Also has RTC which the V5 has by virtue of an extra chip. Tomβs Hardware says it needs active cooling but hopefully the Zynthian heat exchanger will work sufficiently well. (We are already overclocking the Pi4.)
There is Raspberry Pi silicon on the board but they are still using the proprietary Broadcom SoC for core processing.
This is indeed great news and it will be fun to see whether this can be shoehorned into a V5 (or earlier).
Well!! This is super good news, and also extra work for zynthian labs, jeje!
To make life easier to case manufacturers, they have swapped the USB & ethernet connectors again, so we need to modify our case design. The good part is all USB ports will be adjacent in the V5 case.
Regarding computing power, that is what really matters, it seems we will have a nice 2x bump, so we could finally forget about some limitations we currently have, like the down-sampling in Pianoteq, or reducing the jackd latency without having XRuns. Of course, this must be tested.
The GPIO header doesnβt change, so itβs perfectly OK for us. The cooling block has to be totally redesigned. We will be working on it as soon as we get a RPI5 unit or a step (3D) model.
In TomHardwareβs post there is this statement from Gordon Hollingworth, Raspberry Pi CTO:
However, there are a significant number of HATs with software that rely on non-portable interfaces, such as RPi.GPIO, which get broken every time we release a new hardware device.
Thereβll be a new heatsink involved at the very least, and Iβm not impressed by the marketing basically saying βyah donβt expect to actually use that speed too much btw.β As a heavy CPU user, big big BIG thumbs down on this iteration of the franchise from me.
I donβt have the optics to really say, but possibly this is a good moment to consider some of the other [fruit]Pi devices for future revisions - as I recall the Orange Pi has an 8-core version that (if Iβm remembering correctly) smokes the Raspberry in terms of raw performance.
The other platforms lack the Piβs support of course, and the fact that so many hardware platforms have based themselves on it makes it a safe bet to base any new platforms on, performance be damned.
Or maybe itβs time to consider just engineering a single-board device with the brain on the same PCB. It would be a departure from the hobbyist aspect of this project, but I plan to basically shove my v5 in the face of every single keyboard player I see and do the raw math of comparing this to a similarly-capable device from Roland/Korg/Yamaha vs the cost of a v5 - I think Iβm gonna sell one or two in the next year.
Now imagine if v6 was even smaller and twice as powerful and half the price, which is 110% doable on a single-board integrated device. @jofemodo seems like the man to do the job. I would put some money into that kickstarter, if you decided to go that way. Iβd help out if I could, Iβm getting better at C and getting a handle on kernel modules is high on my todo list.
Ok early morning and Iβm letting my imagination go. When it finds something it believes in it starts running scenarios on how to make that happen morebetter. Iβm less and less enthused with the non-open Raspberry every year, quite honestly, but itβs looking like another case of market lock-in. A Zynthian that breaks away from that platform could be a 100% open source device. It would be a happy day for me. :>
This is a reasonable option. Indeed, we strongly considered this (i still have some CM4 modules and CM4 development board) when conceiving the new main board, but we decided to use the βstandardβ Pi4 because the CM4 doesnβt have USB3 ports and only an USB-2 port, what complicates design quit a bit, specially if you want USB-3. Keeping a foot in the tinkering and DIY world was in our minds too, of course.
When CM5 is released, we will reconsider this question. Very probably, USB-3 will be integrated in the CM this time.
Regarding the development of a computing circuitry for zynthian, i donβt consider this in the near-medium term. Itβs too complex and the required financial muscle is huge. And really, from a strictly zynthianic POV, i think there is little reward on it. In spain we just to say: βzapatero a tus zapatosβ
I know you donβt like non-bootstrapped projects from previous conversations, but you would probably need to use some form of funding device if you ever did consider it. Iβm considering it myself actually, I donβt have the hardware skills - yet - but if I did, I would not use any platform but this one and it would, as previously stated, be a 100% open source device, rather than the sorta-kinda-but-not-really open Pi.
Yes. No problem with it. V5 main board power source is perfectly capable.
Regarding the heat dissipation, iβm not really worried about it. The Pi5 is more efficient than the Pi4, so it will generate less heat for the same amount of load. When zynthian is really busy, the thermal block should be capable of managing 5 extra watts. When this happens, if the ambient is hot, the alucase can get really hot (although probably not so hot to cause the Pi to throtle). In such a case, an external fan should be the best solution to keep the case at a lower temperature.
Ahhh, thank you, very interesting indeed, and the correction of the Motorola number also made a search of this forum pop up: Access Virus emulator on Zynthian?
So I can see what I think is the point - the increased performance of RPi5 might allow this DSP56300 to run on Zynthian, and of course @jofemodo 's work on Bookworm-64bit might yield enough performance as well. With both - Wow!
Iβll have a look at what my Alesis Micron-Ion are running, I remember it was some kind of Motorola-NXP Coldfire processor. I would certainly like to have those in Zynthian.