Categories

STM32F103 vs GD32F103

Posted on Friday, January 19th, 2018 in ARM, Chips by DP

stm32vsgd32

Sjaak wrote about a Chinese ARM chip compared to a ST ARM chip:

Most of us do know the ST line of ARM chips called STM32. They come in multiple flavours and the STM32F103 is one of the most common entry level family of chips. They are called by ST as mainstream. They are a full featured 32 bit ARM Cortex M3 chip running at max. 72MHz with all the requisite peripherals like ADC, DAC, USB, CAN, I2C, I2S, SPI, SDIO, PWM, RTC, interrupts and various timers. Lets zoom into the STM32F103C8 chip (which seems the be the go-to choice of the Chinese el-cheapo development breakout boards)

See the full post at smdprutser.nl.

This entry was posted on Friday, January 19th, 2018 at 8:47 pm and is filed under ARM, Chips. You can follow any responses to this entry through the RSS 2.0 feed. You can skip to the end and leave a response. Pinging is currently not allowed.

11 Responses to “STM32F103 vs GD32F103”

  1. Ja says:

    Where these GD chips can be found? Everywhere I look I have found only STM chips.

  2. KH says:

    That’s a good one, ha ha. Let’s see whether ST Microelectronics takes action. I can understand how this can be done legally, ARM is not STM-proprietary, and can one copyright peripheral functionality? Oh, this is going to be interesting.

    How long before someone makes an AVR clone? Yumm. MCHP will blow a gasket. Please, please, make me a quad-core AVR!

    • Bandi says:

      Not sure if you’re trolling or just spent the last few years under a rock…
      GD32 chips have been around since at least 2015.
      AVR clones (like the LGT8F88A) date back to at least to 2013, .

  3. Drone says:

    It seems this GD32F103 part is manufactured by GigaDevice Semiconductor (Beijing):

    http://www.gigadevice.com/product-category/11.html

    The core is an ARM Cortex-M3. ARM Holdings is originally from (and still based in) the U.K., but is now (as of July 2016) majority owned (75%) by SoftBank of Japan. ARM primarily licenses their own Intellectual Property (IP) semiconductor designs which any legitimate license holder (paying royalties to ARM of-cource) may use to build physical devices using a capable fabricator. This is what ST does, and it seems this is what GigaDevice is doing so long as long as GigaDevice is a legitimate holder of an ARM license for this core.

    https://en.wikipedia.org/wiki/ARM_Holdings

    https://en.wikipedia.org/wiki/SoftBank_Group

    GigaDevice should be OK at a high-level if they are a legitimate ARM IP licensee for the core used in this part. But if GigaDevice is using stolen ARM IP (not likely) then it could be in big trouble.

    On the other hand, if GigaDevice is a legitimate licensee of ARM IP for this part, but it is mimicking an ST part close enough to trigger Copyright and/or Trade-Mark challenges, then GigaDevice may also be in trouble. Just the GD32F103 part number may be a red flag because in this case it is obviously targeting the ST product line.

    There may also be licensing restrictions imposed by ARM as to how a licensee may use their IP. Perhaps a restriction would be that a licensee must not closely mimic another licensee’s competitive part based on ARM IP. We have no visibility into this and one could argue that ARM actually wants as many manufacturers as possible to mimic each other.

    Keep one (big) thing in-mind here: There is still little to no rule of law in China today regarding Intellectual Property protection. So even if GigaDevices is doing something “wrong”, good luck holding them accountable!

    Assuming everything is legitimate across the board with this GigiDevice ARM based part, GigaDevice might have a distinct advantage in the Chinese market selling their parts there because (obviously) they have native command of the Chinese language. ST Microelectronics could invest large amounts of money to properly document and support their parts in the Chinese language/market, but obviously this puts them at a cost disadvantage compared with native firms like GigaDevice.

    Mini-Rant: These days it seems to me that regardless of what comes out of China in terms of semiconductor parts, there’s a government driven directive to limit or hamper data on Chinese parts which may lead to companies outside China using these for manufacture outside of China. Or maybe that’s not the case and it is simply too difficult for many/most Chinese firms to produce reasonable quality English documentation, support, and development tools. I dunno…

    • Sjaak says:

      I dunno if you can ‘license’ register definitions, but I’m not a lawyer. Since the chip is available since 2015 and I haven’t heard of any lawsuits I think it is OK. AFAIK they have paid their ARM license fees.

      BTW there is some english documentation on the GD32 mini site: http://gd32mcu.21ic.com/en/index

  4. Bob says:

    That’s what I was wondering, too. Any ARM-based microcontroller is going to have the ARM core itself plus a whole slew of peripherals (DMA engines, SPI/I2C, etc.). It’s remotely possible all the peripherals on the real STM32F103 are licensed from ARM and hence could be (in theory) also licensed by the GD design. But my guess is that at least several of the STM peripherals were designed in-house by STMicro and hence should not be available to the GD design. Something seems suspicious here.

  5. Filker0 says:

    I doubt STMicroelectronics is worried about the ARM Cortex in the GD chips; if the peripherals are cloned from proprietary designs, they may take issue. I use the F4, F7, L4, and soon the H7 from STM, which are much more capable parts with more RAM+Flash and better peripherals, though in my experience far from perfect.

    There’s room in the market for additional chip makers. I’ve not used the parts from GD, I find that the decision comes down to which part has the peripherals and connectivity needed in the power budget and footprint, then the software folks like me help make sure the memory (both flash and sram) is sufficient. Bigger RAM and higher clock rates usually lead to higher power requirements.

    • Sjaak says:

      They are available with more flash then their ST counterparts which makes them more interesting. You don’t have to port the code, but use another linkscript instead.

      They also have a gd32f2 and gd32f4 ARM chips with more memory and speed then the ST counterpart. Havent compared them though…

    • KH says:

      After reading all the comments, I have questions: Does this mean STM32 and AVR boards/parts from dodgy online sellers are suspect? What are your opinions? (I have too many MCU stocked from RS/Farnell so I never bothered to look closely at these things. But I did notice a recent inexplicable upsurge in STM32 proto boards offerings, hmmm.)

      What about Arduino boards? Since these things have been around for a while, why don’t I see LGT8F88A or GD32 on more online seller sites? Are they trying to sell quietly under the radar as legit substitutes or are they laser-remarked and used to substitute for STM32/AVR parts? So far I haven’t exactly noticed EETimes or one of those covering this.

Leave a Reply

Notify me of followup comments via e-mail. You can also subscribe without commenting.

Recent Comments