RISC-V microcontrollers from GigaDevice

[thm_w]

Does it have a similar set of peripherals as well?

Yeah, I guess if a company is producing millions a year of devices using a STM32F103, they could justify the time investment in porting the software to this new chip, and then it would just be a matter of changing a single item in the BOM...

Not very fair to STM, but hey. Who said business was fair, right?

Yeah the peripherals are almost identical, at least for what I looked at, ADC for example (12-bit same sampling times, etc).

There is an opportunity to improve the peripherals (they did bump up the clock speed), but I think that makes things overly complicated, and the vast majority of people buying the chip don't care. Maybe they will come out with another model in the future though.

[Sal Ammoniac]

I've never heard of such a capability!  Sounds neat.
Who is "your credit card company"?

Citi.

Another use for them is when buying something that has an auto-renewal "feature". I've been burned by these in the past, as they're sometimes hidden in the small print, so now I just set the expiration date for the virtual card number to next month, and any auto renewal will fail.

[PCB.Wiz]

Does it have a similar set of peripherals as well?

Yeah, I guess if a company is producing millions a year of devices using a STM32F103, they could justify the time investment in porting the software to this new chip, and then it would just be a matter of changing a single item in the BOM...

That illustrates a problem for any traction.
If the MHz is the same, and the peripherals are the same, then it is only price that might select RISC-V, and I doubt any large company will move to/risk a single-source part, for a few cents saving.

To get traction in RISC-V, you really need to offer something extra, as a motivator.

Is Gigadevice 'Chinese enough' to win favours in China, over ARM ? - I notice Softbank sold ARM china recently, and certainly China would be keen to avoid sending any license fees off-shore  ?

[ataradov]

It eventually will offer lower price due to better licensing. And after that it might offer custom instructions.

This is literally the first real MCU with this core. This is good enough for now.

I would really like to get my hands on both dev kits. But I'm with others - I want an ordering firm in a language I can read.

[oPossum]

If the MHz is the same, and the peripherals are the same, then it is only price that might select RISC-V, and I doubt any large company will move to/risk a single-source part, for a few cents saving.

Also...

• Higher IPC
• Better code density
• Lower mW/MHz

[ebclr]

While those afraid of everything are making complains, I'm playing with my developer board bought from China in Chinese ( I'm smart enough to use google translator and understand ), paid with Paypal with my one-time card.

I hope they understand how limited they are

I guess no one of those was never been in Shenzhen to understand what is going on, and how obsolete they are

[brucehoult]

I checked the LQFP48 pinout and it matches up, so theoretically you can drop this thing into an existing board design for STM32F103, once you figure out the code of course.

This is really smart, first the F103 clone for a lower price to build trust/brand, then this. AFAIK there haven't been any major issues with that design, but I don't know if anyone has done in depth characterizations.

Does it have a similar set of peripherals as well?

Yeah, I guess if a company is producing millions a year of devices using a STM32F103, they could justify the time investment in porting the software to this new chip, and then it would just be a matter of changing a single item in the BOM...

Not very fair to STM, but hey. Who said business was fair, right?

Certainly would be cool if the peripherals are compatible and anyone who wrote their stuff in C pretty much only has to do a recompile.

[brucehoult]

Certainly would be cool if the peripherals are compatible and anyone who wrote their stuff in C pretty much only has to do a recompile.

Good luck on RTOS. That's a lot of works to do, especially if you make your own and don't want to go through hundreds pages of Chinese document on that particular RV implementation.

My understanding is this GD new MCU was not planned ahead of time. It was just an attempt trying to capitalize the fear of Chinese market on dependency to proprietary US technology.

I believe Zephyr and FreeRTOS are both in pretty good shape for RISC-V. Of course you need a little porting work for each chip/board.

[SiliconWizard]

Does it have a similar set of peripherals as well?

Yeah, I guess if a company is producing millions a year of devices using a STM32F103, they could justify the time investment in porting the software to this new chip, and then it would just be a matter of changing a single item in the BOM...

That illustrates a problem for any traction.
If the MHz is the same, and the peripherals are the same, then it is only price that might select RISC-V, and I doubt any large company will move to/risk a single-source part, for a few cents saving.

