AMA Raspberry Pi AMA - Ask your questions now!

[ebenupton]

Why did you decide to add 2 display outputs?

We see Raspberry Pi 4 as a PC, and it feels like the ability to drive two displays is a fundamental PC feature. Certainly if you look around our office most desks have two monitors.

Do the tarrifs on China effect raspberry pi pricing?

No effect, as we manufacture Raspberry Pi in the UK.

Why was it allowed for a more expensive Raspberry Pi this time, but not any other times?

This is the first time we have a chip that can address more than 1GB of RAM, so this is the first opportunity (or requirement) to go beyond $35. If we'd been able to address 2GB of LPDDR2, I'm sure we would have been selling a $45 Raspberry Pi 3 for a while now.

 

[ebenupton]

For machine learning, or in general?

I mean in general. It's not particularly good at "getting the compute out", certainly compared to hand-crafted GPU assembler (which sucks for other reasons).

I see Broadcom is listed as a contributor member of Khronos, but I don't see the Raspberry Pi Foundation.

The Khronos Group

Deploying and developing royalty-free open standards for 3D graphics, Virtual and Augmented Reality, Parallel Computing, Neural Networks, and Vision Processing

www.khronos.org

As Khronos plots the the future evolution of OpenCL (sometimes referred to as "OpenCL-Next"), they seem fairly receptive to input. If you have any suggestions or ideas, perhaps it might be worth reaching out to them.

I didn't say I had answers, just grumbles 😛

I'm all for that, but what I've read about OpenGL Compute Shaders gave me the distinct impression they're essentially a subset of OpenCL.

I don't mean to seem argumentative - just trying to understand your concerns about OpenCL. Thanks.

My impression is that they're the subset of OpenCL that's actually useful/easy to map to hardware.

Perhaps my complaint about OpenCL is that it's too expressive, letting you write complicated stuff that looks good but actually ends up forfeiting 90% of your theoretical compute capability in translation.

 

[ebenupton]

Aside from the extra RAM, does the new Pi4 have better virtualization support? I tried using a Pi3 as an openstack compute node and it turns out that extensive modification would have been necessary to get it to work.

Well, we have a GIC now, which should help. I can't promise everything's wired up appropriately, as it wasn't a target use case, and there's no support for virtualisation at the peripheral level, but we may be in better shape now.

What were the deficiencies that you encountered?

 

[ebenupton]

It looks like you and the team have done a great job with the Raspberry Pi 4. (Big fan of USB 3.0 and videocore VI).

Thanks! We're hoping people like it.

I was just wondering what regrets or what you would do differently if you were where you were back in 2011?

Interesting. I think, in no particular order:

• If we'd been more organised, we might have been able to launch a year earlier, though probably with less RAM.
• It would have been good to jump straight from ARM11 to Cortex-A53, though the team that implemented the A53 was stronger than the one that implemented the A7, and we wouldn't have had access to them at the time we did BCM2836.
• It would have been good to get dual-rank (so 2GB) support into the 40nm chips, though that would have entailed a lot of tear-up of the design (BCM2836 and 2837 are very incremental).

 

[ebenupton]

Good day,

Thanks for bringing the world such a wonderful product. I wish you lots of success for the future.

Thank you!

Relating to the earlier question about tarrifs, what were some of the challenges you encountered in order to keep the price at $35? I recall somebody (might have been yourself) saying GbE and USB 3 being a difficult thing to implement if you wanted to keep the price down.

I think the pressure was pretty much everywhere in the design, but a couple of places where it really showed were connectors: we have an additional HDMI connector, and as type-D (micro) connectors are more intricate than type-A (standard), the unit cost is higher. USB 3.0 connectors are more intricate and cost more than their USB 2.0 equivalents. And USB-C is more expensive than micro-B, although we use a depopulated connector as we don't have high-speed signals on there.

RAM was a bit of a push too, as we're coming off the back end of a price spike.

Do you foresee a situation were the industry's price/performance ratio stagnates somewhat and you're forced to increase the price in order to release better hardware?

It will happen eventually, but I think Moore's Law has more to give. When it happens, stagnation will hit power/performance as much as it hits price/performance, so two things to worry about.

