> "...: It sounds like the key feature will be 'more': a faster CPU and faster IO, rather than new features."
Raspberry Pi Holdings is a embedded systems manufacturer for pity's sake; we don't need more from them, we need less. [EDIT] A faster Raspberry Pi 6 is encroaching on the territory of the Intel N150 and its successors and mainstream Linux distributions and that is a battle they would lose in terms of price and performance.
Give us a Raspberry Pi Zero 3W with proper sleep states to reduce sleep power consumption, lower idle power while awake, and 1 GB of RAM even if it doubles the price.
Maybe a tick-tock release cycle (one with new features and some speed, the next with the ~same features and more speed) is where they're headed, and maybe that makes sense. They wouldn't be the first.
I'd love to see even-lower-RAM versions, though. Most of what I use Raspberry Pis for at home for is not RAM-hungry at all.
My Pi4 network router has 2GB because that was the smallest/cheapest version at release when I got it, but the system itself consistently only uses about 64MB of RAM. It'd do perfectly well and have a ton of breathing room with just 128MB of RAM (which will never happen, but if it did happen...).
I suspect the Pi4 that I use as a set-top box with Kodi would be fine with 512MB.
I've used Zero Ws for all kinds of things over the years and never felt RAM-starved with their little 512MB of RAM.
So I'm learning towards 512MB.
But sure: 1GB options would also be fine even if it does double the price. Our comments serve to demonstrate that there's room in the marketplace for different SKUs with different memory capacities. :)
I think exactly the opposite: we have no shortage of embedded crap we can buy; what is useful is dismembering intel. It would be better if the pi were risc v but this will do for now.
Raspberry Pi isn't in direct competition with N150's.
Their niche is the industrial/embedded space. For that market, power consumption doesn't matter. What matters is that each model is guaranteed to be available till a specific date.
Ideally each RPi generation should keep the same price (or lower now that it's gotten so high) but with better performance. If they can't do that they just shouldn't create a new generation.
Is there some serious astroturfing going on with the N100/N150, or am I just jaded?
I have a bunch of old intel atom boards laying around. The Intel Compute Stick (TM) burnt out its flash root drive in a few months. The C2000 board I had burnt out the clock pin to drive the bios. I have a Clover Trail with a PowerVR GPU (I thought I was getting an intel GPU because it was branded Intel Graphics or similar, but nope!) that lost Windows support very quickly after launch, and has no GPU drivers for any other OS.
Instead of being fooled 4 times in a row, I looked into using an N150 for a NAS, but this time I held off a bit until after launch so I could research it first.
Lo-and-behold, they all have crazy PCIe / memory subsystem data corruption issues. I guess there are some chicken bits for the OS developers to set if the kernel can stay up long enough after boot without a panic.
Why would anyone buy this for a NAS / embedded use case?
I wonder how they are positioned now in the market.
During covid I wanted a small low power always on server. I thought about Raspi, but at the time it was expensive and I went with an intel nuc, for a similar price.
Now if I wanted to do hobby electronics, I heard I should look into esp32 or stm32..
I use a Pi 4B as a 24/7 home server in a country where domestic electricity is expensive (and worsening).
Each Pi release is more powerful, but uses more energy. I found the Pi 4B to be the sweet spot for me, because it is the earliest model to support USB booting, gigabit Ethernet, and offer > 1GB RAM.
Perhaps a used one would fit your purposes and budget?
I currently use it to run PiHole, serve media via SMB, host Postgres & Redis, and run some custom written Dockerized apps. Home Assistant to possibly follow, too. The current load seems reasonable in htop, but I haven’t looked into burst scenarios.
I'd be interested in seeing them partner with some of the other interesting and open-ish players, like Pine64 or Radxa, to push for a more standardised, unified SBC landscape.
But that's not as good for PR as "bigger, faster, better" even if that comes with the problems you mentioned.
> Raspberry Pi Zero 3W with proper sleep states to reduce sleep power consumption, lower idle power while awake, and 1 GB of RAM even if it doubles the price.
Yes, that would be kind of a dream device, perhaps also if it could suspend the os when asleep so it doesn't have to boot every time but I guess that might the standard way of doing it.
