Semiwiki on LinkedIn Semiwiki on Facebook Semiwiki on Twitter Register / Log in 
Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/index.php?threads/mask-rom-footprint-and-read-out-times.10729/
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2021370
            [XFI] => 1050270
        )

    [wordpress] => /var/www/html
)

Mask ROM footprint and read-out times

H

horscht(i)

New member
I am looking for numbers or at least estimation for the footprint (area per bit) and read-out times of mask ROM versus Flash in a given process. The process can be as old as 40nm or even 65nm.
After googling around for an hour the only rather solid statement I could find is "As a consequence, energy/bit can be a factor 20 smaller and read-out 100-fold faster than 16nm flash." (
Chips 2020: A Guide to the Future of Nanoelectronics​
, page 245).

"100-fold faster" read-out sound a Little dramatic to me, though.
If there ever was a significant difference in footprint, I assume flash has now the upper hand since multi-Level cells are around.
 
More considerations than read speed and silicon area

horscht(i) said:
I am looking for numbers or at least estimation for the footprint (area per bit) and read-out times of mask ROM versus Flash in a given process. The process can be as old as 40nm or even 65nm.
After googling around for an hour the only rather solid statement I could find is "As a consequence, energy/bit can be a factor 20 smaller and read-out 100-fold faster than 16nm flash." (
Chips 2020: A Guide to the Future of Nanoelectronics​
, page 245).

"100-fold faster" read-out sound a Little dramatic to me, though.
If there ever was a significant difference in footprint, I assume flash has now the upper hand since multi-Level cells are around.
Click to expand...

Embedded flash can attain sub 10nS read access times. Silicon area competitiveness will hing on the flash technology used and the overall density, since a small (say sub 64kB) block will still carry the same overhead of charge pump etc as a (say) 16Mb block.

Other considerations are IP availability (embedded flash is widely available from 450ns to 28ns today) and security. ROM is more readily reverse engineered whereas embedded flash can contain many DPA/security measures which destroy the information when an attack is sensed.
 
Thanks!
As usual in life the problem solved itself.
There were some timing issues, but it looks like we can make the same timing on some test cases with flash even on older tech nodes.
 
You must log in or register to reply here.
Back
Top