Question Stuck in a Rut: Seeking Advice on My Next Hardware Project

[NOR1NCO]

Hey everyone,

I'm hoping to get some advice from the community because I feel like I've hit a wall and could use some fresh perspective.

I love working on hardware projects, but I'm struggling to land on a viable idea that I can see through to completion. Here's a rundown of what I've been working on:

1. PCIe x16 to 8x M.2 NVMe Card: My first major project was designing a PCIe x16 card that could hold eight M.2 NVMe drives. While I managed to create a functional prototype, I ultimately wasn't satisfied with the performance, and the physical card was just too bulky. I decided to shelve it at the sample stage. Here's a fun fact though: someone actually copied my original design, mass-produced it, and is now selling it at a very low price. Honestly, I have no idea what they're thinking. This is hardly a high-demand product, so their effort will most likely be meaningless. It just adds to my confusion about what makes a project viable.
2. Custom NAS Build: Next, I shifted my focus to building what I thought would be a practical Network Attached Storage (NAS) solution. The hardware side was manageable, but as I mapped out the entire project, I realized the long-term operational costs, especially for high-volume network services after a potential product launch, were far more than I could afford. So, I had to abandon that idea as well.
3. Deep Dive into Server Architecture: More recently, I was fortunate enough to get my hands on incredibly detailed documentation for Intel's Alder Lake and Purley (Xeon Scalable) platforms. The information is fascinating, but the sheer scale and complexity of a project involving these platforms, especially Purley, is overwhelming for a single person. The costs for even a small-batch manufacturing run of anything Xeon-related are astronomical.

I've asked a number of people for their thoughts, but I haven't received any advice that has pointed me in a clear direction.

So, I'm turning to you all. Have any of you been in a similar situation? How do you come up with and validate your project ideas? Are there any niches or opportunities I might be overlooking? I'm open to any suggestions, no matter how "out there" they might seem.

What should I do next?

Thanks for reading.

 

[Lutfij]

This is hardly a high-demand product, so their effort will most likely be meaningless.
I've learned that there's a market for practically anything. There will be people who will buy ridiculous hardware as well.

Are you asking for guidance on how to move forward on the pictures project or projects in general? Just understand that if you've made something that doesn't exist in the field, then you're the guinea pig in this entire scenario. If there's someone else out there who has done it, there's a chance you can reverse engineer it to be better. Case in point, how Asus used to make better boards than Intel(if you look at the board from 2 decades ago).

Once you've established connections with, say, manufacturers, that's when you gain favorable prices.

My 2 cents on the matter.

 

[NOR1NCO]

This is hardly a high-demand product, so their effort will most likely be meaningless.
I've learned that there's a market for practically anything. There will be people who will buy ridiculous hardware as well.

Are you asking for guidance on how to move forward on the pictures project or projects in general? Just understand that if you've made something that doesn't exist in the field, then you're the guinea pig in this entire scenario. If there's someone else out there who has done it, there's a chance you can reverse engineer it to be better. Case in point, how Asus used to make better boards than Intel(if you look at the board from 2 decades ago).

Once you've established connections with, say, manufacturers, that's when you gain favorable prices.

My 2 cents on the matter.

Thanks for your 2 cents!
I do want to emphasize that my M.2 card is, in my view, 100% an absurd and obsolete product (being PCIe 3.0). This is especially true because, in my region, everyone can now buy affordable PCIe 4.0 M.2. That's why I'm still baffled someone would copy it.

You absolutely nailed my main difficulty, which is finding that balance you mentioned: the one between being a "guinea pig" for a totally new idea, and reverse-engineering something that already has clear market support. I feel stuck between "too high risk" and "too much competition."

And regarding connections, my experience is somewhat limited. I've had contact with a Broadcom agent and with Intel, but most of my other interactions have been with distributors, without forming any deep relationships.