I don't quite agree with that. Using STM is also having a single-source solution.
Whereas the "risk" may not be worth it for a typical western company, I think it can be well worth it for a chinese one.
Don't forget that those chinese chips are MAINLY designed to address the chinese industry and help it become fully independent, not to please the western world.

To get traction in RISC-V, you really need to offer something extra, as a motivator.

For a typical western company, again, I'd agree with you. Not necessarily for a chinese one.

Is Gigadevice 'Chinese enough' to win favours in China, over ARM ?

I don't know about Gigadevice as a whole, but any chip designed in China and not depending on any non-chinese IP is probably enough to win favours in China at the moment.

[hansd]

Here is an other one https://www.sifive.com/boards HiFive1 Rev B. Only 16KB of ram. And the Vega board https://open-isa.org/ with 4 cores

[PCB.Wiz]

... And the Vega board https://open-isa.org/ with 4 cores

That's unusual, it has a NXP logo, and claims to have a foot in both camps, with ARM and RISC V cores. RF support and BGA with large memories (1.25MB Flash  384k RAM ) 
However, even with a Nov 2018 NXP data sheet download, a search on NXP.COM for RV32M1, finds nothing at all ?

[hansd]

It must have been initiated by Freescale as and experiment? https://mcuoneclipse.com/2019/03/16/running-freertos-on-the-vega-risc-v-board/ https://mcuoneclipse.com/2019/03/05/debugging-the-rv32m1-vega-risc-v-with-eclipse-and-mcuxpresso-ide/

[coppice]

... And the Vega board https://open-isa.org/ with 4 cores

That's unusual, it has a NXP logo, and claims to have a foot in both camps, with ARM and RISC V cores. RF support and BGA with large memories (1.25MB Flash  384k RAM ) 
However, even with a Nov 2018 NXP data sheet download, a search on NXP.COM for RV32M1, finds nothing at all ?

That device makes huge sense. The core is only a few percent of the die on a chip like that. Adding a second core lets them dip a toe in the RISC-V waters cheaply, while having a ARM core that ensures market acceptance. Making a complex chip with only a core whose market acceptance is unproven is risky. Making a simple chip with the unproven core means it won't do enough to attract much of an audience.

[PCB.Wiz]

That device makes huge sense. The core is only a few percent of the die on a chip like that. Adding a second core lets them dip a toe in the RISC-V waters cheaply, while having a ARM core that ensures market acceptance. Making a complex chip with only a core whose market acceptance is unproven is risky. Making a simple chip with the unproven core means it won't do enough to attract much of an audience.

Yes, I agree it is a clever approach. Given the large Flash and RAM, and Peripherals, the cores will be relatively small % of area.

[langwadt]

Certainly would be cool if the peripherals are compatible and anyone who wrote their stuff in C pretty much only has to do a recompile.

Good luck on RTOS. That's a lot of works to do, especially if you make your own and don't want to go through hundreds pages of Chinese document on that particular RV implementation.

My understanding is this GD new MCU was not planned ahead of time. It was just an attempt trying to capitalize the fear of Chinese market on dependency to proprietary US technology.

ARM is from the UK

[langwadt]

ARM is from the UK

ARM is owned by Softbank, which is known to be very keen on following US rules. Apple is another owner of ARM, which naturally, follows US rules.

Per US DOC regulation, all companies with more than 25% US capital or US technology fall within DOC's jurisdiction. If such a company defies US regulation, it will face sanction, even if the major stake is owned by foreign entities.

Softbank is Japanese and ARM is headquartered in the UK, How much stock does Apple own in ARM?

[Sal Ammoniac]

How much stock does Apple own in ARM?

Probably not a lot these days. Fun fact: ARM was originally started as a joint venture between Acorn Computers, Apple, and VLSI Technology.

[chickenHeadKnob]

... And the Vega board https://open-isa.org/ with 4 cores

That's unusual, it has a NXP logo, and claims to have a foot in both camps, with ARM and RISC V cores. RF support and BGA with large memories (1.25MB Flash  384k RAM ) 
However, even with a Nov 2018 NXP data sheet download, a search on NXP.COM for RV32M1, finds nothing at all ?

