Author Topic: STM32V - cortex M85  (Read 5734 times)

0 Members and 1 Guest are viewing this topic.

Offline dmendesfTopic starter

  • Frequent Contributor
  • **
  • Posts: 432
  • Country: br
STM32V - cortex M85
« on: May 08, 2025, 06:32:42 pm »
A bird told me today that ST will launch a cortex M85 chip (even said it will be called STM32V). Anybody got more info? Seems it will be launched in July...
 

Offline mark03

  • Frequent Contributor
  • **
  • Posts: 783
  • Country: us
Re: STM32V - cortex M85
« Reply #1 on: May 08, 2025, 11:45:22 pm »
There have been "leaks" claiming that it is coming for at least a year.  See the "STM32 roadmap" thread.  It ST confirms that it is real, that would be interesting.  Perhaps they need to "freeze the market" a bit to stave off an exodus to other vendors.

If they announce this summer I would be shocked if silicon hits Digikey/Mouser before mid-2027.  For comparison, STM32N6 was first *demonstrated* way back in Spring 2023, and possibly announced well before that---I can't remember.  The timeline from public announcement to general availability is quite a bit longer than it used to be.
 

Offline Datom

  • Contributor
  • Posts: 11
  • Country: cn
Re: STM32V - cortex M85
« Reply #2 on: May 09, 2025, 08:04:03 am »
yes,STM32V8
 

Offline dmendesfTopic starter

  • Frequent Contributor
  • **
  • Posts: 432
  • Country: br
Re: STM32V - cortex M85
« Reply #3 on: May 10, 2025, 07:36:34 pm »
Thank you for the pointers!
 

Offline Sal Ammoniac

  • Super Contributor
  • ***
  • Posts: 1813
  • Country: us
Re: STM32V - cortex M85
« Reply #4 on: May 20, 2025, 08:54:20 pm »
Anyone else have a CM85 in production yet? The only one I'm aware of is from Renesas.
"That's not even wrong" -- Wolfgang Pauli
 

Offline dmendesfTopic starter

  • Frequent Contributor
  • **
  • Posts: 432
  • Country: br
Re: STM32V - cortex M85
« Reply #5 on: May 21, 2025, 12:37:30 am »
It's the only one I know about.
 

Offline Shalomie42

  • Contributor
  • Posts: 15
  • Country: us
Re: STM32V - cortex M85
« Reply #6 on: May 21, 2025, 06:10:25 pm »
The hardest part of being under NDA with ST is that you can't share information with anybody that would actually care to talk about technical details with you. From my experience in industry, there are very few fellow engineers who get giddy about cool new MCUs and their features. I have flowed through 5 fortune 500 companies and engineering teams as a CEN and I think I might have interacted with one or two true comrades.

Take what I said as you will, but I will say several vendors including ST are not sitting around. MCU silicon in general is a lot more time consuming compared to say a desktop PC where new chips abound every 6 months. MCU fabrication just doesn't have that momentum. However it is gaining year over year as vendors switch to smaller nodes.

Quote
A bird told me today that ST will launch a cortex M85 chip (even said it will be called STM32V). Anybody got more info? Seems it will be launched in July...

That bird is outdated. Don't hold your breath.
« Last Edit: May 21, 2025, 06:16:05 pm by Shalomie42 »
 

Online SiliconWizard

  • Super Contributor
  • ***
  • Posts: 17842
  • Country: fr
Re: STM32V - cortex M85
« Reply #7 on: May 21, 2025, 06:23:14 pm »
That bird is outdated. Don't hold your breath.

Oh really. Does that mean much later than July, or never at all? :popcorn:
 

Offline Shalomie42

  • Contributor
  • Posts: 15
  • Country: us
Re: STM32V - cortex M85
« Reply #8 on: May 21, 2025, 06:47:09 pm »
Let me give you some examples and perspective about ST.

ST is one of the companies that cares about performance crowns. You can see what some of the FAEs I have talked to refer to as "one-up manship" in just about all of their newer MCUs. NXP releases an M33 @ 150MHz, ST releases an M33 @ 160MHz. That's absolutely not coincidence. Trust me. A substantial (tens of MHz) safety margin is built into the silicon literally just for marketing wank, "one ups manship" opportunities like this.

NXP releases a much faster i.mx RT than they previously had by several hundred MHz. Our H7 competed well with what they had, but we need to do something for that crown! Our next designs are years away, so what can we do now as a stop gap? Queue H7's releasing @ 550MHz compared to 480MHz. The H7S/R are exactly in that same boat. They are literally squeezing the MHz out of the existing design because that's all they have right now. That and minor silicon enhancements to gain a few MHz.

I believe DMIPS for DMIPS, the recently released N6 is similar in performance to the RA8, but still loses by some.