Finally, do you have any thoughts on RISCV and if it can potentially be used on your products?

It's exciting, and I think there are probably near-term applications in the microcontroller space, but we're quite a long way from having a properly mature software ecosystem, or licensable cores in the A72 class, so I don't expect to see a RISC V Raspberry Pi in the near future.

 

[jprodrigues]

Hi Eben,

Thank you for the continued work on creating an amazing learning tool. I'm fascinated with baremetal programming and have a RTOS and I'm thrilled to port it to the new SOC

Is there a timeframe for a proper HW documentation release?

 

[ebenupton]

Hi Eben,

Thank you for the continued work on creating an amazing learning tool. I'm fascinated with baremetal programming and have a RTOS and I'm thrilled to port it to the new SOC

Excellent news 😀

Is there a timeframe for a proper HW documentation release?

We'll have an update to the venerable "BCM2835 Peripherals" document in the next week or so. Docs for the PCIe and GMAC (which you may want to play with) may take a little longer to get out, but we're working on it.

 

[chithanh]

One of the really, really great things about the previous Raspberry Pis is that you could take the bare board to basically any place in the world that has electricity, hook it up using technology which is omnipresent, IOW already there (microUSB for power, HDMI for display, SD/microSD cards for storage, USB for keyboard/mouse) and get going.

Now with dual Micro HDMI, this ability is somewhat compromised (you have to always bring a cable along, or ensure that one is present at your destination). Are you sure that this was worth it and going to stick with that decision, or are you open to make a quick board revision in case the feedback about this is negative?

We did look at stacked type-A, but it would have dominated the board z height and we were a bit squeamish about the torque on the board when inserting the upper connector, and the EMC implications of routing high-speed signals to it.

Did you also look at upright HDMI Type-A connectors?

 

[will9183]

Hi Eben, thanks for your response, I just have another question. Will we see an official port of Android/Android TV to Raspberry Pi in the near future?

 

[ebenupton]

One of the really, really great things about the previous Raspberry Pis is that you could take the bare board to basically any place in the world that has electricity, hook it up using technology which is omnipresent, IOW already there (microUSB for power, HDMI for display, SD/microSD cards for storage, USB for keyboard/mouse) and get going.

Now with dual Micro HDMI, this ability is somewhat compromised (you have to always bring a cable along, or ensure that one is present at your destination). Are you sure that this was worth it and going to stick with that decision, or are you open to make a quick board revision in case the feedback about this is negative?

I definitely think this was the right decision; in any case quick board revisions don't really exist when you're shipping ~500ku a month and need to ensure continuity of supply. I don't think it's a significant burden to carry a $3 micro HDMI male to HDMI female adapter if you want to be able to integrate with arbitrary bits of kit at your destination.

Did you also look at upright HDMI Type-A connectors?

I don't believe we did.

 

[ebenupton]

Hi Eben, thanks for your response, I just have another question. Will we see an official port of Android/Android TV to Raspberry Pi in the near future?

Not from us, but I believe the move to the V3D Mesa driver, KMS, and ARM-side control of the HEVC decoder, makes a high-performance community port much more viable.

 

[chithanh]

Thank you for your reply.

in any case quick board revisions don't really exist when you're shipping ~500ku a month and need to ensure continuity of supply.

Of course by "quick" I don't mean shipping tomorrow or even next month. I am aware that such changes have a many-month lead time from the time the decision is made to actual changed products in users' hands.

I don't think it's a significant burden to carry a $3 micro HDMI male to HDMI female adapter if you want to be able to integrate with arbitrary bits of kit at your destination.

It certainly is not if you live in a developed country and can just walk into the next electronics mega-store and buy one.
It is if you are in a rural place in some less developed country and the adapter cable just gave in to flexion-based damage, or the local fauna has decided to gnaw on it.

I don't believe we did.

If at any time you get second thoughts about the Micro HDMI thing, please consider upright connectors. They would not solve the problem of Z height, but somewhat reduce the stress on the PCB compared to the stacked HDMI.

 

[ebenupton]

It certainly is not if you live in a developed country and can just walk into the next electronics mega-store and buy one.
It is if you are in a rural place in some less developed country and the adapter cable just gave in to flexion-based damage, or the local fauna has decided to gnaw on it.

