3nm Intel chips

[pshufd]

Just market share and applications.

I wouldn't mine a x86 core or two, but I don't think it will happen. What I'm hoping for is some kind of peripheral that would run x86/64 Windows.

Absent that, I need a computer to run x86 or my software vendors to port to M1. I don't see one of them doing that in the next decade. The other could do it easily but it's not in their current plans. So I use two machines now. I'd really like to not have to do that when mobile.

 

[robco74]

Back in the day, there were expansion cards that put an x86 chip into a Mac to run Windows. But that would only work on the Mac Pro. Depending on performance need, you could pick up a SFF PC or compute stick, or look to cloud-based solutions.

If Qualcomm can come close to ASi performance next year, as they claim, and MS does go all-in on supporting ARM, then some software vendors may simply have no other choice than to support the platform.

We're not even a year into a two-year transition. There are still a number of development tools that aren't yet ready for ARM. That will have impacts downstream. I hope that by this point next year, that situation will have improved, both for macOS and Windows. There will be some holdouts, and hopefully others will see opportunity to step in and pick up the slack. However, there will undoubtedly be some users who will have no choice but to switch platforms and remain on legacy systems. Omelet, eggs, and such.

 

[pshufd]

Back in the day, there were expansion cards that put an x86 chip into a Mac to run Windows. But that would only work on the Mac Pro. Depending on performance need, you could pick up a SFF PC or compute stick, or look to cloud-based solutions.

If Qualcomm can come close to ASi performance next year, as they claim, and MS does go all-in on supporting ARM, then some software vendors may simply have no other choice than to support the platform.

We're not even a year into a two-year transition. There are still a number of development tools that aren't yet ready for ARM. That will have impacts downstream. I hope that by this point next year, that situation will have improved, both for macOS and Windows. There will be some holdouts, and hopefully others will see opportunity to step in and pick up the slack. However, there will undoubtedly be some users who will have no choice but to switch platforms and remain on legacy systems. Omelet, eggs, and such.

No, they don't. Software vendor can just say: if you want to run our software, then you have to run Windows x86.

I used Medved Trader ten years ago. He got asked regularly to do a Mac version. This is from his FAQ.

No. Medved Trader is a Windows application and requires Windows 7 or higher (Windows 10 is recommended)

If you need the software, then you buy the hardware that it runs on. Period.

 

[robco74]

No, they don't. Software vendor can just say: if you want to run our software, then you have to run Windows x86.

I used Medved Trader ten years ago. He got asked regularly to do a Mac version. This is from his FAQ.

No. Medved Trader is a Windows application and requires Windows 7 or higher (Windows 10 is recommended)

If you need the software, then you buy the hardware that it runs on. Period.

Then that is an opportunity for someone to make a competing application. That's how capitalism is supposed to work, so I'm told. If WoA catches on, the dev will either update their app, or lose customers.

 

[bobcomer]

Back in the day, there were expansion cards that put an x86 chip into a Mac to run Windows. But that would only work on the Mac Pro. Depending on performance need, you could pick up a SFF PC or compute stick, or look to cloud-based solutions.

The Amiga had those too, both internal and external. It actually worked pretty well and that's what I was thinking about.

A compute stick is more for plugging into a monitor rather than another computer, no interaction basically. There's no reason it couldn't be done, and at a decent price.

If Qualcomm can come close to ASi performance next year, as they claim, and MS does go all-in on supporting ARM, then some software vendors may simply have no other choice than to support the platform.

That wont help me at all, I have to run old business applications, some of them custom. No way Microsoft's going to ditch x64/x86 in my lifetime, so a Windows machine it is -- unless I can cheat and run x86/64 Windows on my M1 Mac.

 

[pshufd]

Then that is an opportunity for someone to make a competing application. That's how capitalism is supposed to work, so I'm told. If WoA catches on, the dev will either update their app, or lose customers.