A refresh for a Pi Zero with real sleep. lower power consumption and moving away from GPIO would be awesome. Stuff like a dedicated i2c connector, PoE support and display over USB-C.
M5Stack and other esp32 based ecosystems rule the world for MCU so the Zero just need to be linux SOC with all the fancy bells and whistles with good driver support
It's not $0.50 extra. It's $0.16 extra for a USB-C port assuming you bought the USB-C port on mouser at 10k quantities and threw the micro USB port away.
I just don't get it. Anyone who wants to save a few pennies just buys the chip directly. Their Pico board is primarily for prototyping and one off products, where quality of life is everything and 16 cents is nothing. The adapter cable probably costs more than the amount they saved. That's a dick move.
You can at least buy USB-C boards from other vendors since they sell the rp2040/rp2350 separately. If you want wifi it gets a little more complicated unfortunately.
Would love to see actual security focused hardware/software features, like full OP-TEE, fTPM (or a more ideally a real physical TPM), and similar. For example, so that the OTP isn't the only way to store a disk encryption unlock key.
The existing secure boot mechanisms aren't bad, but allowing for more than one public key hash in OTP would be nice, too.
These kinds of things are expected to be on modern embedded SOCs and SOMs now.
The 8GB Pi 5, at $170 [1], is encroaching on Jetson Orin Nano Super's $240 price point [2]. But the Jetson has a faster CPU (newer a78ae cores rather than a76) and, obviously, a whole-ass GPU.
I have bought an rpi at every generation. And I still have yet to find an actual use for them.
Everything they do from a compute perspective is just better with a mini pc or old laptop with a mobile spec chip.
Everything they do from a programmability perspective is just better with a microcontroller specific to the task.
I just don't see the actual market position for these things. They were supposed to be a cheap board, but you can't actually buy them cheaply because the vendors upcharge so much.
The only way I'll buy another raspberry pi is if they come with a power supply that's guaranteed to work with them. I got tired of the random reboots in the night and replaced my media center/NAS with an old Nuc.
I run the media lab at one of Europe's must prolific art universities. The variant I tend to use most is the 3B+.
Reasons:
- full sized HDMI connector
- headphone connector
- good bang for the buck
If I had one wish for any new product in the Raspberry line it would be: Do the Raspberry Pi 3++ or something. Same thing. Faster, but with USB-C power connector, 4K Video resolution, 2× USB-C I/O, 2× USB-A peripherals and maybe M.2 support.
I recently found out about the Radxa Dragon Q6A. A Qualcomm chip with faster CPUs, a good GPU, a DSP and AI accelerator, and a hardware video encoder seems very compelling. It even supports Windows if you want that for some reason.
My opinion is that Raspberry Pi has to release an NPU, and start/revolutionise the open source NPU communities and tech. Raspberry Pi's has to find itself useful in vision AI applications (extremely common in industries these days). Without it I think Arduino with its new Qualcomm boards will kill it.
The PI5 was a real downgrade for me with its lack of proper hardware h264 decoding as in the earlier versions (playing back h264 is my primary use case.) I will buy a PI6, if that comes back, else I stay with my reliable, and passively cooled, PI3.
The flagship Pi boards are also hitting the thermal design ceiling of what makes sense for this size or platform. Pi 5 is IMO already on that edge because it pretty much needs active cooling and a dedicated power supply. Going past this point means it will compete with much more powerful platforms and likely loose due to it’s architectural limitations.
They aren't actually shipping the new RP2350 silicon revision on Pico 2 boards yet. If you want the errata fixes, you've gotta source the chips and make your own boards.
For me, Pi6 shall focus on NPU to run edge models instead of getting into minipc space, e.g. 10TOPS built-in, that will be slightly better than Rockchip 3588 which was produced 4 years back at 6TOPS, and way more powerful than NXP's 2TOPS. 10TOPS is the sweet spot for edge AI as far as I can tell.
It used to be that raspberry pi was a cheap pc. Well it's not longer cheap.
And at their price point, you could just get a mini PC and have better performance, or if you want to use it as a microcontroller, you can just use an arduino, esp32, or an actual microcontroller for a fraction of the price and power consumption.
> When asked about the Pi Zero 2W, Eben said the substrate supply is constrained—basically, so many AI chips are being made that even older chips using older process nodes have to fight for the actual silicon wafers to use to make the chips.
