Ask Paul: June 26 (Premium)

Happy Friday! Not surprisingly, this week’s Apple announcements have triggered a wave of questions. So let’s dive right in.

PC performance

ErichK asks:

Paul — do you think we’ll ever hit a wall with how sophisticated and fast computers will get? After all these days we almost take for granted that even an average PC has more power than a mainframe from the early ’90s. Engineers have to work so hard to give us these gains. I just wonder if it’s ever going to come to an end, or at least plateau a great deal. Plus, it takes a large team of people nowadays just to make one chip. It’s like an operating system — I sometimes wonder if Windows will keep expanding to a point where the collective IQ of the people that work on it will not be enough to handle it. (Of course, there’s some hope — seems like every year, there’s a new contender for the top of the supercomputer list.)

In writing the article about AMD’s energy efficiency milestone, I had to look up Moore’s Law because I was curiously vague on the timeframe for that prediction. As it turns out, that is because the “law” has changed over time: When it was first introduced, Moore said that the number of components in integrated circuits would double every year. But that was modified over time to 18 months and then 2 years. Today, we pretty much don’t really talk about Moore’s Law anymore, in part, I think, because the performance gains are really slowing down. Even a several year-old PC would probably be fine for virtually anyone.

These days, the talk is about performance-per-watt, which his a deliciously skewed way to describe performance, because it values low-power—i.e. ARM chips—over actual power. But PCs and workstations aimed at gamers, engineers, developers, and other power users literally just need performance, energy-efficiency be damned. And this is one of the biggest unknowns of the Apple transition, how it will maintain/improve on performance on desktop systems.

But with regards to Windows, Microsoft, like Apple, has long understood that mainstream personal computing was all heading to thin, light, and portable, ideally in devices with no fans. And that was what was behind the ARM push, fundamentally, to force Intel to compete effectively in that space and have multiple suppliers of chipsets that could meet that need. That Apple can just go it alone in this area has got to be frustrating to Microsoft.

The next year or so is going to be very interesting.

Windows on Intel silicon

lvthunder asks:

Do you think Microsoft will make Windows 10 work on the Apple based silicone? If they want competition in the chip space pitting Apple, Qualcom, Intel, and AMD go at it would be great.

So there are two possibilities here. That Microsoft would literally partner with Apple to make it a fourth supplier for chips that come inside Windows-based PCs, and that Microsoft would port Windows 10 on ARM to Apple silicon so that it could run on Macs.

But I don’t see either happening, no.

First of all, Apple’s use of its own chips is a strategic and marketing advantage, and Apple would never give that up for Microsoft. Imagine if Windows actually ran better than macOS on the same hardware. Apple can’t take that chance.

In the second instance, it should be noted that Windows 10 on ARM is misnamed; it should really be called Windows 10 on Snapdragon. (Incidentally, this is what Qualcomm calls it.) That’s because there are multiple variants of ARM chipsets out in the world, and while Microsoft has pretended that it plans to bring WOA to other ARM chipsets, there is zero business case for doing so. Sans the Apple tie-up, which will never happen. So Microsoft would need to spend time/resources modifying WOA to run on Apple silicon and even if it was wildly successful, it would then run only virtually under macOS since Apple is not adding Boot Camp to the new platform.

Anyway, this will never happen.

Future of Windows 10X

TheJoeFin asks:

Do you have any Windows 10X information? (public beta? is progress looking good?) As a Windows fan watching the Big Sur announcements I kept thinking that Windows 10X is Microsoft’s only real opportunity to breathe new life into their only successful operating system.

Nothing new, sorry. But Microsoft’s decision to delay Surface Neo because 10X wasn’t ready is telling, in my opinion. And while I’m all for a simpler version of Windows that finally offers a solution for moving forward while containing legacy technology, this thing looks sickly compared to macOS Big Sur. If you look at the two side-by-side, you can see the problem immediately.

And that’s kind of interesting because Apple is providing a model by which a legacy system can be moved forward. Apple’s schedule may be too aggressive for Microsoft, but surely there is a middle ground between the Apple approach and Microsoft’s “no code left” behind strategy. I bet this Apple event triggers some rethinking in Panos’s group in Redmond.