Given that ST cares deeply about being on top and its been a while since the last high performance class design released (H7 - CM7), do you think they will or won't release something that competes with the RA8?

Those XML files seem to be interestingly accurate at predicting future devices don't they? I think one of the ones I saw awhile ago had the H7R/S and some of the lower variants of the H5 series in it which hadn't released yet. When those lower variants popped out a few months later, it even caught me off guard at how quickly they came after the H56/7 launched.

I have NDA / FAE relationships with many of the major MCU players. When I want to discover new MCUs or solutions that may be coming in the future, I always go to ST. Not NXP, not Renesas, etc. Why? Look at their track records and cadence history for MCUs.

  • ST is releasing new chips every 6 months, regardless of performance level. I know I said MCU silicon takes time and doesn't have momentum. That is true, but what is also true is that ST executes their designs with years of planning such that they can game the system to do releases like that. They are very strategic. They release one variant of a new design and then trickle out the rest as they become available - all using the same blueprint they planned over 5 years. They release the variants that will have the largest ROI first.
  • NXP is very much so a joke with regards to MCU cadence. They are not even on the map for me. I look at their chips every once in awhile when they happen to release something, because they are so slow and uninteresting. Serious question, can you actually hold MCX silicon in your hands right now? I knew of those chips 4 or 5 years ago. They "released" it 3? years ago?
  • Renesas has had a very horrible reputation in many circles with regards to their MCUs and chips in general, including mine. My current company won't even touch them. I am the crazy one for bringing up GreenPak every once in awhile. Ah right! That one wasn't Renesas - it was Silego. That whole R series Cortex M lineup they spent years formulating and are just now releasing? Those chips are very cute as display pieces. Many many many many many many, did I mention many? promises that are never fulfilled. Forge FPGA anyone? If I had a penny for every whizbang new MCU technology they tried to sell people on over the years that were duds, I would have about $0.5. Also, their MCU peripherals have a horrendous reputation.

« Last Edit: May 21, 2025, 07:34:34 pm by Shalomie42 »
 

Offline Sal Ammoniac

  • Super Contributor
  • ***
  • Posts: 1813
  • Country: us
Re: STM32V - cortex M85
« Reply #9 on: May 21, 2025, 09:50:18 pm »
Renesas ... Also, their MCU peripherals have a horrendous reputation.

How so? What don't you like about their peripherals?
« Last Edit: May 22, 2025, 06:59:58 pm by Sal Ammoniac »
"That's not even wrong" -- Wolfgang Pauli
 

Offline rf-fil

  • Regular Contributor
  • *
  • Posts: 140
  • Country: au
    • VK2ZJ at QRZ
Re: STM32V - cortex M85
« Reply #10 on: May 22, 2025, 03:45:24 am »
    Renesas ... Also, their MCU peripherals have a horrendous reputation. [/li][/list]

    How so? What don't you like about their peripherals?

    One anecdote here: It was a major ordeal to configure just basic peripherals. Their documentation is complete shit and takes huge effort to decode. In one case, I ended up bit-banging I2C because I couldn't get their peripheral to work after a few days. This was on their RX2xx MCU. I will never use Renesas MCUs again.
    -VK2ZJ
     

    Offline Shalomie42

    • Contributor
    • Posts: 15
    • Country: us
    Re: STM32V - cortex M85
    « Reply #11 on: May 22, 2025, 10:00:15 am »
    I have personally not used Renesas MCUs. I am only pointing out their reputation. I had a few colleagues years back that were using their proprietary cored MCUs (not Cortex M) and they spoke about how bad the peripherals were. Not only were they a dog to configure and get working, but they always had crazy errata you had to get around.

    The point is I don't work with Renesas MCUs because of how bad their reputation is that preceded them. Has it changed with the Cortex M chips? Maybe. There peripherals look "ok" on paper as you read the RM. Unfortunately, even if I wanted to give them a chance, I couldn't because the rest of my business doesn't want to touch Renesas. They have poor reputations across the board.

    The Cortex M cores that Renesas is spending so much time to build out I think will be too little too late. I also have to point out those MSPM0's from TI. Those things really jostled ST. A Cortex M0+ analog chip with dual CAN FD, dual ADCs, a trig unit, 80MHz core, etc.. All for a few dimes. The interesting part is where those things came from and why TI released them. When I was working at Stanley, their tools division wanted a cheap MCU to replace their current MCU they were using for their brushless drill drives with the specific bells and whistles they wanted. You know those cheap cordless Dewalt drills you can buy. I have a bunch of Dewalt tools myself. TI designed those chips with the Stanley tools team. You can do that kind of stuff when your volumes are millions of units per year.
    « Last Edit: May 22, 2025, 10:24:47 am by Shalomie42 »
     


    Share me

    Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
    Smf

     

    - - -
    -->