I keep hearing voices invalidate each other, is the bottleneck the raw silicon substrate, or fab capacity?
What purity levels are required for say Pi Zero 2W?
The volume of monocrystalline silicon used in solar panels is orders of magnitude greater than the volume used in IC's / RAM production.
Is the actual bottleneck 11N + grade silicon wafers while 6N to 9N grade used in solar panels remains unaffected?
I was thinking about the RPi 6 yesterday whilst realising I couldn't set up my RPi Zero 2W anymore - the OS has become burdensome - tied strictly to an imager, that gives me an allergic reaction. Yes - they did all this for the uninitiated - but for Raspberry Pi OS Lite - bring back this experience: dd the image, write ssh into the boot drive, SSH in - change password, fully set up in almost zero fuss or effort.
Then I actually couldn't set the thing up because of the mini HDMI connection - I have a mini to HDMI cable, but to use my portable screen with it I need mini HDMI to MINI HDMI. Don't get me started on micro HDMI - almost everyone of of those connectors I've bought slips off or breaks in the device. Every time I go to set up an RPi5 I end up having to order another one of those tiny connectors.
Full HDMI for all new devices please. Even if the second display can't be connected.
These days a 175 GBP N95 from a no-name Chinese OEM on Amazon, with 16 GB of RAM and a 500GB SATA SSD is way better value and performance - and importantly - zero fuss - standard setup.
> __These days__ a 175 GBP N95 from a no-name Chinese OEM on Amazon, with 16 GB of RAM and a 500GB SATA SSD is way better value and performance - and importantly - zero fuss - standard setup.
"These days" seems to be over for now. The prices went up there, and now the only affordable option is an old laptop's motherboards with DDR3 sticks. But passively cooled PCs are rare, so the Raspberry Pi still makes sense. As a bonus, it is easier to power and find a small UPS that will keep it running for hours.
Hard to care anymore. They made it abundantly clear they dont care about their original audience of makers and education anymore now that the industrial market are so reliant on them. Not to mention the whole crapping all over their customers on social media and then accusing them of some sort of coordinated 'attack' just made it clear they're clueless about their audience.
The price increases were the end of the Pi being a viable option for most application in my eyes.
>When asked about why Picos use micro USB and not USB-C, he said it's a cost issue. USB-C connectors are more expensive than micro USB, while also taking a tiny bit more board space. That said, USB-C will probably happen someday.
Glad I'm not alone in caring about this. All I want from them is a Pico with a USB-C port. When using them for keyboards and arcade controllers you either need to wire up a separate port (and USB-C is awful to solder) or use someone else's variant with one on there. I have had issues in both cases. The Open-Frame1 leverless controller needed me to either pay for assembly or fail at soldering on the extra port myself (did the former for batch 1, got cocky when ordering batch 2 and didn't pay assembly, never got any of the new boards working). Flatbox rev 5 used an RP2040-Zero from Waveshare, which initially seemed fine but later turned out to have major bouncing issues. Typing on a virtual keyboard with it was nearly impossible, all the extra inputs being detected. The amount of debounce needed to be added in GP2040-CE settings to completely solve it resulted in it being much higher latency. I heard a theory that it was due to a lack of filtering on the MCU. Meanwhile Haute42 started pumping out incredibly cheap leverless controllers in all sorts of designs within a year of me building 10 Flatboxes. They're so cheap you can't really DIY one cheaper anymore unless you need them in bulk, and they have no bouncing/input issues. They even have extension ports to help deal with console auth. Their non-3d-printed buttoncaps also don't break as much, though I did eventually have one break after living in my backpack a while.
I'd like to take another crack at building a solid leverless someday. A new Pico with a USB-C port would probably be enough motivation.
Also I should note that part of why USB-C is such a big deal on a game controller is because they all use it now, and I've got lots of long USB-C to A cables connected to consoles and my PC, I can easily switch controllers at the user-facing end without having to re-run a cable. I can go from my Open-Frame1 playing Rivals of Aether to my Steam Controller playing Crab Champions to my Haute42 M16 playing Melty Blood all with the same cable routed under my desk, often without leaving my chair.
