Double in format of iEEE754
Paul Stoffregen
Why I can not get a number of double format into 64bit format of ieee754?
Rick Swan
Regarding “Why I cannot get a number of double format into 64bit format of ieee754?”
On 8-bit Atmel CPUs, floating point (FP) is provided via software libraries. These libraries have been compiled for 32-bit FP support only. I think there may also be a “#define double float” in a header file somewhere (or perhaps I’m confusing with different vendor).
I suspect that supporting 64-bit FP is more complex than simply changing a preprocessor constant and rebuilding libraries because in many cases the FP libraries are hand optimized assembler for performance and the 64-bit routines may not be implemented.
Dave Gordon
Have you tried this to see what you should be getting?
--
You received this message because you are subscribed to the Google Groups "Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to developers+...@arduino.cc.
Netza Guadarrama
Thank you, I run it on arduino due and it had Ieee754 format of 64bit structure.
But why arduino uno does not have this?.
Sent: Wednesday, July 20, 2016 4:41 PM
Why I can not get a number of double format into 64bit format of ieee754?
https://www.arduino.cc/en/Reference/Double
Jim Leonard
The atmega processors have lots of limitations including space limitations and so this was one of the things that was left out.
On Fri, Jul 22, 2016 at 07:00:08PM +0000, Netza Guadarrama wrote:
> Thank you, I run it on arduino due and it had Ieee754 format of 64bit structure.
>
> But why arduino uno does not have this?.
>
> what is the reason?
> ________________________________
> From: Paul Stoffregen <pa...@pjrc.com>
> Sent: Wednesday, July 20, 2016 4:41 PM
> To: devel...@arduino.cc
> Subject: Re: [Developers] Double in format of iEEE754
>
>
>
> Why I can not get a number of double format into 64bit format of ieee754?
>
> https://www.arduino.cc/en/Reference/Double
> Reference Language | Libraries | Comparison | Changes. double Description. Double precision floating point number. On the Uno and other ATMEGA based boards, this ...
>
>
>
>
> --
> You received this message because you are subscribed to the Google Groups "Developers" group.
Netza Guadarrama
Thank you for the answer I understand a little more.
Arduino deside use ARM-M-3
* Arduino Due is using ARM-M
* Beaglebone is using ARM-A-v8
* RaspBerry is using ARM-A-V15
Why Arduino Due dont use ARM-A- V73 ?
I mean ARM-A instead of M.
Sent: Friday, July 22, 2016 7:06 PM
The atmega processors have lots of limitations including space limitations and so this was one of the things that was left out.
On Fri, Jul 22, 2016 at 07:00:08PM +0000, Netza Guadarrama wrote:
> Thank you, I run it on arduino due and it had Ieee754 format of 64bit structure.
>
> But why arduino uno does not have this?.
>
> what is the reason?
> [cid:c96890e3-0457-43a0-bcd0-97ed885fc472]
>
> ________________________________
> From: Paul Stoffregen <pa...@pjrc.com>
> Sent: Wednesday, July 20, 2016 4:41 PM
> To: devel...@arduino.cc
> Subject: Re: [Developers] Double in format of iEEE754
>
>
>
> Why I can not get a number of double format into 64bit format of ieee754?
>
> https://www.arduino.cc/en/Reference/Double
Paul Stoffregen
Why Arduino Due dont use ARM-A- V73 ?
The Cortex-A processors are generally meant for single board computers, where memory is provided by by separate DRAM and Flash chips (or a flash chip within a SD card). Software is usually based on a powerful but complex operating system like Linux.
Cortex-M are generally meant to be microcontrollers, where the processor and all memory usually is fabricated on a single piece of silicon. Compute performance is usually lower, because the silicon must implement flash memory. Power consumption is also usually much less, and typically the chips implement many low-power modes. Usually these microcontrollers have more on-chip peripherals and better on-chip analog features. Cortex-M also features very low-latency interrupts. Software is usually "bare metal" or RTOS-based, for superior responsiveness to I/O events.
ARM makes the Cortex-M cores and microcontrollers exist in the market because responsive I/O, simpler hardware (and software), and lower power consumption matter much more than high CPU computing power for many real world applications
Traditionally, most Arduino compatible boards and most Arduino software and libraries are based on microcontrollers, not single board computers. That's why you tend to see Cortex-M chips, not Cortex-A.
Netza Guadarrama
Thank you, for your time and response.
My best regards.
Sent: Friday, July 22, 2016 9:00 PM