In a related question, MartinusV2 asks:

Since WWDC this week have shaken us hard the Microsoft fans, did Microsoft knew beforehand that Windows 10x would not be good enough to compete against MacOS Big Sur? Pretty sure they knew what was coming with their work with Office.

While I’m sure Microsoft had an “oh shit” moment internally with regards to macOS Big Sur, it’s pretty clear that Windows 10X simply wasn’t ready and that Panos killed/delayed it for good reason. Maybe Big Sur was just the icing on the cake, so to speak.

Optimistic for the future

brisonharvey asks:

At a higher level, did you come away more optimistic for Microsoft’s future following Build or Apple’s future following WWDC this year?

Interesting. As someone firmly in the Microsoft camp, I do tend to view its competitors’ events (like Google I/O and WWDC) through that lens. That is, I think about whether they are doing things better/worse than Microsoft, whether Microsoft has anything to learn from them, and so. This WWDC was particularly acute because of the whole ARM thing, of course. I almost don’t care at all about a thing they said about iPad, Apple TV, or Apple Watch, and just a tiny bit about iPhone. But the Mac thing. That was fascinating.

I came away from both shows impressed by how these companies handled the virtual information dump. Both had pros and cons from a presentation perspective—Apple’s was slicker, as expected—but overall both were great. As far as being optimistic for their futures, I feel that both companies are well-positioned for the future. But these companies are not standing toe-to-toe. There’s some overlap, but Microsoft’s primary business segment is business while Apple’s is consumer. Those are very different.

It seemed like Microsoft focused on expanding their new platforms (Teams, PowerApps, Fluid Framework) and built a bridge for older apps in Project Reunion and (to some extent) Windows Package Manager. Apple, as usual, iterated their platforms forward with some new features. I really was excited about what the Fluid Framework might mean for the future of Office and I think that Project Reunion could go a long way in building some design continuity between old and new Windows apps. But with the big hardware platform jump from Apple, it seems like they are better prepared for the future of computing. Maybe next year, Microsoft?

I agree with all that, but my overarching concern for Microsoft on the client is what I voiced previously: It has a bad habit of bragging about what it’s going to do, slowly dripping out parts of what it promised, and then drifting off and never finishing it. Apple, meanwhile, does that all internally, so by the time it announces or releases anything, it tends to be complete, thoughtful, and full-featured. It’s hard not to prefer Apple’s model given how terrible Microsoft has been at finishing the job.

I guess I’d say that Apple is better positioned for the future personal computing, especially the hardware bit. Microsoft is better positioned for the cloud, including online services, especially those aimed at productivity.

Future of Apple’s platforms

ggolcher asks:

Knowing the transition for Mac to ARM, the coming ability to run iPad apps on Mac, and the fact that the new control center has big items that look clearly tappable, I suspect future developments are afoot and I’d love to get your perspective:

Obviously, I can just voice my opinions/expectations here…

Do you think we’re slowly edging towards supporting touch on macOS?

Anything is possible, but I don’t see this happening. Apple has been weirdly anti-touch on the Mac and there’s no indication that’s changing. More likely is…

Do you think there is any chance of converging iPadOS and macOS?

Yeah. I’ve long felt that the iPad could replace the low-end portable Macs with just a few changes, but that future assumed Intel was needed for the more powerful Macs. Now it seems that’s not the case, and this further blurs the line between the two platforms.

So here’s my guess. Step one is what we know now: Mac users will be able to run iPad (and iPhone) apps natively on their Macs. Step two, which could come as soon as WWDC21, sees Apple bringing MacOS capabilities to the iPad, allowing for such things as Xcode, full Office, full Photoshop, and so on to run on those devices. After all, these apps will be Apple silicon native. It will be possible.

Any other long-term vision perspective you think may happen?

If the above happens, I could see iPad slowly replacing the Mac in portables and Apple retaining non-touch macOS for its workstation-class devices only. Once the iPad can do what the Mac does, keeping both just makes less sense.