47 comments
[ 3.1 ms ] story [ 54.5 ms ] threadRaspberry Pi Holdings is a embedded systems manufacturer for pity's sake; we don't need more from them, we need less. [EDIT] A faster Raspberry Pi 6 is encroaching on the territory of the Intel N150 and its successors and mainstream Linux distributions and that is a battle they would lose in terms of price and performance.
Give us a Raspberry Pi Zero 3W with proper sleep states to reduce sleep power consumption, lower idle power while awake, and 1 GB of RAM even if it doubles the price.
Maybe a tick-tock release cycle (one with new features and some speed, the next with the ~same features and more speed) is where they're headed, and maybe that makes sense. They wouldn't be the first.
I'd love to see even-lower-RAM versions, though. Most of what I use Raspberry Pis for at home for is not RAM-hungry at all.
My Pi4 network router has 2GB because that was the smallest/cheapest version at release when I got it, but the system itself consistently only uses about 64MB of RAM. It'd do perfectly well and have a ton of breathing room with just 128MB of RAM (which will never happen, but if it did happen...).
I suspect the Pi4 that I use as a set-top box with Kodi would be fine with 512MB.
I've used Zero Ws for all kinds of things over the years and never felt RAM-starved with their little 512MB of RAM.
So I'm learning towards 512MB.
But sure: 1GB options would also be fine even if it does double the price. Our comments serve to demonstrate that there's room in the marketplace for different SKUs with different memory capacities. :)
How do you cope with high resolution videos?
Their niche is the industrial/embedded space. For that market, power consumption doesn't matter. What matters is that each model is guaranteed to be available till a specific date.
I have a bunch of old intel atom boards laying around. The Intel Compute Stick (TM) burnt out its flash root drive in a few months. The C2000 board I had burnt out the clock pin to drive the bios. I have a Clover Trail with a PowerVR GPU (I thought I was getting an intel GPU because it was branded Intel Graphics or similar, but nope!) that lost Windows support very quickly after launch, and has no GPU drivers for any other OS.
Instead of being fooled 4 times in a row, I looked into using an N150 for a NAS, but this time I held off a bit until after launch so I could research it first.
Lo-and-behold, they all have crazy PCIe / memory subsystem data corruption issues. I guess there are some chicken bits for the OS developers to set if the kernel can stay up long enough after boot without a panic.
Why would anyone buy this for a NAS / embedded use case?
During covid I wanted a small low power always on server. I thought about Raspi, but at the time it was expensive and I went with an intel nuc, for a similar price.
Now if I wanted to do hobby electronics, I heard I should look into esp32 or stm32..
Each Pi release is more powerful, but uses more energy. I found the Pi 4B to be the sweet spot for me, because it is the earliest model to support USB booting, gigabit Ethernet, and offer > 1GB RAM.
Perhaps a used one would fit your purposes and budget?
I currently use it to run PiHole, serve media via SMB, host Postgres & Redis, and run some custom written Dockerized apps. Home Assistant to possibly follow, too. The current load seems reasonable in htop, but I haven’t looked into burst scenarios.
> encroaching on the territory of the Intel N150.
It’s nerdier to say you built your homelab on a Raspi, and that’s what keeps the foundation afloat.
But that's not as good for PR as "bigger, faster, better" even if that comes with the problems you mentioned.
Yes, that would be kind of a dream device, perhaps also if it could suspend the os when asleep so it doesn't have to boot every time but I guess that might the standard way of doing it.
M5Stack and other esp32 based ecosystems rule the world for MCU so the Zero just need to be linux SOC with all the fancy bells and whistles with good driver support
I seriously cannot fathom being someone doing development who wouldn't pay $0.50 extra to purge the last micro USB from their desktop.
I just don't get it. Anyone who wants to save a few pennies just buys the chip directly. Their Pico board is primarily for prototyping and one off products, where quality of life is everything and 16 cents is nothing. The adapter cable probably costs more than the amount they saved. That's a dick move.
The existing secure boot mechanisms aren't bad, but allowing for more than one public key hash in OTP would be nice, too.
These kinds of things are expected to be on modern embedded SOCs and SOMs now.
[1] https://www.microcenter.com/product/673711/raspberry-pi-5
[2] https://www.microcenter.com/product/691058/nvidia-jetson-ori...