Jerry Medved wrote QuoteTracker in the 1990s and it was unique. It was broker-neutral so that you could simply switch brokers if one went down. It was the same interface if you used multiple brokers. He sold it to Ameritrade around 2000. Ameritrade put it out but didn't maintain it. Medved started writing a new version around 2007 as his non-compete would expire a few years later. And he launched it around 2009 or 2010.

There's nothing else on the market like it. It's a fairly small Windows executable, lightweight and fast. I installed Windows via Bootcamp to run it.

He has the multi-broker knowledge of how to hook into their streaming and trading systems around the world and that knowledge seems rare.

 

[leman]

There are still a number of development tools that aren't yet ready for ARM.

What tools are you still missing? I think most of the stuff works quite well. There is still no official GCC support, but I official patches exist and appear to work reliably.

 

[pshufd]

The Amiga had those too, both internal and external. It actually worked pretty well and that's what I was thinking about.

A compute stick is more for plugging into a monitor rather than another computer, no interaction basically. There's no reason it couldn't be done, and at a decent price.


That wont help me at all, I have to run old business applications, some of them custom. No way Microsoft's going to ditch x64/x86 in my lifetime, so a Windows machine it is -- unless I can cheat and run x86/64 Windows on my M1 Mac.

I think that you can run QEMU but I'd guess that the performance would be horrible. Especially since you'd probably want a decent amount of RAM in your virtual machine. I plan to give it a try on an M1X.

 

[JMacHack]

Jerry Medved wrote QuoteTracker in the 1990s and it was unique. It was broker-neutral so that you could simply switch brokers if one went down. It was the same interface if you used multiple brokers. He sold it to Ameritrade around 2000. Ameritrade put it out but didn't maintain it. Medved started writing a new version around 2007 as his non-compete would expire a few years later. And he launched it around 2009 or 2010.

There's nothing else on the market like it. It's a fairly small Windows executable, lightweight and fast. I installed Windows via Bootcamp to run it.

He has the multi-broker knowledge of how to hook into their streaming and trading systems around the world and that knowledge seems rare.

Sounds to me like a business opportunity. If demand for M1 Macs is strong enough, then some up and comer could theoretically write a similar program for Mac.

People throw money at dumber startup companies.

 

[pshufd]

Sounds to me like a business opportunity. If demand for M1 Macs is strong enough, then some up and comer could theoretically write a similar program for Mac.

People throw money at dumber startup companies.

Are you volunteering? It took him three years to write it and beta test it.

Lots of people say that someone should do this. But I used this back then and Medved Trader is the only product that I'm aware of that does this.

 

[JMacHack]

Are you volunteering? It took him three years to write it and beta test it.

Lots of people say that someone should do this. But I used this back then and Medved Trader is the only product that I'm aware of that does this.

I’d volunteer if I wasn’t stupid. I’m incapable of writing such a program, but it’s silly to say a team of people couldn’t write a similar app. I’m guessing that it hasn’t been tried yet.

 

[bobcomer]

I think that you can run QEMU but I'd guess that the performance would be horrible.

I've tried, and networking isn't good enough to even be useful. If you run an older version of Windows, the performance is only bad -- not horrible.

Especially since you'd probably want a decent amount of RAM in your virtual machine. I plan to give it a try on an M1X.

That's in my plans as well. Hopefully they can get the networking stack working better. Right now it's just a pho-NAT implementation, not useful for a local LAN.

 

[JMacHack]

Addendum to what I posted. We’re into less than a year of this transition, it’s been 15 years of x86 dominance and it’s not going to change overnight.

 

[pshufd]

I’d volunteer if I wasn’t stupid. I’m incapable of writing such a program, but it’s silly to say a team of people couldn’t write a similar app. I’m guessing that it hasn’t been tried yet.

Who's going to pay the team of people while the software is being developed? Who's going to pay for the hardware and software? Where are you going to get the expertise to write this software. Who's going to get the contracts done with brokerage firms around the world to access their streaming quotes and trading API?