I hear you on that last point. I have enough trouble in the UK with my local fauna (species homo sapiens, age 2): cost me $50 in lightning cables before I trained her they weren't chewable treats.

If at any time you get second thoughts about the Micro HDMI thing, please consider upright connectors. They would not solve the problem of Z height, but somewhat reduce the stress on the PCB compared to the stacked HDMI.

You're right - they do look like they should have lower bending moment than the top socket of a stack.

 

[ElDani]

Hi Eben, a question somewhat in reply to the recent few posts. Did you consider making some more radical changes to the board design of the Pi 4 that you always wanted to since designing the very first Pi, or at least the B+? If yes, why didn't you go farther than you have?

I'm asking because some of the changes present (ethernet/USB, HDMI/double HDMI Micro) already make it incompatible with previous generation cases for example.

 

[james.Harking]

Good morning again Eben,

Thank you for your previous answers to my questions, particulary the one about integrating a Raspberry Pi in to a monitor as a single device, I really do think there is a lot of value there. Particulary with the move to a more destop focused paradigm. A Raspberry PI like iMac would be very nice indeed and save us a lot of clutter and lost parts! After all, 21.5inch 1080p panels are cheap now

But on to my additional questions for you. The Raspberry Pi has clearly been a success by almost any metric with millions of units being sold. But do you feel that it has fulfilled the original goals as originally envisioned?

“further the advancement of education of adults and children, particularly in the field of computers, computer science and related subjects”.

With a large proportion of units going in to industrial computing or being used as 'Kodi boxes'/ media centres and Emulation machines (not to speak of the dubious legal nature of these shipping with thousands of copyrighted encumbered ROMs). Indeed, with success unfortunately comes people that will use tools quite nefariously.

For many the Pi has become a 'black box' for that just serves a single purpose, do you have any regrets/concerns about this? I suppose this still brings money in to the foundation, so this is not all bad.

In my experience the Code Club and CoderDojo 'mergers' have been successful for the foundation and should be commended, can you speak to the impact and success that has been made to the original mission statement and the number of children becoming involved with programming? Is there any outreach to governmental departments to inform on the national curriculum?

Over the years that the Raspberry Pi has been in the market place has there been a shift in the need for single board computer? Is this still as relevant as invisioned in 2008? If not, as relevant what do you think should be the market to move to? Do you for example have any evidence that children are using their tablet or mobile phones as comuputing devices for learning purposes?

Would a software release incorporating programming and other tols from the foundation (similar to the Raspbian x86 release) for these types of platforms, primarily iOS and Android be more beneficial in the long run as you would potentially increase the market to billions of end users?

Looking over the next 5 years do you expect that the devices released by the foundation will change significantly or be incremental upgrades over the current devices we have today? What do you see as being the largest challenge for the foundation? How closely do you follow 'competitors' in the SBC market and what have you learnt from their releases? Certainly, from my perspective their software support has been considerably lacking compared to the Raspberry Pi even if the hardware is on paper better and cheaper.

Regarding manufacturing, do you see it as a competitive advantage that the Pi is assembled in the UK? Do you have any stats on how much cheaper the Pi could be manufactured if it was made in Shenzan for example? If prices are lower could a Pi be offererd that would reflect that difference in cost?

With the movement from 40 to 28 nanometer production what advantages have you been able to capitalise on? As 28 is now quite old and mature are you happy to stay on this node for the forseeable future? Would costs increase considerably if moving to a lower power node such as TSMCs 14nm? Do you even have any control there if you continue to use Broadcom SOCs? Do the foundation have any relationship with Arm holdings directly? I'm sure you must have one eye eventually on future designs using newer platforms such as Cortex A76/A77.

I know that is a lot of questions but really, I am a huge fan of the work that you do and just want to say thank you.

Kind regards,

Jim

 

[MisfitFTW]

Hello Eben,

Although it may sound weird considering that the Raspberry Pi 4 is recently released, I would like to ask this anyway.

What features would you like to have on the Raspberry Pi 5? Also are there any features in your mind that make you say "it would be great if we had it but alas, we cannot." kind of thing?