Everything they do from a compute perspective is just better with a mini pc or old laptop with a mobile spec chip.
Everything they do from a programmability perspective is just better with a microcontroller specific to the task.
I just don't see the actual market position for these things. They were supposed to be a cheap board, but you can't actually buy them cheaply because the vendors upcharge so much.
The moment you want to use a dynamically linked library, you're out of luck.
The vast MCUs don't have enough RAM or flash storage. You'll have to trim everything down so it fits.
Reasons: - full sized HDMI connector - headphone connector - good bang for the buck
If I had one wish for any new product in the Raspberry line it would be: Do the Raspberry Pi 3++ or something. Same thing. Faster, but with USB-C power connector, 4K Video resolution, 2× USB-C I/O, 2× USB-A peripherals and maybe M.2 support.
https://www.notebookcheck.net/Raspberry-Pi-discusses-Zero-3-...
The vast majority of NPUs are a poorly documented mess. The only one with good documentation is the one from AMD.
And at their price point, you could just get a mini PC and have better performance, or if you want to use it as a microcontroller, you can just use an arduino, esp32, or an actual microcontroller for a fraction of the price and power consumption.
So, what do people actually do with these pies?
I keep hearing voices invalidate each other, is the bottleneck the raw silicon substrate, or fab capacity?
What purity levels are required for say Pi Zero 2W?
The volume of monocrystalline silicon used in solar panels is orders of magnitude greater than the volume used in IC's / RAM production.
Is the actual bottleneck 11N + grade silicon wafers while 6N to 9N grade used in solar panels remains unaffected?
Then I actually couldn't set the thing up because of the mini HDMI connection - I have a mini to HDMI cable, but to use my portable screen with it I need mini HDMI to MINI HDMI. Don't get me started on micro HDMI - almost everyone of of those connectors I've bought slips off or breaks in the device. Every time I go to set up an RPi5 I end up having to order another one of those tiny connectors.
Full HDMI for all new devices please. Even if the second display can't be connected.
These days a 175 GBP N95 from a no-name Chinese OEM on Amazon, with 16 GB of RAM and a 500GB SATA SSD is way better value and performance - and importantly - zero fuss - standard setup.
"These days" seems to be over for now. The prices went up there, and now the only affordable option is an old laptop's motherboards with DDR3 sticks. But passively cooled PCs are rare, so the Raspberry Pi still makes sense. As a bonus, it is easier to power and find a small UPS that will keep it running for hours.
The price increases were the end of the Pi being a viable option for most application in my eyes.
Glad I'm not alone in caring about this. All I want from them is a Pico with a USB-C port. When using them for keyboards and arcade controllers you either need to wire up a separate port (and USB-C is awful to solder) or use someone else's variant with one on there. I have had issues in both cases. The Open-Frame1 leverless controller needed me to either pay for assembly or fail at soldering on the extra port myself (did the former for batch 1, got cocky when ordering batch 2 and didn't pay assembly, never got any of the new boards working). Flatbox rev 5 used an RP2040-Zero from Waveshare, which initially seemed fine but later turned out to have major bouncing issues. Typing on a virtual keyboard with it was nearly impossible, all the extra inputs being detected. The amount of debounce needed to be added in GP2040-CE settings to completely solve it resulted in it being much higher latency. I heard a theory that it was due to a lack of filtering on the MCU. Meanwhile Haute42 started pumping out incredibly cheap leverless controllers in all sorts of designs within a year of me building 10 Flatboxes. They're so cheap you can't really DIY one cheaper anymore unless you need them in bulk, and they have no bouncing/input issues. They even have extension ports to help deal with console auth. Their non-3d-printed buttoncaps also don't break as much, though I did eventually have one break after living in my backpack a while.
I'd like to take another crack at building a solid leverless someday. A new Pico with a USB-C port would probably be enough motivation.
Also I should note that part of why USB-C is such a big deal on a game controller is because they all use it now, and I've got lots of long USB-C to A cables connected to consoles and my PC, I can easily switch controllers at the user-facing end without having to re-run a cable. I can go from my Open-Frame1 playing Rivals of Aether to my Steam Controller playing Crab Champions to my Haute42 M16 playing Melty Blood all with the same cable routed under my desk, often without leaving my chair.