Home > On-Demand Archives > Q&A Sessions >

Live Q&A with Jacob Beningo - Running DSP Algorithms on Arm® Cortex®-M Processors

Jacob Beningo - Watch Now - DSP Online Conference 2021 - Duration: 28:56

Live Q&A with Jacob Beningo for his talk titled "Running DSP Algorithms on Arm® Cortex®-M Processors"
M↓ MARKDOWN HELP
italicssurround text with
*asterisks*
boldsurround text with
**two asterisks**
hyperlink
[hyperlink](https://example.com)
or just a bare URL
code
surround text with
`backticks`
strikethroughsurround text with
~~two tilde characters~~
quote
prefix with
>

jmdebalasy
Score: 0 | 3 years ago | 1 reply

Awesome talk. Very informative, concise and clear. Thank you!

Jacob_BeningoSpeaker
Score: 0 | 3 years ago | no reply

You?re welcome!

15:02:59	 From  John Phillips : Digital signal processing always seems to require a largish number of input samples. What if these are not available ? For example, explosive decompression. What typical minimum input sample size is neessary for most solutions ?
15:07:07	 From  Radu Pralea : I'm not familiar with this notion, "explosive decompression". What is it?
15:08:01	 From  Alex : How does the ARM compiler choice affects performance of DSP algorithms running on Cortex-M processors?
15:12:40	 From  Emanuele Ziglioli : Cortex M4s are everywhere. Haven't seen many M7s, what happened?
15:14:16	 From  Emanuele Ziglioli : Have seen some parts with external RAM, maybe that was the blottleneck, clock speed with onboard RAM
15:15:22	 From  Michael Kirkhart : Yes, I have that same problem - lots of interesting things to look into, but only so much time!
15:15:52	 From  Robert Edwards : What low cost/entry evaluation/development board do you recommend for "running DSP Algorithms on Arm® Cortex®-M Processors"?
15:18:14	 From  Emanuele Ziglioli : There's been so much integration in the vendor space. I love NXP's LPC boards, hate legacy boards from Freescale (Kinetik, much harder to use, different peripherals). Kudos to NXP for trying merging tools, still painful though, what do you think?
15:19:51	 From  Radu Pralea : There was a trend (or at least that's how I perceived it) towards multicore MCUs, usually heterogeneous, but I fell it hadn't enough traction really. Currently there are (at least) a couple of pretty popular pieces on the market, but I'm not sure how much they're used in real projects. From your perspective, do you feel there's maybe no applications requiring such real-time computing power in the general purpose space, or...? (I know there are many pieces targeted for automotive and other use-cases, but those aren't really COTS, so...)
15:19:54	 From  Dave Comer : If you are using Bluetooth/WiFi, SiLabs has some very interesting offerings as well. And Cheap boards!
15:20:27	 From  Robert Edwards : thanks, lots of choice, I have STM-H board so I'll try that
15:21:46	 From  John Phillips : I haven't seen much other than specialist core/ADC/DAC ASIC chips perform SDR. Do you know if Cortex M devices can perform useful SDR with the current available libraries ?
15:23:34	 From  Devendra Chaudhari : Is there any scope of further optimisation for the code generated from modeling tools like matlab for cortex m?
15:26:30	 From  Dave Comer : With ZYNQ, you can get up to quad core (A9)'s and have an FPGA with trust zone. You get the power, the DSP, and security in one package.
15:29:38	 From  Dave Comer : Thanks. Great discussion.
15:30:30	 From  Alex : Thanks a lot Jacob.

OUR PARTNERS