But again, this is all just speculation.

AMD v. Intel

PhilByPond asks:

Do you know of a PC maker (Dell, HP, Lenovo, Microsoft, …) that offers the same laptop or 2-n-1 with both AMD and Intel and have they been compared in a review? It would be a way to give a boost to AMD’s traction. It may be that the OEM agreements preclude that happening. I took a quick look at Lenovo ThinkPads shows different model numbers (like T490 & T495) for Intel vs. AMD. My guess, but not sure, is that these are “equivalent” machines that could be compared.

The only one I’m aware of is Surface Laptop 3, which is available in both AMD and Intel configurations in the 15-inch version; the problem is that the Intel versions are only available through resellers since they’re part of the Surface for business offerings.

Unfortunately, the AMD part used there is lackluster and you will routinely find reviewers dinging its performance and recommending that readers get the Intel version instead. I suspect that a future version (Surface Laptop 4) that offers the Ryzen 4000-series chips would be better competition.

Do you know if Intel and/or AMD impose restrictions/incentives on companies so that they do not offer “identical” machines with comparable processors from both Intel and AMD? The odd way that the Surface Laptop 13.5″ and 15″ could be configured on the Microsoft site strongly suggests this is the case.

No, I’m not aware of anything like that. But it is striking that there are so few PCs with the same physical design that don’t come with either chipset.

Is Mac the future?

will asks:

I think and believe that your comments on Apple this week have been correct in every way. Right now we do not know what chips Apple has for their desktops and laptops, outside of the iPad chip running in a dev kit. Going forward there has to be some reason someone would buy a MacBook Pro vs a MacBook Air, both with Apple chips, or an iPad. Same goes for all their Pro vs consumer line of desktops. But for enterprise, does a Mac mini, say the size of an Apple TV, that runs almost any iPad or iPhone app, be a good fit for a business with an MDM solution that basically locks it down easier and better than a Windows machine? I know not all business programs in use are iOS or web apps, but 2-3 years down the road what would this look like for organizations?

There is a line of speculation now that Apple can/will lower prices on Macs because they are saving money by using their own chipsets, and that should such a thing happen, the Mac could start making inroads again against Windows and better leverage the iPhone halo.

And there is a precedent for this belief: Remember that iPad sales fell for six straight years before Apple woke up and lowered prices. And iPad sales have been up ever since: You can buy a great iPad now for just $329. That is … literally reasonable.

I do think Apple should lower prices on its ARM-based Macs, and that doing so could really improve unit sales and marketshare. But I’m not sure it can cut enough to see mass deployments with businesses or a true surge with consumers. Today, the cheapest portable Mac, the MacBook Air, starts at $999 and it is woefully underpowered. And the cheapest desktop Mac, the Mac Mini, is $799, again in a non-ideal configuration. Even a 20 percent price drop to each would keep them out of the entry-level part of the market. But would that be enough to trigger a Mac sales renaissance?

I don’t see it. But the advantages to the Apple ecosystem are real, and with macOS getting more consistent with the portable platforms and Apple ramping up its privacy messaging, I guess anything is possible.

Electron vs. React Native

dtatgenho asks:

With Electron proving to be an increasingly bloated resource hog, and with Microsoft announcing the new Xbox beta app has been rebuilt from Electron to React Native (specifically citing performance issues as a big reason), do you see Visual Studio Code at some point also getting rebuilt to ditch Electron?

I never understood Microsoft’s embrace of Electron, but they’re actually using it more now in some cases, not less. What’s interesting is that the extensible and cross-platform Microsoft apps—Teams and Visual Studio—are built with Electron. I guess I see them sticking with it.

I would think all the moving parts around the VS Code ecosystem (extensions, themes, console integrations, etc.) would make such a move a much more complex undertaking compared to the Xbox app, but I don’t know enough about the differences between Electron and React Native. Is React Native different enough from Electron that such a move would be a breaking change for VS Code’s existing ecosystem?

