I love AMD a lot - my desktop is a 5800 with 128GB RAM in part because I didn’t want to play silly market segmentation games with IOMMU/VT-D. The company is great, and Lisa Su has some rare brilliance.
But she’s not the only business savvy tech director in the room.
It would be malpractice for a company to promise two platforms that support the same software library; for Framework to take their team and slice it in two in order to support both Intel and AMD fans would basically be doubling R&D and gain nothing in the process; sure, they might pickup some customers on the way, but boards would need to be designed and QA testing would need to happen. It’s not easy, either; different bit layouts means testing against different levels of impedance and crosstalk, further complicating designs.
Framework needs to sell a widespread accessible notebook computer that is, at the very least, credible. They need to make the case to businesses as much as to users that this computer is appealing to their needs.
We don’t know what FrameWork’s numbers look like, but possible scenarios might go like this:
- They announce AMD-exclusive as their next gen. Sales die as people anticipate; others go an by a Ryzen system today. Some decide that Ryzen doesn’t have efficiency cores and were gonna buy Intel anyway.
- They announce both Intel and AMD. R&D is divided and the budget is increased. 90% of systems are Intel purchases; lots of Ryzen stock fills a crate in the warehouse and is quietly discarded in a few years.
- They announc Intel and AMD. Both are equally successful - 50-50 split. In the end there’s no real gain as no more computers are sold than what would have been sold.
- AMD utterly knocks it out of the park, crushing Intel sales. A warehouse is filled with Intel parts. R&D budget was doubled, but returns don’t.
There are few scenarios where FW comes out ahead, but if we’re thinking optimistically then perhaps FW’s addition of Ryzen-based boards manage to sell as much as Intel boards and also increases install base, resulting in incredibly high revenue. This assumes every buyer is perfectly informed and also has FW on their radar, and they’ve managed to make inroads to business sales, beating Dell, Lenovo, and Apple. Microsoft, who despite having dedicated business account sales teams, evangelists and Azure integration built-in to Windows, has barely cracked it, and MS offers Ryzen and Intel options (so does Lenovo and HP, while Dell barely has Ryzen on consumer models).
Long story short, if I were an investor and saw MS’s own limitations at the hands of their partners, I’d be weary about Ryzen being a selling point to FW customers (even if I knew better).
I’m not hating on the idea of FW making a Ryzen system, and perhaps they could shake things up by giving the market something new. But the reality is that by now we’ve been clamoring for an AMD model with vague promises like “I’d buy one if you made it”, when a handful of direct and non-committed promises aren’t business building blocks. Some of us, myself included, are fans who totally would drop in a replacement board. But given that the install base for most FW devices is about one year old, they’re not even likely to sell a giant number of boards, and there would need to be a compelling reason to get us to remove our perfectly good boards as it is, leaving new customers to be the majority buyers in a split niche market.
TL;DR: I want it too, but as an outside speculator, I think it’s far more likely that we’ll get an Fn lock indicator than see an AMD-based FW board this soon. Maybe in 7K/8K generation, but not likely for 6K (or whatever the upcoming notebook gen is).
Addendum:
Any approach that FrameWork makes will ultimately also beg the question about what to do with discarded motherboards. If FW did offer a shiny, new motherboard, regardless of platform, I’d also fully expect to see a FrameWork CyberDock - a new home for old boards with modular slots for power and gfx output, as well as a means for mounting Wifi antennas (perhaps an “extended EGPU mount” that’s internally routed to a GPU slot might get attention, too!). Assuming that old boards would be thrown out seems wasteful, contrary to FW’s mission, but they likely won’t have much secondary market value on their own, and the third option - buying a new laptop shell - is just buying another laptop with more steps. The i5s and i7’s are fantastic boards in their own right, and there is value to a functional board that can easily handle the needs of SteamOS with the option of slotting in a 1050-2070 eGPU, getting locked into a RAID storage friendly case, and heck, even racked in and attached to an iSCSI array and low-key running a hundred Docker containers from a 1u chassis (lets get creative with the “Wifi” slot while we’re at it!).
I know that for me, each of those scenarios are pretty compelling. I rather like the idea of firing up a new dedicated Linux server that uses half the space of a typical rack system, and could even see “HomeRacks” being a thing, and that might be enticing to those otherwise less likely to upgrade so soon after receiving their FrameWork computers. The HomeRack idea wouldn’t even need an Fn Lock indicator!
I don’t mean to shift the conversation too far away from wanting a Ryzen drop-in, but I think it’s something to think about: how do we answer that question in a way that aligns with the overall company mission around sustainability?