Prime95

Prime95 30.8 build 17

  • Huge improvements to stage 2 of P-1 when lots of memory is available.
  • Warning: upgrading in the middle of P-1 stage 2 will restart P-1 stage 2 from scratch
  • Support for Ryzen 7000?
New features in Version 30.8 of prime95.exe
-------------------------------------------

1) P-1 stage 2 using fast polynomial multiplication similar to the GMP-ECM program. If lots
of memory is available for stage 2 this implementation will be substantially faster.
  • Same as build 14 with a benchmarking bug fixed.
  • Same as build 13 (post #490) with the memory over-allocation on small exponents fixed.
  • Same as build 12 with the few reported bugs fixed.
30.8 build 12

30.8 is now feature complete! Let's call it a beta release.

Changes:

  1. Fixed nordi's stage 2 init bug (D > 510510).
  2. hanged the way "FFT switching" occurs at stage 2 init. You'll no longer get these messages unless "PolyVerbose=1" is set. There is a small chance this version will switch to a smaller FFT size resulting in a negligible speed increase.
  3. This version should not over-allocate memory. For small exponents, say below 50000, the previous version could over-allocate memory by several GB.
  4. Three new D values added to cover a big gap between 10 million and 29 million.
  5. Fixed bug where MaxHighMemWorkers=1 was not obeyed.
  • Trial factoring crashes. Fixed in build 9.
  • Options/Benchmark tries to run a hyperthreaded benchmark on non-hyperthreaded CPUs. Fixed in build 6.
  • Another possible crash bug in stage 2 init when memory settings change. Fixed in build 6.
  • ECM sometimes generated excessive roundoff error, usually at start up which then forced using a larger FFT size than necessary. Fixed in build 6.
  • On stage 2 restart due to more memory now available, stage 2 % complete was erroneously reported to be 100%. Fixed in build 8.
  • On stage 2 restart due to less memory being available, stage 2 might restart from scratch. Fixed in build 8.
  • Rare radix conversion excessive roundoff error affecting PRP of non-base-2 numbers. Fixed in build 8.
  • Assume CERTs will complete before all other work types in computing estimated completion dates. Fixed in build 5.
  • A low-memory situation during stage 2 init of ECM could lead to a crash writing a save file. Fixed in build 5.
  • During stage 2 init, checking for a restart due to a reduction in available memory was infrequent. Fixed in build 5 - might reduce chance of an out-of-memory event.
  • Hyperthreaded torture tests not setting affinity properly for small FFTs. Fixed in build 4.
  • Hyperthreaded in-place torture tests crash for small FFTs. Fixed in build 4.
  • Semi-obscure ECM crash. If an ECM curve needed modular inverses in stage 2 and a subsequent curve needed none (more memory available), then a crash occurred. Fixed in build 4
Zurück
Oben Unten