https://www.cnx-software.com/2020/09/04/picorio-linux-risc-v-sbc-is-an-open-source-alternative-to-raspberry-pi-board/

@hardwaretech
Runs with 500+ Mhz.

I don't think switch on risc-v would be a good option for amiga.
At this time due lack of any kind of resources could be smarter continua to concentrate on power/powerpc boards.
Without reinventing the weel we have already a full 'open' board like Raptor Talos II

Everything is open source, from firmware to drivers, you have only to code and port all stuff under amigaos.

But if I have to choose between Risc-v and RPi... I prefer RPi

_________________
Pegasos II G4@1GHz 2GB Radeon 9250 256MB
AmigaOS4.1 fe - MorphOS - Debian 9 Jessie

@hardwaretech

It's a promising start, but... no GPU, so it's not really an alternative to the RPi.

Last edited by bison on 05-Sep-2020 at 04:23 PM.

_________________
"Unix is supposed to fix that." -- Jay Miner

following with one with PowerVR GPU in 2021. The current one might be well paired with a pie 4 for display and emulation of older software. If some ported newer os to RISC-v (morphos or aros, even os4 would make a cheap alternative to current systems at about $250.

@hardwaretech

Is there any practical reason for all this pointless hype towards unknown, unused, untrusted exotic platforms no-one other cares about? Why amigans are so fond of obscure, forgotten architectures that try to re-emerge from the dustbin of history?

What may be the point into fragmenting once more the community? There is a break-even line between "let's port AmigaOS/MOS/AROS to another platform" and "Don't waste time and human resources into completely missable objectives".

@paolone

RISC-V is an up-and-coming thing, so it makes sense to keep an eye on it.

_________________
"Unix is supposed to fix that." -- Jay Miner

@paolone

Quote:

Is there any practical reason for all this pointless hype towards unknown, unused, untrusted exotic platforms no-one other cares about?

Europe needs to become less dependent of the US and China regarding its IT infrastructure, and a CPU architecture not controlled by US or Chinese companies is a major factor in that regard,

Not to mention that an open architecture is always a good thing as far as competition and security is concerned.

And RISC-V is everything but a "forgotten architecture", it's actually quite a success story so far. I don't know if it will ever be useful on the desktop, but there are plenty of other areas where RISC-V is having a major impact.

Porting AmigaOS to RISC-V is a ridiculous idea though.

@cgutjahr
Porting AmigaOS to RISC-V is a ridiculous idea though.

Maybe but the CPU os 4 is currently on is a dead-end street, been dead for over 10 years you have not noticed. Since you still using CPUs that date from that time. Sooner or later the software must jump to another CPU or die. Whatever they jump to must be cheap and readily available to live. You kidding yourself otherwise.

Quote:

bison wrote: @paolone RISC-V is an up-and-coming thing, so it makes sense to keep an eye on it.

Maybe I overlooked the technology and I should give it more attention, however I could not miss the details about the 'features' of the PicoRio.

1) 4-core 500 MHz processor: seriously? You have to perform at least twice the operations of a competing ARM core or 4 times the ones of a x86-64 core for every clock cycle, if you wish to catch up with the speed of a Raspberry Pi 4 (not counting that we might have Pi 5 in 2021) or a very cheap PC-based board.

2) no GPU for the first generation. Well, let's turn back to how video output was intended to be in 1995!

3) 64 bit processor with single-channel 16 bit memory interface. No need to comment.

4) 28nm fabrication process: AMD abandoned 14nm two years ago. But this is not so important, as long as frequencies are low and power consumption is not so important.

Definitely, this can be a good candidate for some Linux based embedded project, but not something worth a costy port of a whole commercial operating system (AmigaOS), nor the time needed to adapt and compile AROS on it. Our general purpose computing does not fit very well with similar hardware specs.

Quote:

paolone wrote: Maybe I overlooked the technology and I should give it more attention, however I could not miss the details about the 'features' of the PicoRio. 1) 4-core 500 MHz processor: seriously? You have to perform at least twice the operations of a competing ARM core or 4 times the ones of a x86-64 core for every clock cycle, if you wish to catch up with the speed of a Raspberry Pi 4 (not counting that we might have Pi 5 in 2021) or a very cheap PC-based board.