 

[ebenupton]

Hi Eben, a question somewhat in reply to the recent few posts. Did you consider making some more radical changes to the board design of the Pi 4 that you always wanted to since designing the very first Pi, or at least the B+? If yes, why didn't you go farther than you have?

I'm asking because some of the changes present (ethernet/USB, HDMI/double HDMI Micro) already make it incompatible with previous generation cases for example.

I think we're happy with the basic form factor: big z-height ports on the right, GPIO at the top and everything else along the bottom.

One amusing observation that has come up recently (while shooting the videos for the launch) is that the unit often ends up on the desk "upside down", so that the HDMI and power go towards the back of the desk. It did occur to us that we really should flip the board, or just the silk screen, to account for this.

 

[winkleink]

Hi Eben,

With Raspberry Pi growing it seems like the focus is moving away from the Community and more towards Industry.
I expect this to be the case for Trading as that is where the majority of sales are coming from.

Do you know for the Foundation what the Community support plans are?
NCCE/|Picademy supporting teachers and Code Club and Coder Dojo are well resourced (relatively speaking) for their volunteer lead activities.
So, I suppose my question is more specifically related to Community events like Raspberry Jams?

Albert.

 

[will9183]

Thank you Eben for answering my previous questions, I just got one more question, Target in the US has been selling the Raspberry Pi 3 for a while now, will more Raspberry Pi stuff (like different models and accessories) be coming to Target and other stores like it in the US?

 

[ebenupton]

But on to my additional questions for you. The Raspberry Pi has clearly been a success by almost any metric with millions of units being sold. But do you feel that it has fulfilled the original goals as originally envisioned?

“further the advancement of education of adults and children, particularly in the field of computers, computer science and related subjects”.

With a large proportion of units going in to industrial computing or being used as 'Kodi boxes'/ media centres and Emulation machines (not to speak of the dubious legal nature of these shipping with thousands of copyrighted encumbered ROMs). Indeed, with success unfortunately comes people that will use tools quite nefariously.

For many the Pi has become a 'black box' for that just serves a single purpose, do you have any regrets/concerns about this? I suppose this still brings money in to the foundation, so this is not all bad.

Overall sales have so exceeded our original estimates, that we're having more impact than we'd hoped for even with "only" 10-20% of sales going into education. With respect to our original parochial goal, we see ~1100 applicants per year to CS at Cambridge, up from a low of ~200 in 2008, and substantially above our previous dot-com era peak of 600 in 1999.

In my experience the Code Club and CoderDojo 'mergers' have been successful for the foundation and should be commended, can you speak to the impact and success that has been made to the original mission statement and the number of children becoming involved with programming?

These mergers have both been great successes for the Foundation. We're looking at something like 15,000 clubs globally, reaching 250,000 students.

Is there any outreach to governmental departments to inform on the national curriculum?

Raspberry Pi is part of a consortium delivering the National Centre for Computing Education, a £84m UK government program to train teachers to support the new Computer Science curriculum in England.

Over the years that the Raspberry Pi has been in the market place has there been a shift in the need for single board computer? Is this still as relevant as invisioned in 2008? If not, as relevant what do you think should be the market to move to? Do you for example have any evidence that children are using their tablet or mobile phones as comuputing devices for learning purposes?

If anything I think we're seeing an increased level of interest in the sorts of physical computing activities at which Raspberry Pi excels. There's definitely some interesting work going on with teaching on tablets and mobiles, but ultimately they're very limited platforms compared to a real PC.

Would a software release incorporating programming and other tols from the foundation (similar to the Raspbian x86 release) for these types of platforms, primarily iOS and Android be more beneficial in the long run as you would potentially increase the market to billions of end users?

It's an interesting idea, but probably not something we are best placed to deliver.

Looking over the next 5 years do you expect that the devices released by the foundation will change significantly or be incremental upgrades over the current devices we have today? What do you see as being the largest challenge for the foundation? How closely do you follow 'competitors' in the SBC market and what have you learnt from their releases? Certainly, from my perspective their software support has been considerably lacking compared to the Raspberry Pi even if the hardware is on paper better and cheaper.

