data61 cuda-fixnum issues

data61 / cuda-fixnum

Extended-precision modular arithmetic library that targets CUDA.

Other

40 stars 28 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Sorry, cudaMallocManaged() is not implemented in the current version.

#72 guiming-shi opened 2 years ago
1
gentests.py doesn't generate tests by default?

#71 jkrauska closed 5 years ago
2
cmake and autodetect gpu

#70 jkrauska opened 5 years ago
15
Exponentiation benchmark pegs CPU, hangs forever

#69 imeckler closed 5 years ago
5
At Master branch, make failed with errors

#68 Chenfengldw closed 5 years ago
3
Fix build

#67 unzvfu closed 5 years ago
0
Benchmarking program fails when no parameter specified

#66 unzvfu closed 5 years ago
1
Build failure on cuda 10.1 with RTX 2080

#65 imeckler closed 5 years ago
5
Document required prerequisites

#64 jkrauska closed 5 years ago
1
Implement faster Newton-Raphson

#63 unzvfu opened 5 years ago
1
Investigate specialised mulmod for base*cuml in modexp

#62 unzvfu opened 5 years ago
1
Rewrite test suite to run via Python interface

#61 unzvfu opened 5 years ago
1
Re-jigger the API to ease writing HLL interfaces

#60 unzvfu opened 5 years ago
1
Implement Python interface

#59 unzvfu opened 5 years ago
1
Investigate different Python interfaces and pick one

#58 unzvfu closed 5 years ago
2
Implement REDC-based modnum

#57 unzvfu closed 5 years ago
0
Implement independent MODNUM concept

#56 unzvfu closed 5 years ago
0
Wrap everything in cuFIXNUM namespace.

#55 unzvfu closed 5 years ago
0
Faster full width multiplication and squaring

#54 unzvfu closed 5 years ago
0
Faster, smarter specialised sqr_wide implementation

#53 unzvfu closed 5 years ago
1
Faster mul_wide implementation

#52 unzvfu closed 5 years ago
1
Allow checking for overflow without penalising fast path

#51 unzvfu opened 5 years ago
1
Overhaul benchmarking system

#50 unzvfu opened 5 years ago
1
Use warp votes to branch on argument size to select fastest algo

#49 unzvfu opened 5 years ago
1
Work out why 32-bit digits is slower than 64-bit digits with same fixnum size

#48 unzvfu opened 5 years ago
1
Put all cuda-fixnum code in its own namespace

#47 unzvfu closed 5 years ago
1
Implement and test REDC-based mulmod

#46 unzvfu closed 5 years ago
2
Review consequences of Cuda 7 Independent Thread Scheduling on warp-synchronicity assumptions

#45 unzvfu opened 5 years ago
1
Benchmark against AccelerateHS "bignum" implementation.

#44 unzvfu opened 5 years ago
1
Understand why CLNW sliding-window is faster than k-ary in the tests

#43 unzvfu opened 5 years ago
1
Implement VLNW and compare to the current CLNW sliding-window method

#42 unzvfu opened 5 years ago
1
Enforce exponent sharing in sliding-window modexp function

#41 unzvfu opened 5 years ago
1
Modular exponentiation: sliding window implementation and window size selection

#40 unzvfu closed 5 years ago
0
Assess potential utility of CUDA warp matrix functions

#39 unzvfu opened 5 years ago
1
Consider reëxpressing fixnum layout in terms of CUDA's "coöperative groups"

#38 unzvfu opened 5 years ago
1
Investigate, document and incorporate "secure" arithmetic implementations

#37 unzvfu opened 5 years ago
1
Clean up warp_fixnum division code

#36 unzvfu opened 5 years ago
1
Ensure arguments from user functions are always read into registers

#35 unzvfu opened 5 years ago
1
Implement bignum arithmetic in terms of fixnum arithmetic

#34 unzvfu opened 5 years ago
1
Speed up test case generation using multiprocessing

#33 unzvfu opened 5 years ago
2
Combine multiple test case vectors in one prior to submission to the GPU

#32 unzvfu closed 5 years ago
1
Store test cases in compressed format

#31 unzvfu opened 5 years ago
1
Remove GMP dependency

#30 unzvfu closed 5 years ago
1
Investigate using Montgomery multiplication in the division algorithms

#29 unzvfu opened 6 years ago
1
Consider storing precomputed values in shared memory

#28 unzvfu opened 6 years ago
1
Investigate use of other PTX instructions in arithmetic implementations

#27 unzvfu opened 6 years ago
2
Feature Monty modular exponentiation

#26 unzvfu closed 6 years ago
0
Specialise multi_modexp to case where all exponents are small

#25 unzvfu opened 6 years ago
1
Support different moduli in different slots

#24 unzvfu closed 6 years ago
2
Ensure that device-side assertions are disabled in non-debug builds

#23 unzvfu opened 6 years ago
1