That device makes huge sense. The core is only a few percent of the die on a chip like that. Adding a second core lets them dip a toe in the RISC-V waters cheaply, while having a ARM core that ensures market acceptance. Making a complex chip with only a core whose market acceptance is unproven is risky. Making a simple chip with the unproven core means it won't do enough to attract much of an audience.

These multi-core, multi instruction set franken-monstrosities don't appeal to me at all. TI has something  like this in the AM574x sitara.
Which has, count em,
2 Arm Cortex A15, 2 C66x DSP, 2 cortex M4, 2 PRU's and a partridge in a pear tree!  Just Crazy.
Each  different ISA requiring a different compiler and tool chain. who needs that complexity headache.

I see at least 4 market segments where RISC-V has a unique value proposition.
1. Very low cost high volume commodity controllers where the ARM licence cost is a dead weight and every penny counts.
2. companies like Western Digital who have high volume internal needs and want to be free from ARM licencing control and cost.
3. commodity manufactures who want to be immune to capricious changes in export law/sanctions
4.High security PC applications and server market, those that  want to be free from back-doors like  Intel's management engine

[brucehoult]

... And the Vega board https://open-isa.org/ with 4 cores

That's unusual, it has a NXP logo, and claims to have a foot in both camps, with ARM and RISC V cores. RF support and BGA with large memories (1.25MB Flash  384k RAM ) 
However, even with a Nov 2018 NXP data sheet download, a search on NXP.COM for RV32M1, finds nothing at all ?

That device makes huge sense. The core is only a few percent of the die on a chip like that. Adding a second core lets them dip a toe in the RISC-V waters cheaply, while having a ARM core that ensures market acceptance. Making a complex chip with only a core whose market acceptance is unproven is risky. Making a simple chip with the unproven core means it won't do enough to attract much of an audience.

These multi-core, multi instruction set franken-monstrosities don't appeal to me at all. TI has something  like this in the AM574x sitara.
Which has, count em,
2 Arm Cortex A15, 2 C66x DSP, 2 cortex M4, 2 PRU's and a partridge in a pear tree!  Just Crazy.
Each  different ISA requiring a different compiler and tool chain. who needs that complexity headache.

I see at least 4 market segments where RISC-V has a unique value proposition.
1. Very low cost high volume commodity controllers where the ARM licence cost is a dead weight and every penny counts.
2. companies like Western Digital who have high volume internal needs and want to be free from ARM licencing control and cost.
3. commodity manufactures who want to be immune to capricious changes in export law/sanctions
4.High security PC applications and server market, those that  want to be free from back-doors like  Intel's management engine

I'd say you missed a few.

1. anything where a custom instruction tightly integrated into the processor (i.e. take one or two register values, compute something, write the result to a register, all in 1 clock cycle) can significantly improve performance and energy efficiency on some task. ARM won't allow you to do that at any price. A coprocessor or memory-mapped device are the best they'll allow. This is ARM policy rather than physics or mathematics, so they could choose to change this.

2. anything that wants a microcontroller level of area / performance / energy consumption but wants 64 bit addressing, for example because you're an I/O controller in the corner of a bigger chip. ARM doesn't offer small 64 bit cores comparable to Cortex M0 or even M3. Again, this is more policy than anything, though they'd also want to define a subset of Aarch64, which is not currently allowed. RV64I cores can be very small -- and there seems to be demand for RV64E (16 registers only) also.

3. anything that wants 64 bit with code density comparable to Thumb2. Aarch64 code is much bigger than Thumb2 (though smaller than original ARM). This is hard to fix. There is insufficient room in the A64 encoding to add on 16 bit opcodes so you'd need either an awkward mode change as with original Thumb, or else an entirely new instruction encoding. MIPS, incidentally, seem to have done a good job of this with NanoMIPS which has 2, 4, and 6 byte instructions and both 32 bit and 64 bit. It looks to be virtually stillborn though, with only a single 32 bit chip announced 15 months ago, and no inclusion in their Open Source offering.