I expect we'll continue to invest in software support for all Raspberry Pi platforms, back as far as Raspberry Pi 1B from 2012. There's a lot of work to be done to enable some of the more advanced features of the new platform. I guess the challenge will be to allocate resources to this work while also (in a year or so) starting to think about what a future hardware platform might look like.

Regarding manufacturing, do you see it as a competitive advantage that the Pi is assembled in the UK? Do you have any stats on how much cheaper the Pi could be manufactured if it was made in Shenzan for example? If prices are lower could a Pi be offererd that would reflect that difference in cost?

We save roughly $1-2 by manufacturing Raspberry Pi in the UK.

With the movement from 40 to 28 nanometer production what advantages have you been able to capitalise on? As 28 is now quite old and mature are you happy to stay on this node for the forseeable future?

28nm brings quite substantial improvements in energy. It's also the current "value node" (lowest cost per transistor), and I think will remain so for the foreseeable future.

Would costs increase considerably if moving to a lower power node such as TSMCs 14nm? Do you even have any control there if you continue to use Broadcom SOCs? Do the foundation have any relationship with Arm holdings directly? I'm sure you must have one eye eventually on future designs using newer platforms such as Cortex A76/A77.

Yes, we would currently expect unit costs to increase if we moved to a smaller (16nm, 12nm, 7nm) process. This won't be the case forever, but I could see us staying on 28nm for another five years at this point.

I know that is a lot of questions but really, I am a huge fan of the work that you do and just want to say thank you.

You're most welcome. Thanks for your support.

 

[ebenupton]

Hello Eben,

Although it may sound weird considering that the Raspberry Pi 4 is recently released, I would like to ask this anyway.

What features would you like to have on the Raspberry Pi 5? Also are there any features in your mind that make you say "it would be great if we had it but alas, we cannot." kind of thing?

The interesting thing is that really our feature set hasn't changed since Raspberry Pi 3B in 2016: newer models just have "more" of everything, even if all of that "more" sometimes stacks up to a more qualitative change (e.g. from "not a PC" to "a PC").

I suppose Raspberry Pi 5, whenever it arrives, is likely to just have more of everything. One hope is that over the next few years we see enough decline in RAM prices to allow us to fit more memory into the baseline $35 device.

 

[ebenupton]

Hi Eben,

With Raspberry Pi growing it seems like the focus is moving away from the Community and more towards Industry.
I expect this to be the case for Trading as that is where the majority of sales are coming from.

Do you know for the Foundation what the Community support plans are?
NCCE/|Picademy supporting teachers and Code Club and Coder Dojo are well resourced (relatively speaking) for their volunteer lead activities.
So, I suppose my question is more specifically related to Community events like Raspberry Jams?

Albert.

Hey Albert
You're right that Trading hasn't historically been the home of community-support activities at Raspberry Pi.

There's currently an active debate going on about how we can best support the Jams movement: we need to acknowledge that when you have these large networks of events (Clubs, Dojos, Jams), there's going to be a certain amount of churn at the individual club level, but there may be things we can do to help support volunteers to prevent attrition. An example would be helping create governance structures that avoid events becoming too dependent on the time and commitment of their original founders: lots to learn from the Scouts in this regard.

In any case, it's on our radar, and we're going to try to come to a decision on what support we should be providing, and from what organisation, very soon.

 

[ebenupton]

Thank you Eben for answering my previous questions, I just got one more question, Target in the US has been selling the Raspberry Pi 3 for a while now, will more Raspberry Pi stuff (like different models and accessories) be coming to Target and other stores like it in the US?

I don't have any info here I'm afraid, as this isn't a direct engagement, but rather is run by (I think) Canakit, one of our Approved Resellers. Having said that, getting product into bricks-and-mortar retail is a priority for us, and the work Target have been doing in the maker space is very impressive, so I'm hopeful we'll see growth here.

 

[winkleink]

What kinds of Industrial applications are you seeing the Raspberry Pi hit the sweet spot for?


With the regular Raspberry Pi certified and dual display ports on the Pi4B is the Compute Module still in demand for new industrial users or is the expectation that the Pi4B will become their primary platform.

 

[winkleink]

There's currently an active debate going on about how we can best support the Jams movement:

If you can pass the message it would be great if the Jam community to be involved in these discussions.