History
My previous computers have often been refurbished or highly subsidized by the fact that the parts were at least >3 years old and didn’t have resale value standalone. It wasn’t till around the Zen launch (back when Kaby Lake was new) that I was in a position to buy a computer… for work.
As a result of collecting money, circumstances allowed me to finally go from my Core 2 Quad/[insert i3 chip that i bought for <$100] and iGPU/[insert the cheapest NVIDIA graphics card was still being supported and had HDMI] to a Ryzen 7 1800X and a Radeon RX 480.
I primarily needed the machine for (what I then envisioned would be on Linux) development and went Ryzen/Radeon for that reason, maximum price-to-performance and while I didn’t get the motherboard everyone was recommending (the ASUS PRIME X370, as it was out of stock everywhere), I found myself with a decent MSI Krait Gaming instead.
Where I screwed up
I found myself platform hopping (macOS, Windows, Debian-derivatives, Arch-derivatives, GhostBSD, even) as I was trying to figure out a workflow that best suited me (a combination of collaborative creative work and development where the ability to get into the weeds was a must) and three years after purchase, realized that Linux-based operating systems won’t do it for me. Windows requires me throw away everything I’ve learned on Linux and WSL isn’t making the cut for what I do.
Naturally, the only compromise between the app library of Windows and the *nix heritage of Linux, macOS… is where I feel at home, the ability to run creative collaborative software. So now we’re entering Hackintosh territory.
First screw up? I bought a Ryzen CPU. Now Ryzentoshing is mature to a point where most can daily drive it and calling it a screw up might even be dramatic but it’s definitely different from running an Intel CPU (Hackintosh work on OPEMU has been retired ever since macOS went “goodbye 32-bit” even though I need 32-bit macOS software) and I want macOS Mojave because I need to run legacy macOS software, the more recent versions feel too constrained for me.
Additionally, I really don’t like having to deal with Intel MKL related crashes in the middle of work or having to re-patch Discord after every update.
Second screw up? Timing. This wasn’t in my control, I needed a machine ASAP and got one, my new job depended on it and a slow machine wasn’t helping. I bought into Zen, not Zen 2 and didn’t benefit from architectural improvements, I was also an early adopter of DDR4 and didn’t benefit from improvements and price efficiency that subsequent memory modules got.
TL;DR: I was an early adopter and I had to pay the early adopter tax and got less performance for my money. I also picked the wrong chip for the job that I didn’t know I would be most comfortable doing when buying it
The Problem
I’ve been tolerating these inconveniences for as long as I can but I’m reaching a breaking point. VirtualBox is not a viable alternative to Hypervisor.framework (which will not work on Ryzen chips), I need Docker and VMWare Fusion and I have to be on Monterey as my job requires it (even though I love Mojave, Xcode 13 doesn’t)
I’ve tried running a thin Linux-based hypervisor and running macOS in a KVM instance but my motherboard’s IOMMU grouping makes it less conducive for actual use (even with patching the kernel to split each device into its own IOMMU group) and audio crackling (even with patching QEMU) is a consistent problem despite owning an external USB audio device.
With DDR5 on the horizon (well, it’s out but it’s still time before it becomes optimized cost-effective), I see myself in the same position as I was in when DDR4 came out and I was compelled to get a computer.
Back then, Zen vs Kaby Lake was the start of something interesting in the desktop computer space. Now, that year is 2023, the year when Meteor Lake will come out (and hopefully, by then DDR5 will become the norm and the economies of scale would make it affordable) and the year when Apple Silicon would reach levels of maturity that would make it worth it for me to make the jump (the M2 Max is where I’m comfortable giving Apple my money, even then, only for a Mac Mini, not for a Macbook Pro)
Conclusion
I don’t mind spending a pretty penny on good computers but I can do so only if I spend once per generation. There is no used market for high end components here so “buy now, flip before launch and get the latest gen” isn’t an option for me.
Hackintoshing is getting old and the future for me seems to be within the Apple ecosystem (for better or for worse), I want to give Apple my money but Apple Silicon isn’t mature enough for the work that I do and their devices aren’t easily repairable (which is why I turned to Hackintoshing in the first place, I need hardware I can repair), their SoC being a single point of failure is a dealbreaker for me. I already pay the developing country tax buying an Apple product, having it fail on me due to Apple’s engineering deficits (Rossman’s videos on Macbook repairs do not inspire any confidence in me) just ramps up the hurt.
Despite all that, I might bend if matured, professional-grade Apple Silicon (basically, the hypothetical M2 max) comes to the Mac Mini. But that’s 2023.
Alder Lake is out but I see myself repeating the same mistake as I did with my first big desktop computer with Zen/DDR4 if I were to buy today. I bought at the beginning, when prices are highest and performance is lowest (compared to the rest of the DDR* lifecycle) but had no choice then.
Meteor Lake is scheduled for 2023. It’s the Zen 2 of the generation I bought into. The DDR5 RAM I buy now is the DDR4 RAM that I bought then which costed a lot for a lot less performance.
What do I do?
This is my first post. Sorry for the walls of text.