The PicoRio core performance is likely worse in comparison to the Raspberry Pi than you expect. It looks like the cores are tiny simple scalar cores. Perhaps they can gain some improvements in the ASIC conversion but the FPGA core benchmark only gave 0.516 DMIPS/MHz. The 68060 was 1.466 DMIPS/MHz at 600nm and the Apollo Core in FPGA likely exceeds this. A 500MHz Apollo Core in an FPGA would likely destroy the PicoRio in performance. Unless newly added, there is no FPU or SIMD unit for the PicoRio although it has a co-processor interface for adding them.

Quote:

2) no GPU for the first generation. Well, let's turn back to how video output was intended to be in 1995!

I expect there is at least a screen buffer and display controller (sometimes licensed separately from GPU). It would be difficult to market the product as a Pi competitor without at least 2D support on the SBC. The Amiga custom chips handle 2D support and display in an SoC even in FPGA form.

Quote:

3) 64 bit processor with single-channel 16 bit memory interface. No need to comment.

With the slow clocked CPU cores, banked L2, compressed code and limited graphics capabilities, maybe it wouldn't limit performance too much. Even 16 bit LPDDR4 has more memory bandwidth than Amiga classics which also didn't have an L2 yet had responsive performance for small footprint hardware. Once the graphics capabilities are improved, the lack of memory bandwidth is likely to become more of a bottleneck.

Quote:

4) 28nm fabrication process: AMD abandoned 14nm two years ago. But this is not so important, as long as frequencies are low and power consumption is not so important.

The QorIQ P1022 used in the Tabor is probably a 28nm process. This is about where PPC will be stuck until it is no longer available. Competing with the Raspberry Pi requires choosing a process which offers the most value (performance/price and transistors/price for example) instead of the best performance and lowest power. The reduction in power will likely be welcomed for embedded use but the reduction in performance, lack of hardware features and lack of software are bigger concerns. A classic Amiga SoC ASIC would have more performance and software but we have other hurdles like SMP which may restrict current software to the first core.

Quote:

Definitely, this can be a good candidate for some Linux based embedded project, but not something worth a costy port of a whole commercial operating system (AmigaOS), nor the time needed to adapt and compile AROS on it. Our general purpose computing does not fit very well with similar hardware specs.

The PicoRio has the right idea of making a low cost SoC. In my opinion, creating a product which is price competitive with the Raspberry Pi is more important than creating a product which is performance competitive. A cheap price is more conducive to creating or enlarging a user base while performance and the more important performance/price (for Pi competition) can be improved in later versions. The PicoRio is not a good target for porting yet but it gives a chance for RISC-V enthusiasts and embedded customers to cheaply enlarge the user base and allows for future improvements which is more than the Amiga has.

Last edited by matthey on 08-Sep-2020 at 10:05 PM.

@hardwaretech

Intel invests in Risc V:

https://www.google.com/amp/s/www.zdnet.com/google-amp/article/intel-invests-in-open-source-risc-v-processors-with-a-billion-dollars-in-new-chip-foundries/

@bennymee

A foil against ARM, probably. It makes sense.

_________________
"Unix is supposed to fix that." -- Jay Miner

bennymee Quote:

Intel invests in Risc V:

bison Quote:

A foil against ARM, probably. It makes sense.

Intel has invested in RISC-V developer SiFive before and maybe even tried to acquire them.

https://en.wikipedia.org/wiki/SiFive#Growth Quote:

In April 2018, SiFive received $50.6 million Series C funding including a major amount from Intel Capital. ... In June 2021, SiFive launched a new processor family with two core designs: P270, a Linux-capable CPU; and P550, the highest-performing RISC-V CPU. At the same time, Intel’s Foundry Service adopted P550 for use in its Horse Creek platform, a RISC-V development platform built on Intel’s newest 7 nm process node, Intel 4. The announcement furthered speculation of a potential acquisition of SiFive by Intel, which reportedly offered to acquire SiFive for $2 billion.

