[Cryptech-Commits] [user/shatov/modexpng] branch master updated: One more cosmetic fix.
git at cryptech.is
git at cryptech.is
Tue Nov 26 12:27:09 UTC 2019
This is an automated email from the git hooks/post-receive script.
meisterpaul1 at yandex.ru pushed a commit to branch master
in repository user/shatov/modexpng.
The following commit(s) were added to refs/heads/master by this push:
new 6791175 One more cosmetic fix.
6791175 is described below
commit 6791175e7466ec09bf8464be99078ef33fe8da23
Author: Pavel V. Shatov (Meister) <meisterpaul1 at yandex.ru>
AuthorDate: Tue Nov 26 15:26:40 2019 +0300
One more cosmetic fix.
---
README.md | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/README.md b/README.md
index 17e19ee..a66c2f8 100644
--- a/README.md
+++ b/README.md
@@ -11,7 +11,7 @@ Compile-Time Settings
The core has one synthesis-time parameter:
- * **NUM\_MULTS** - Sets the number of DSP slices to use per modular multiplier, which must be a power of 2. Each multiplier consumes NUM\_MULTS + 1 slices, since one additional multiplier is required to eliminate the final conditional subtraction step of the Montgomery modular multiplication routine. The core internally consists of a pair of _dual modular multipliers_, thus the total number of occupied DSP slices is 4 * (NUM\_MULTS + 1). When in CRT mode, both _dual multipliers_ operate [...]
+ * **NUM\_MULTS** - Sets the number of DSP slices to use per modular multiplier, which must be a power of 2. Each multiplier consumes NUM\_MULTS + 1 slices, since one additional multiplier is required to eliminate the final conditional subtraction step of the Montgomery modular multiplication routine. The core internally consists of a pair of *dual modular multipliers*, thus the total number of occupied DSP slices is 4 * (NUM\_MULTS + 1). When in CRT mode, both *dual multipliers* operate [...]
Combined DSP slice utilization is outlined in the following table:
@@ -97,7 +97,7 @@ The core has the following registers:
Length of operand bank in bits. This read-only parameter returns the length of internal operand bank and allows the largest supported operand width to be determined at run-time. Currently BANK\_BITS is hardwired to always return 4096.
- * **NUM_MULTS**
+ * **NUM\_MULTS**
This read-only parameter returns the width of the internal parallel multiplier, that was specified at compile-time. This parameter is currently 8.
@@ -187,7 +187,7 @@ The core uses the Montgomery modular multiplier, which can only operate on numbe
F = (F2 < 0) ? F1 : F2
return F
-The final step of Montgomery modular multiplication is Montgomery modular reduction. It is done by adding a multiple of the modulus to the intermediate product. The multiple is selected in such a way, that the lower half of the sum is all zero bits, so it can be safely reduced by just a trivial right shift. This speeds things up, since there's no more need to do computationally difficult division anymore. To determine the multiple of the modulus, another quantity is required, which is th [...]
+The final step of Montgomery modular multiplication is Montgomery modular reduction. It is done by adding a multiple of the modulus to the intermediate product. The multiple is selected in such a way, that the lower half of the sum is all zero bits, so it can be safely reduced by just a trivial right shift. This speeds things up, since there's no more need to do computationally difficult division anymore. To determine the multiple of the modulus, another quantity is required, which is th [...]
R = 1
B = 1
--
To stop receiving notification emails like this one, please contact
the administrator of this repository.
More information about the Commits
mailing list