[coppice]

... And the Vega board https://open-isa.org/ with 4 cores

That's unusual, it has a NXP logo, and claims to have a foot in both camps, with ARM and RISC V cores. RF support and BGA with large memories (1.25MB Flash  384k RAM ) 
However, even with a Nov 2018 NXP data sheet download, a search on NXP.COM for RV32M1, finds nothing at all ?

That device makes huge sense. The core is only a few percent of the die on a chip like that. Adding a second core lets them dip a toe in the RISC-V waters cheaply, while having a ARM core that ensures market acceptance. Making a complex chip with only a core whose market acceptance is unproven is risky. Making a simple chip with the unproven core means it won't do enough to attract much of an audience.

These multi-core, multi instruction set franken-monstrosities don't appeal to me at all. TI has something  like this in the AM574x sitara.
Which has, count em,
2 Arm Cortex A15, 2 C66x DSP, 2 cortex M4, 2 PRU's and a partridge in a pear tree!  Just Crazy.
Each  different ISA requiring a different compiler and tool chain. who needs that complexity headache.

I see at least 4 market segments where RISC-V has a unique value proposition.
1. Very low cost high volume commodity controllers where the ARM licence cost is a dead weight and every penny counts.
2. companies like Western Digital who have high volume internal needs and want to be free from ARM licencing control and cost.
3. commodity manufactures who want to be immune to capricious changes in export law/sanctions
4.High security PC applications and server market, those that  want to be free from back-doors like  Intel's management engine

I agree with you about some of the weird mixed core chips on the market, requiring a variety of software tool chains and other complexities. However, the NXP chip is designed as an either/or device. You can choose which core to use, and completely ignore the other one. Its completely benign.

[SiliconWizard]

ARM has design centers in the US and other countries that are part of NATO, so many ARM IPs, partly or as a whole, have been designed in those. That alone can enable the US to set the rules when it comes to dealing with "unwanted" countries.

[brucehoult]

Sipeed are releasing a $4.90 board called the Longan Nano using this GD32VF103CBT6 chip. Board and chip specs in English and SDK are all here:

http://dl.sipeed.com/LONGAN/Nano/

Orders here: https://www.seeedstudio.com/Sipeed-Longan-Nano-RISC-V-GD32VF103CBT6-Development-Board-p-4205.html

Good reliable company.  Many here will already be familiar with their boards using the Kendryte K210 RISC-V chip.

They say the chip has 32 KB SRAM (double the largest mentioned in the first post in this thread), and 128 KB Flash.

The standard SDK is Eclipse-based, but they are also supporting PlatformIO, which is now sponsored by SiFive and Western Digital making the full version with debugging free for everyone (not only RISC-V users).

[nctnico]

... And the Vega board https://open-isa.org/ with 4 cores

That's unusual, it has a NXP logo, and claims to have a foot in both camps, with ARM and RISC V cores. RF support and BGA with large memories (1.25MB Flash  384k RAM ) 
However, even with a Nov 2018 NXP data sheet download, a search on NXP.COM for RV32M1, finds nothing at all ?

The layout of the documents looks exactly like the documents from Freescale. This is very interesting because it may hint NXP is working on a new breed of non-ARM microcontrollers and maybe even SOCs.

[brucehoult]

A story on the Longan Nano here, although I don't think there's any new info: https://www.cnx-software.com/2019/08/29/longan-nano-gd32v-risc-v-development-board-comes-with-lcd-display-and-enclosure/

And apparently the "Bumblebee" core is actually Andes N22. There's a story here: https://www.cnx-software.com/2019/08/28/andescore-n22-risc-v-core-rv32imac-rv32emac-instruction-sets/

This story doesn't seem to quite match the Gigadevice/Sipeed docs, for example the Sipeed docs say that PMP (Physical Memory Protection .. basically a set of several base&bound registers so you can detect things such as stack overflow or protect machine mode things from user mode code) is not implemented.

[mac.6]

Here is an other one https://www.sifive.com/boards HiFive1 Rev B. Only 16KB of ram. And the Vega board https://open-isa.org/ with 4 cores

RV32M1 is not a true 4 core, it's 2+2, ie you use either ARM CM4/CM0+ or RI5CY/ZeroRISCY pair (and I think ARM/RISC-V hybrid also).