Software developers like to get paid while they are working unless they are moonlighting.

Software developers don't run on magical thinking.

 

[pshufd]

Addendum to what I posted. We’re into less than a year of this transition, it’s been 15 years of x86 dominance and it’s not going to change overnight.

I used to work for a company and they made minicomputers and mainframes and had a ton of customers for their software. This was in the 1990s. Software is still being run on that architecture. Fidelity Investments still runs COBOL on mainframes. I don't think that they use punched cards but there are companies that use old architectures and hardware because the cost of transition can be really expensive. Or they failed so many times trying to do previous migrations.

 

[JMacHack]

Who's going to pay the team of people while the software is being developed? Who's going to pay for the hardware and software? Where are you going to get the expertise to write this software. Who's going to get the contracts done with brokerage firms around the world to access their streaming quotes and trading API?

Software developers like to get paid while they are working unless they are moonlighting.

Software developers don't run on magical thinking.

Allow me to clarify:

Yes, I understand that people like to be paid money for programming.

I also understand that this is a non trivial task.

And I understand that there would be hoops to jump through.

But surely it is not beyond the realm of possibility, that a software company could develop such a program. And the capital could be sourced to do so from investors.

There have been other startups where money was thrown at them and they produced far less (Nikola and Lordstown Motors come to mind)

Is there hurdles? Yes, and I understand that. Are they impassable? No. And maybe in time such a program might exist. Again, we’re less than one year into the ARM transition.

 

[JMacHack]

I used to work for a company and they made minicomputers and mainframes and had a ton of customers for their software. This was in the 1990s. Software is still being run on that architecture. Fidelity Investments still runs COBOL on mainframes. I don't think that they use punched cards but there are companies that use old architectures and hardware because the cost of transition can be really expensive. Or they failed so many times trying to do previous migrations.

I would posit that user-facing software is more likely to change than back end software. But I’m not sure we would agree on this.

 

[pshufd]

Allow me to clarify:

Yes, I understand that people like to be paid money for programming.

I also understand that this is a non trivial task.

And I understand that there would be hoops to jump through.

But surely it is not beyond the realm of possibility, that a software company could develop such a program. And the capital could be sourced to do so from investors.

There have been other startups where money was thrown at them and they produced far less (Nikola and Lordstown Motors come to mind)

Is there hurdles? Yes, and I understand that. Are they impassable? No. And maybe in time such a program might exist. Again, we’re less than one year into the ARM transition.

That it hasn't been done in the past decade implies considerable difficulty or lack of investor interest. There's plenty of startup money available but apparently there is no interest. At any rate, that there is no competition indicates that it isn't being done.

 

[robco74]

What tools are you still missing? I think most of the stuff works quite well. There is still no official GCC support, but I official patches exist and appear to work reliably.

There's a rundown on the YouTubes. Some will run under Rosetta, but still waiting for native versions. AFAIK, there's still no officially supported JDK.

I should be OK personally, but I'm still rocking Intel with no immediate need to upgrade.

 

[robco74]

The Amiga had those too, both internal and external. It actually worked pretty well and that's what I was thinking about.

A compute stick is more for plugging into a monitor rather than another computer, no interaction basically. There's no reason it couldn't be done, and at a decent price.


That wont help me at all, I have to run old business applications, some of them custom. No way Microsoft's going to ditch x64/x86 in my lifetime, so a Windows machine it is -- unless I can cheat and run x86/64 Windows on my M1 Mac.

I would think that unless you need strong GPU performance, that something like an Intel NUC would be able to run older Windows apps just fine. Likewise, if the emulation layer is good enough, then running Windows in a VM should work well too. Eventually the ASi Macs should be able to run apps quickly enough through sheer brute force. M1 is limited to 16 GB of RAM and 4 performance cores, but I imagine future chips will be able considerably more capable. Intel will be supported in macOS for some time, so there's still plenty of opportunity to get everything sorted.

