Yep, that is correct.
What’s the status here? I’m ready to buy as soon as users of the initial boards are able to confirm success. (This seems like a relatively low-risk project given the mostly mechanical / pass-through nature as I [limited] understand it - but this is my production machine so I’m not able to ~alpha test)
Thanks for your work!
@Josh_Cook - saw your reply on my other thread about framework curating stuff like this; replying here just to keep that about the general problem.
I’m still confused. Totally get things take time, I’m just hoping to understand what the actual status of this is - is there a concise way to describe what is/isn’t done, and whatever your current (spitballed) guestimate of a realistic ship date?
1 Like
@Josh_Cook I have standard expansion bay shell Framework
If I want to use your m2 expansion bay module, then I need to order your module and Graphics Module Interposer Framework
Am I correct?
Yep you are.
Hi Josh, given that it appears that Framework will be selling a similar module, are you still planning to produce yours? I would like to support you (edit to add - I have ordered your module and have no plans to cancel that order) - I may look into getting one from Framework in addition to the one that I have ordered from you. I have to check how many spare nvme drives I have laying around. Thank you, and have a great day!
Similar with one glaring limitation:
It’s only up to 2280.
2280 is the most common size so I don’t see the problem. Are you expecting one of the less common size such as 22110 ?
I was hoping / expecting to have 22110 support in the expansion bay, if it’s physically possible. Being able to run enterprise drives that have PLP (power loss protection) in a consumer laptop would be a killer feature, and those are (as far as I can find) exclusively in the 22110 form factor.
Drives with PLP tend to have substantially better write performance compared to drives without it, which is a major consideration with developer/sysadmin workloads such as running databases, multiple VMs, etc.
Yea, I’ll produce them, but probably only do 1 batch.
1 Like
I have a bunch of 22110 drives, so I do see a problem.
1 Like
Unless there are enterprise 22110 drives with PLP and working apst now, all that would kill is battery live.
I don’t know why anyone needs PLP in a laptop, nor am I in the situation where I’d care. But this is Framework. Isn’t this community in the spirit of “what more can I do with this hardware [outside of the box that it came in]?”
Nothing is stopping you from connecting them with adapters (or your own pcb) and stuff but it makes sense to not waste valuable space to support hardware that would not make much sense in the laptop shell anyway.
I have seen this mentioned several times in various topics on here, and I agree with you. Why have PLP when you have a built in UPS in the form of the internal battery? Maybe they don’t charge their battery and rely on mains power only.
But in the context of this thread,
- The space exists.
- Doesn’t seem like a monumental addition:
Yeah, but this statement is the very kind of sentiment I was addressing in my sentence you quoted. Creative license is an encouragement to think outside the box―or in this case, the laptop shell. There are all sorts of weird ideas in other threads. I don’t see the appeal for my uses, nor do I think anyone else has as any stock in those ideas. But some idea is going to make sense to somebody.
One could argue that many ideas exact development, production, and support costs. But what are the costs of supporting an extra 3 cm that’s earning your negative feedback? I’m curious to hear your side of the story.
A lot stops a lot of people. I’m not a hardware engineer. Rather, I work with the more squishy innards that reside in storage and CPU registers. But advocacy is also a valid way to engender support for something. How does anyone with the means to make something know what the market wants if the market doesn’t speak?
On further looking into it it looks like the reference design even supports them, not sure what’s going on there. I initially looked at the reference design assuming it only supports 2280s in which case it looks like there really isn’t any extra space but looks like that’s already with bigger ssds in mind.
In this case there is no cost as it’s already done and filling out the already available space so there is no need to find an extra 3cm that comes with drawbacks. I am all for add extra options if convenient and not making the intended use worse.
If however doing 2280s side by side would fill out the whole width, supporting 22110s would require a stacked diagonal setup which is pretty much universally worse as long as you are not using 22110s and I’d be against that.
You don’t need to be one, all you need (assuming hypothetically the adapter only supported 2280s) is a pair of wired m.2 extensions you can buy of the shelve and you can put your (hypothetically) oversized ssds wherever you want.
I don’t think I ever said advocacy is bad, apart from vocal minorities having a disproportionate impact on the silent majority sometimes it is a very important tool over all.
I figured out why 22110 was not present on Framework’s version, because of a few areas with 0mm height clearance. Shown in the screenshot below has the maximum height in those areas, this also means I will not be able to support 22110 M.2 devices, but I’m not sure why there is the limited clearance in these particular areas.
I see U shaped H:0 zones on the lower left and right edges. Those are the ones?
Could you do one 22110 capable slot, one 2280? Maybe not ideal for those that would like a pair of matched 22110s, but hey, better than nothing.