Yeah, I don’t know either. This is something I’m actively researching now.

Apple TV+ and a local dining experience

sabertooth920 asks:

Is the upcoming Foundation show reason enough for you to turn on Apple TV+?

Yes.

I grew up on Asimov and his Foundation and Robot books and short stories in particular. And I always think back to seeing the first Jurassic Park movie and realizing that it was now possible to make anything into a movie. The Lord of the Rings is a good proof point, but I prefer binge shows to big movies, so this ideal. It could literally go on forever, they’ll never run out of story.

And for whatever it’s worth, there are some great shows on Apple TV+. If you not seen them, I can recommend Servant and For All Mankind highly.

Have you experienced Shady Maple, since being in PA?

Yeah, and we were not impressed at all. We live in a place that has amazing and abundant food sources, and Shady Maple was like a giant Golden Corral or whatever. It was terrible.

WOA vs. macOS on Apple Silicon

OldITPro2000 asks:

How is it possible that Microsoft is seemingly able to compile a full ARM version of Office for Mac, but has yet to offer the same for Windows? How is it possible that Apple is able to emulate 64-bit x86 code on ARM, but Microsoft can only do the same for 32-bit code? How is it possible that Apple appears to have a hypervisor for ARM already while Hyper-V for WOA was released in preview only last month? Is there even a roadmap for WOA enhancements?

Speculating, my guess is that this has to do with the respective strategies here. Remember, macOS is literally transitioning from Intel to Apple silicon. It’s not a port that will appear on some devices, as with Windows 10 on ARM (WOA). And that reality has a few side effects. If you’re a Mac developer, you can’t ignore this, so you will be transitioning your apps to be native on Apple silicon. It has to happen.

On WOA, the picture is very different. WOA has to account for some tiny sub-1 percent of market share in the Windows PC market, so porting your apps to the platform is less important and easily ignored. This is particularly true for legacy/Win32 apps like Office, where it’s not like you can check a box in a dialog and just compile it to ARM. These ports would require months and months of work and, for something as complex as Office, may literally be impossible. So the focus on WOA is on improving emulation performance. Existing apps have to just work.

Knowing that it was transitioning to its own chipsets years ago, Apple set in motion multiple parallel efforts. They made macOS and the apps running on it 64-bit only. They improved their chipsets to run macOS and those apps better. They improved Xcode so that it could compile natively to the new universal app type, with support for both binaries. Apple was going to war. It mobilized that entire part of the company.

With WOA, none of that is happening, or if it is, it’s just a tiny part of the Windows org. There is no pressure for this to be perfect on day one and no incentive for developers to even think about it, let alone make the effort.

Apple has been working on this longer than Microsoft, by the way. It’s just that Microsoft went public with it first.

Microsoft custom chipsets

MartinusV2 asks:

Since AMD and Microsoft have a good relationship for building customs ships for xBox and Surface Laptop 3. Do you think Microsoft would go farther with a custom CPU that could potentially be less compatible with regular CPU? Could not Microsoft dictate, demand changes to Intel / AMD for better integration of Windows with the CPU?

This is the dark side to Microsoft’s partner-centric strategy. They can ask, and they can collaborate, but they can’t really go as far as Apple can in literally designing their own custom CPUs/SoCs. It’s not coincidental that Microsoft is partnering with the three biggest non-Apple silicon makers (Qualcomm, Intel, and AMD) on custom CPUs. I’m sure they’re playing each off of the other.

But you raise an interesting point. Could this Apple move trigger a deeper collaboration between Microsoft and at least one of these chipset makers—and AMD would be ideal—to create truly optimized chips for this new mobile future? Maybe this will be the push they need to get this done.

Going it alone

bschnatt asks:

I have long thought that Microsoft should STOP selling Windows to 3rd party manufacturers and just refine it to work on their own hardware. This can’t be illegal, because Apple is already doing this.

Well, such a decision would destroy the PC businesses at HP, Dell, Lenovo, and several other companies, and there would certainly be legal challenges. Also, Microsoft accounts for some tiny percentage of the PC industry and hasn’t shown it can scale from single digit millions of PCs per year to over 250 million units per year.