I do wonder if MS will hold to the CPU requirements for Windows 11, or if the blowback will be enough to get them to reconsider. I can see them halving the ecosystem again, like it was back in the Win 9x/NT era. Some folks will need backward compatibility for some time, but many consumers could do with a version with less cruft. Users of legacy software should be able to get by the security updates and major bug fixes.

 

[dgdosen]

Absent that, I need a computer to run x86 or my software vendors to port to M1. I don't see one of them doing that in the next decade. The other could do it easily but it's not in their current plans. So I use two machines now. I'd really like to not have to do that when mobile.

We're all free to choose software that doesn't react to the market in a decade's time.

No, they don't. Software vendor can just say: if you want to run our software, then you have to run Windows x86.

I used Medved Trader ten years ago. He got asked regularly to do a Mac version. This is from his FAQ.

No. Medved Trader is a Windows application and requires Windows 7 or higher (Windows 10 is recommended)

If you need the software, then you buy the hardware that it runs on. Period.

The countless millions of users of this software will have to use Windows 10. What does he say about efficient markets?

 

[pshufd]

We're all free to choose software that doesn't react to the market in a decade's time.

The countless millions of users of this software will have to use Windows 10. What does he say about efficient markets?

Windows 10 works for most people.

Running on macOS M1 is a want for me. Clearly it isn't for the vast majority of users.

It would be nice if Apple fixed it but it's not in their financial interest to do so.

 

[bobcomer]

I would think that unless you need strong GPU performance, that something like an Intel NUC would be able to run older Windows apps just fine.

I don't need GPU performance at all, but a NUC is a separate PC, and this would be for traveling and that's what I bought my MBA for. I'm wanting a NUC that is small enough and can just plug into my MBA. Say a USB<-Ethernet<-headless NUC, and I'd connect to it via RDP. (for the times when there isn't a decent internet connection.) I'd also like to be able to turn my PC's at home off when I'm gone, rather that keeping them powered so i can use them remotely.

Likewise, if the emulation layer is good enough, then running Windows in a VM should work well too. Eventually the ASi Macs should be able to run apps quickly enough through sheer brute force.

And someone writes a good emulator!

I do wonder if MS will hold to the CPU requirements for Windows 11, or if the blowback will be enough to get them to reconsider.

I think that will be the case. The PC vendors messed up on this one, the hardware needed has been suggested for quite some time! Some do it well like Lenovo, others, not so much. Anyway, it's not really a biggie for me, the ones I want to run W11 already have the needed hardware and I tend not to upgrade older PC's OS-wise.

Users of legacy software should be able to get by the security updates and major bug fixes.

Business users get by with far less than that, especially with custom software. It's just not cost effective to rewrite things.

 

[AgentMcGeek]

People also tend to forget that TSMC has massive expansion plans in Taiwan, the US (namely Arizona) and a bunch of European countries. By 2024 their production capacity will have expanded to better satisfy the increased demand in silicon.

 

[thekev]

I would think that unless you need strong GPU performance, that something like an Intel NUC would be able to run older Windows apps just fine. Likewise, if the emulation layer is good enough, then running Windows in a VM should work well too. Eventually the ASi Macs should be able to run apps quickly enough through sheer brute force. M1 is limited to 16 GB of RAM and 4 performance cores, but I imagine future chips will be able considerably more capable. Intel will be supported in macOS for some time, so there's still plenty of opportunity to get everything sorted.

I do wonder if MS will hold to the CPU requirements for Windows 11, or if the blowback will be enough to get them to reconsider. I can see them halving the ecosystem again, like it was back in the Win 9x/NT era. Some folks will need backward compatibility for some time, but many consumers could do with a version with less cruft. Users of legacy software should be able to get by the security updates and major bug fixes.

They aren't terrible, but at least some generations have significant fan clogging problems, and the fan is below the motherboard. They're also far from silent with the stock case. I would think twice before buying one.