SPEED-CC registration deadline
	If you intended to come to Berlin for SPEED-CC please register asap. Registration officially closes tomorrow, I can still keep it open for 1-2 days. [link] By Tanja Lange  - Sep 28 2009 - 1 new of 1 message

	SPEED-CC Call for papers
	Dear all, You might be interested in the upcoming workshop "SPEED-CC -- Software Performance Enhancement for Encryption and Decryption and Cryptographic Compilers" which will take place October 12&13, 2009 in Berlin, Germany. The web page is at [link] Please find the call for papers below.... more » By Tanja Lange  - Aug 4 2009 - 1 new of 1 message

	New AMD Packed Integer Rotates and Shifts instructions
	Hi, I have just read that AMD are introducing new set of instructions for "Packed Integer Rotates and Shifts". That is good news for x86 family of new processors. In my opinion many cryptographic algorithms can benefit from those instructions (at least several SHA-3 hash candidate functions that... more » By Danilo  - May 27 2009 - 1 new of 1 message

	constant-time extended GCD?
	I've been doing an x86/SSE2 implementation of curve25519 using ideas similar to the ones in "Fast elliptic-curve cryptography on the Cell Broadband Engine" ([link]). (It currently takes less than 260,000 cycles on Core 2 32-bit mode for a scalar multiplication using Montgomery form, not counting the final modular... more » By Wei Dai  - May 5 2009 - 1 new of 1 message

	secrets of GMP 4.3
	I noticed that GMP 4.3.0, released 2 days ago, contains new optimized assembly code. In mpn/x86_64/addmul_2.asm, there is this comment: C cycles/limb C K8,K9: 2.375 C K10: ? C P4: ? C P6-15: 4.45 C This code is the result of running a code generation and optimization tool C suite written by David Harvey and Torbjorn Granlund.... more » By Wei Dai  - Apr 16 2009 - 1 new of 1 message

	Jacobi quartic form curves
	I've been updating myself on the latest elliptic curve techniques, and found a March 2009 update of [link], which claims (based on theoretical calculations) that elliptic curve scalar multiplication should be fastest in Jacobi quartic form. Looking at the latest eBATS results, the fastest curves that have been implemented are gls1271 and... more » By Wei Dai  - Apr 16 2009 - 6 new of 6 messages

	SSE2 Comba multiplication
	eBATS results seem to show that Crypto++ in x86 32-bit mode ([link]) is competitive in speed with GMP in 64-bit mode ([link]) for some operations. This note describes a couple of the techniques used by Crypto++... more » By Wei Dai  - Apr 10 2009 - 1 new of 1 message

AES register saving tricks

The straightforward way of computing the AES round function requires 8 32-bit registers, but the x86 architecture provides only 7 useable ones. I noticed a neat trick that Brian Gladman used in his AES x86 assembly code to avoid spilling a register, and have an improvement upon it of my own. The basic problem is, for each of the four 32-bit register representing the... more » By Wei Dai  - Apr 6 2009 - 1 new of 1 message

	GCC compile farm
	Some of you probably know this already, but if you need to test your crypto code (and if it's free software) on different CPU architectures, you can get a shell account on the GCC Compile Farm. The URL is [link]. In the past I've used the Sourceforge Compile Farm and HP Test Drive for the... more » By Wei Dai  - Apr 3 2009 - 1 new of 1 message

	tips for instruction scheduling?
	Does anyone have tips for how to schedule a sequence of assembly language instructions, in order to maximize the instructions per cycle (IPC) executed by a CPU? What I've been doing so far is just simple trial and error. I'd move a few instructions around more or less at random, benchmark the resulting code... more » By Wei Dai  - Apr 3 2009 - 6 new of 6 messages