When you think about it, it would make more sense to kill Surface than to stop licensing Windows to third parties. (And no, I’m not advocating such a thing.)

Doing this would make it *vastly* easier for them in so many ways. They’d only have to test code changes on their own hardware, and could do the same fine-tuning that Apple is doing. It wouldn’t be so good for the many people Microsoft would lay off who are doing all that now-unnecessary coding, testing and documenting, but it seems like the obvious thing to do considering the nightmare they’re putting themselves and us through right now.

But we would also collectively lose the diversity and choice, not to mention the pricing competition, that makes the Windows ecosystem so special. I just don’t see this happening. You go to war with the army you have. This is the army we have, and it’s winning. The Mac is still just 6.8 percent of the PC market. Let’s not lose our minds here.

I’m not advocating this per se, but what just jumped into my mind was that maybe Microsoft should instead just open source Windows and let the PC makers do what Android hardware makers do and customize the user interface to their needs. We’d have a bunch of software compatible versions of Windows out there, and our PC choices would be based in part of the software we got, and not just on the hardware. Surface could be the Pixel version, with a clean software image, and the PC makers could go to town.

Given what they do with crapware, this might seem fanciful. But Windows licensing costs is what drove them to that.

Removing apps from the Microsoft Store

hrlngrv asks:

Does MSFT have clear policies for removing software which isn’t maintained from its app store? There are some examples of Desktop Bridge titles in the store which are a few versions behind the same titles from other sources.

Not that I’m aware of, and I could imagine the outrage if apps just started disappearing. Cleaning up the store is a great idea, but I’d go with a quality/value bar: At this point, it should be about quality not quantity. Quantity is not working. (More crap is still crap, and still a negative.)

Microsoft v. Windows

brothernod asks:

If Microsoft abandoned Windows (day today or 5 years from now) do you think they could maintain their place as one of the top 5 most valuable companies in the world?

Today, no. I’m guesstimating that Windows is responsible, conservatively, for about $8 billion in revenues every quarter. They can’t just walk away from that.

If Microsoft said, hey, we’re going to open-source this thing over time, we have a 5-10 year transition in place, and we’re going to place it in the stewardship of some organization similar to Mozilla that it would fund, then yeah, maybe they could maintain that, assuming that their cloud revenues grew during that time to replace the lost Windows revenjues.

I’ve been day dreaming about Apple’s ARM move and wondering if they hypothetically have a significant performance and/or margin benefit over Windows, could they massively grow their laptop/desktop market share? Do you think they’ll move in to selling servers?

No, Apple tried the server thing in the early days of OS X, and that’s just not their expertise. They’re making a killing on hardware and consumer services.

Honestly, I really wish my Surface Pro was as fast as my $300 iPad feels in general and in games. I’d find an iPad Pro that actually let you develop software would be incredible.

The day that Xcode and Android Studio run on iPadOS is going to be a big day. There’s no way it doesn’t happen. Eventually.

Xbox Series S and Series X pricing

spacecamel asks:

Just finished FRD from yesterday where you talked about the Xbox Series “S” and was wondering what you thought the price will be for the “S”. I am trying to understand the value of the “S” vs. a current “1-X”. It seems a refurbished “1-X” will be cheaper for the same performance. Any I missing something new with the “S”?

My general thought on Xbox pricing is that Microsoft will undercut Sony. And if they announce first and Sony comes in lower, Microsoft will adjust accordingly. They’ve made it clear that they will not lose on price, and Sony has literally said that they will market the value of what they’re selling, which to me indicates higher prices.

With regards to the actual pricing, a little under a month ago, I tweeted that my guess was that Xbox Series X would cost $499 at launch and that the PS5 would be $599. That still seems right to me. So Xbox Series S, which is roughly on par with the Xbox One X, would then come in around $349 or perhaps even $299.

These are just guesses, of course.

The value of a Series S is that it allows Microsoft to remove the Xbox One from the market and proceed forward with a new family of products with the same architecture.