it’s bad.

Apple wasted hundreds of billions buying back stock and have contributed little to moving technology forward in any meaningful way under Tim‘s watch. Sure, they did well in wearables, but I would argue that’s an outgrowth of Steve’s recipe and low risk ventures. Imagine if they had spent just a fraction of the billions they spent on stock buybacks (and beats by Dre) on building out massive data centers with amazing capabilities for consumers, developers and businesses and sat tech so we can have access to data anywhere in the world. As for advertising, I’m not a huge fan of the industry, but it could have been used as a strategic weapon against Google and the billions they could have generated could have subsidized cloud services, which in turn could have sold more hardware.

 

[Unregistered 4U]

The only way they'll do stuff like this is with a fairly pricey subscription.

If pricey means less than $200 per year, then sure. Especially if means you don’t have to pay for a dev subscription separately.

 

[chucker23n1]

If pricey means less than $200 per year, then sure. Especially if means you don’t have to pay for a dev subscription separately.

$200/yr for unlimited continuous integration would be a steal.

Look at these guys discussing that they're expecting $75k in CI costs this year, and $90k next year.

At $200, they'd have to massively throttle it.

 

[Unregistered 4U]

At $200, they'd have to massively throttle it.

If they were getting $200 a year from everyone using it, it’d pay for itself, though, right? I mean isn’t their problem that they’re trying to scale something that is effectively free? Apple wouldn’t have that problem.

 

[chucker23n1]

If they were getting $200 a year from everyone using it, it’d pay for itself, though, right? I mean isn’t their problem that they’re trying to scale something that is effectively free? Apple wouldn’t have that problem.

Wait, are we talking about the same thing? Putting a whole lotta CPU power in a data center is hardly free.

If you mean offering a macOS license for someone else’s existing machine, yeah, maybe.

 

[Unregistered 4U]

We might just be. I might not properly understand the link. I read that post and what I took from it was “WE NEED MOAR MONIES!” The amount of monies freedesktop.org needed were $90k every year. I assumed from the “free” that their problem is a LOT of people using a service, but few paying for it. So, I figured if the folks using it all paid $200 a year, then freedesktop.org’s money problems would no longer be money problems. They’d likely have way more than $90k a year.

 

[chucker23n1]

We might just be. I might not properly understand the link. I read that post and what I took from it was “WE NEED MOAR MONIES!” The amount of monies freedesktop.org needed were $90k every year. I assumed from the “free” that their problem is a LOT of people using a service, but few paying for it. So, I figured if the folks using it all paid $200 a year, then freedesktop.org’s money problems would no longer be money problems. They’d likely have way more than $90k a year.

Oh, sure. But that's not the same kind of thing.

freedesktop.org pays that money to a cloud service provider (appears to be Google), who charge based on CPU usage.

Apple would be insane to offer the same service for just $200/yr (when others take five figures) and not at least heavily throttle it. If they do want to enter that market (which I really don't see happening), they'll similarly have a scaling model where you can start out very cheaply, but costs can also go way up. A flat fee just doesn't make sense for this kind of service.

(If you do want a flat fee, you can do that at MacStadium starting at $79/mo, or almost five times your cited $200/yr. It's probably possible to undercut them, but nobody seems to want to do so.)

 

[I7guy]

As usual, Tim Cook (and Eddy Cue) is late to the party. I wrote to him 7 years ago imploring him to think big and invest more in cloud technology, satellites and to expand iAds. Now, Amazon is raking in billions in 2 of those categories while Apple, once the undisputed leader in digital media, plays second fiddle to Spotify and is wasting billions on not so great video content and lags half a dozen other companies in that space, including AT&T... and if you’re behind AT&T in anything, you know it’s bad.

Apple wasted hundreds of billions buying back stock and have contributed little to moving technology forward in any meaningful way under Tim‘s watch. Sure, they did well in wearables, but I would argue that’s an outgrowth of Steve’s recipe and low risk ventures. Imagine if they had spent just a fraction of the billions they spent on stock buybacks (and beats by Dre) on building out massive data centers with amazing capabilities for consumers, developers and businesses and sat tech so we can have access to data anywhere in the world. As for advertising, I’m not a huge fan of the industry, but it could have been used as a strategic weapon against Google and the billions they could have generated could have subsidized cloud services, which in turn could have sold more hardware.

Apple would never develop an enterprise cloud solution like you are suggesting. And who says apple plays second fiddle to Spotify? As far as the rest, the direction Apple took is not the one you wanted. Tim made Apple into a behemoth, and attributing it to Steve is folly.

 

[chucker23n1]

As usual, Tim Cook (and Eddy Cue) is late to the party. I wrote to him 7 years ago imploring him to think big and invest more in cloud technology, satellites and to expand iAds. Now, Amazon is raking in billions in 2 of those categories while Apple, once the undisputed leader in digital media,

Sorry, uh, when was Apple ever a leader in server hosting?

And what does AWS have to do with "digital media", whatever that even is?

plays second fiddle to Spotify

I do think Apple missed the subscription boom for a few years, but Apple Music is massive now.

and is wasting billions on not so great video content and lags half a dozen other companies in that space, including AT&T... and if you’re behind AT&T in anything, you know it’s bad.

Er. AT&T is only "ahead" (are they, though?) through an acquisition. Apple could've bought a back catalog and chose not to. And AT&T actually has a difficult task ahead of building up a brand without tarnishing the good name of HBO. So far, they seem to be doing a so-so job at that.

Apple wasted hundreds of billions buying back stock and have contributed little to moving technology forward in any meaningful way under Tim‘s watch. Sure, they did well in wearables, but I would argue that’s an outgrowth of Steve’s recipe and low risk ventures.

Sorry, what? You don't get to have it both ways.

If they're "an outgrowth of Steve's recipe", then how is Tim doing anything wrong?

Imagine if they had spent just a fraction of the billions they spent on stock buybacks (and beats by Dre) on building out massive data centers with amazing capabilities for consumers, developers and businesses and sat tech so we can have access to data anywhere in the world.

OK, but… why?

This seems like a "be more like every single other tech company out there, Apple!" argument, but also "be more like Steve, who was… uhhh… not at all like every single other tech company out there, but also be like that!!"

We've had plenty of that in that in the 90s ("Apple needs to license macOS or they're doomed"), and it was precisely not doing what everyone else did that helped Apple survive.

 

[Unregistered 4U]

Apple would be insane to offer the same service for just $200/yr

GOTCHA! I see where you’re coming from. As Apple goes into things from a profit perspective, I’d agree they’d have to limit it in some specific way to ensure an appropriate ROI.