2010-11-25 10:50 AM
ST's Cortex-M4.
2011-05-17 05:16 AM
There is a dual core Cortex-A9 part.
What specific features of the M4 are you looking for?2011-05-17 05:16 AM
2011-05-17 05:16 AM
What does your distributor say?, they often have some visibility of future product releases......
Chris.2011-05-17 05:16 AM
''What does your distributor say?, they often have some visibility of future product releases......''
Indeed. And, if yours is a really serious enquiry, they might be able to tell you something under NDA...
Otherwise, if ST have not yet anounced it, they're hardly likely to say anything much on a public internet forum - are they?!2011-05-17 05:16 AM
Subject Body Category Tags New STM32F2xx zero-wait program execution up to 120MHz Subject Body Category Tags
STM32 embedded Flash performance gets double boost with 90nm production availability and Adaptive Real-Time accelerator enabling zero-wait program execution up to 120MHz
Today, we announce two significant advances that further improve the performance and power consumption of our successful STM32 family: availability of production devices featuring embedded Flash at 90nm process technology; and the industry’s first Adaptive Real-Time (ART) memory accelerator optimized for the STM32’s industry-standard ARM® Cortex™-M3 processor.
The first production STM32 microcontrollers leveraging ST’s 90nm embedded Flash technology, which is already proven in smart card and automotive ICs, deliver faster operation, increased peripheral integration, lower power consumption, and increased on-chip memory densities. ST announced it was sampling these devices in 2
The proprietary ART memory accelerator balances the inherent performance advantage of the ARM Cortex-M3 over Flash memory technologies, which normally requires the processor to wait for the Flash at higher operating frequencies. The CPU can now operate up to 120MHz without waiting, thereby increasing overall system speed and efficiency.
To release the processor’s full 150 DMIPS performance at this frequency the accelerator implements an instruction pre-fetch queue and branch cache, enabling program execution from Flash at up to 120MHz with zero wait states. Competing Cortex-M3 MCUs can now only outperform the STM32 by operating at frequencies above 120MHz, which will increase power consumption and heat dissipation.
With this performance now available, developers can host extra elements of a system on the microcontroller, saving the need to use a more expensive microcontroller or a companion DSP. An example is in multimedia applications, where customers will be able to implement an audio codec, video processing functions, data encryption, digital filtering, and a multi-protocol gateway, with sufficient remaining resources to manage other tasks.
The latest STM32 variants benefiting from the 90nm process and the ART memory accelerator have been verified according to CoreMark™ tests created by the Embedded Microprocessor Benchmark Consortium. CoreMark results verify the STM32 executes 8% faster than Cortex-M competitors at 100MHz clock speed. The performance advantage is greater still at 120MHz. CoreMark analysis also confirms dynamic power consumption of only 188µA/MHz (98µA/CoreMark). This is equivalent to drawing 5mA at 120MHz (executing from Flash memory, with ART accelerator enabled and all peripherals off).
The new 90nm devices featuring the ART memory accelerator are sampling at lead customers. Product details will be announced later this year.
http://www.st.com/stonline/stappl/cms/press/news/year2010/t24htm
Subject Body Category Tags Product Subject Body Category Tagshttps://my.st.com/public/STe2ecommunities/mcu/Tags.aspx?tags=STM32
2011-05-17 05:16 AM
I'd still like to know the intrinsic speed of the flash, the current iteration is in the 35-40 ns range. Most of the techniques to speed things up involve wider memory lines, buffering and prefetch. Which is great for linear flow, but less so in the usual mix. It is not clear to me if this ''zero-wait state'' is actually that, or that it is more dynamic, where it's zero most of the time, but stalls on a ''miss'' condition.
Fujitsu has some 15-17 ns flash memory in it's FM3 part (almost certainly a 90 nm geometry design), so I somehow doubt that ST can sustain the 7-8 ns in less than optimum conditions.2011-05-17 05:16 AM
Hi,
This one taken from MCU newsletter (read below) JanToday is a big day for STMicroelectronics which has released its roadmap for its STM32 family of 32-bit MCUs based on ARM Cortex-M processor cores. This for the first time includes devices centered on the Cortex-M4 and the Cortex-M0 cores from ARM as well as the introduction of STM32 F-2 series MCUs based on the Cortex-M3 core.
The Cortex-M4 from ARM is an upwardly compatible version of the Cortex-M3, offering DSP instructions and a floating point unit (FPU). The core will be used in STM32 products to be sampled in 2011. The STM32 portfolio will also be widened with the sampling in 2011 of MCUs developed around the Cortex-M0. ST says all future STM32 Cortex-M4 and -M0 based MCUs will remain fully pin- and software compatible with each other and the STM32 Cortex-M3 devices. They will benefit from the same tool and software ecosystem, allowing seamless moves within the full STM32 product range. The news and links to the technical specifications are availablehttp://newsletters.eetimes.com/portal/wts/ccmciv2cnViLecwzm6rU-x2j08mvi
. Colin Holland Editor - MCU Designlinemailto:colin.holland@ubm.com
2011-05-17 05:16 AM
So the answer to the original question, ''When ST will announce its Cortex-M4?'' is:
''Today!''2011-05-17 05:16 AM
''The news and links to the technical specifications are available
http://newsletters.eetimes.com/portal/wts/ccmciv2cnViLecwzm6rU-x2j08mvi
.''Errr... no they're not! That link just goes to a page with no content - lots of adverts, but no actual content!