The SiFive CEO was Naveed Sherwani, an Intel Veteran, up until late 2020 and he is still Chairman of the board of directors.

https://venturebeat.com/2017/08/15/custom-processor-maker-sifive-appoints-intel-veteran-as-ceo/
https://www.sifive.com/press/sifive-appoints-patrick-little-as-president-and-chief

Fabless semiconductor tech startups are hot right now. Too bad Amiga developers develop FPGA only CPUs, use decade old PPC embedded technology and believe emulation will revive the Amiga. The Amiga used to be a technology leader and now the products trail leading edge technology by at least a decade and usually with premium prices.

Last edited by matthey on 08-Feb-2022 at 06:57 PM.

@bison

Quote:

@paolone RISC-V is an up-and-coming thing, so it makes sense to keep an eye on it.

And where did you pull that stellar piece of info from, your arse? The only place that RISC-V is going is to the same place that all its other incarnations went, which is the dust bin. Paolone is correct. RISC-V is run by a non-profit organization so it'll never even make it to the engineering samples stage let alone be manufactured for consumer devices. Adopting a RISC-V architecture makes about as much sense as putting a G-5 PCC CPU onto a modern motherboard and expecting miracles. It's nice to wax nostalgic about old architectures, but PPC and RISC are dead for very good reasons.

Quote:

matthey wrote: The QorIQ P1022 used in the Tabor is probably a 28nm process

By the QorlQ, You mean the single one cpu the Tabor Team managed to secure.

Rumors say they shot for a second piece, but by then funds had been exhausted by a trip to McDonald's for a couple Happy Meals.





/mega

_________________
I HAVE ABS OF STEEL
--
CAN YOU SEE ME? CAN YOU HEAR ME? OK FOR WORK

@ferrels

You should get in touch with Intel and tell them that you know more than they do.

_________________
"Unix is supposed to fix that." -- Jay Miner

More risc news
https://youtu.be/z89ysU4RwIg
https://youtu.be/yU2MHsRMtZE
https://youtu.be/yTMRGERZrQE
https://forums.macrumors.com/threads/raspberry-pi-vs-power-mac-g5.2111057/
A pi 3 was slightly weaker than a Mac g5 does that mean a pi 4 is equal or more powerful? If so what does it say about Amiga ppc? All I know risc cpus are getting more powerful and cheaper each year. It takes time to make a product and sell it. If they started now they cpu leave open be able to select it the last year before making the product. Software like Aros is designed to port to any hardware. I am sure the rest of Amiga OS could do the same with that much lead time.

@paolone

Quote:

Is there any practical reason for all this pointless hype towards unknown, unused, untrusted exotic platforms no-one other cares about? Why amigans are so fond of obscure, forgotten architectures that try to re-emerge from the dustbin of history?

Because we're fond of s/architectures/OS:es/g I guess

_________________

Quote:

paolone wrote: Maybe I overlooked the technology and I should give it more attention, however I could not miss the details about the 'features' of the PicoRio. 1) 4-core 500 MHz processor: seriously? You have to perform at least twice the operations of a competing ARM core or 4 times the ones of a x86-64 core for every clock cycle, if you wish to catch up with the speed of a Raspberry Pi 4 (not counting that we might have Pi 5 in 2021) or a very cheap PC-based board. 2) no GPU for the first generation. Well, let's turn back to how video output was intended to be in 1995! 3) 64 bit processor with single-channel 16 bit memory interface. No need to comment. 4) 28nm fabrication process: AMD abandoned 14nm two years ago. But this is not so important, as long as frequencies are low and power consumption is not so important.

I heard it runs AMIGA RACER really well, has a dedicate ARU just for that.

/mega

Last edited by MEGA_RJ_MICAL on 09-Feb-2022 at 01:36 PM.

_________________
I HAVE ABS OF STEEL
--
CAN YOU SEE ME? CAN YOU HEAR ME? OK FOR WORK