• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

New Prime95- version 26.3, build 3

Prime95v.28.1 Build 1 (64-bit)

http://www.mediafire.com/?awambpo41sy6qns

From developer:
For any Haswell owners that are interested, an evaluation version 28.1 is available. I have some ideas to improve it further, but they will take some time to implement.

I sure hope it works correctly because I've started using it on my Haswell box
 
thanks, does this also detect correctly vishera cpus?
 
Tried with my Haswell and it is harder to pass than prime95 with AVX. My previous stable setting on Prime95 AVX wont stable with this which means more time for tinkering :D
 
Prime95v.28.3 Build 1 Beta (32/64-bit)

ftp://mersenne.org/gimps/p95v283.win32.zip
ftp://mersenne.org/gimps/p95v283.win64.zip
ftp://mersenne.org/gimps/p95v283.linux64.tar.gz
ftp://mersenne.org/gimps/p95v283.MacOSX.zip

Prime95 version 28.3 is available. This version is for you intrepid beta testers. I've run the torture test and done some QA, but I haven't run any double-checks yet. I don't expect any problems, and I am now using it on my computers.

This version includes nearly all the Haswell optimizations I plan on implementing. A few of these changes may also benefit Sandy/Ivy Bridge machines. On my Sandy Bridge I'm seeing a 1.5% improvement.
 
Last edited:
Should be called Prime14
 
Why?

Stasio, are you registered on ALL IT related forums around the world? I honestly can't think where I didn't see you yet, lol.
 
Or Prime 8. Because we are past "95"
 
Prime95v.28.5 Build 1 Beta

ftp://mersenne.org/gimps/p95v285.win32.zip
ftp://mersenne.org/gimps/p95v285.win64.zip
ftp://mersenne.org/gimps/p95v285.linux64.tar.gz
ftp://mersenne.org/gimps/p95v285.MacOSX.zip

1) Changed the output to the worker windows during LL and PRP tests. The new output includes
the estimated time to complete the test. There are two new options described in undoc.txt:
ClassicOutput and OutputRoundoff.
2) Added some new options described in undoc.txt: ScaleOutputFrequency, TitleOutputFrequency,
and SilentVictoryPRP.
3) Benchmarking on hyperthreaded machines now times only the most common cases. Specifically,
hyperthreading is used only in the one cpu and all cpu cases.
4) Benchmarking trial factoring is now off by default. Prime95 should not be used for trial
factoring. GPUs are about 100 times more efficient at that task.
5) On multi-core machines, benchmarks are now run on multiple workers. This measures the effect
of memory bandwidth during testing and helps you select the setup that gives you the most throughput.
6) There are many new options described in undoc.txt to customize the benchmarking process
 
Last edited:
@stasio : Your last few posts with links to ftp are broken. The link itself (not the text,) is missing a colon after ftp so links are coming up ftp//mersenne.org/gimps/p95v285.zip. Try clicking on the links and you'll see what I mean.
 
Prime95v.28.6 Build 1

ftp://mersenne.org/gimps/p95v286.win32.zip
ftp://mersenne.org/gimps/p95v286.win64.zip

1) The "days between sending new end dates" preference now accepts values between 0.125 and
7.0 (was between 1 and 7). This lets you contact the server as frequently as every three
hours. This can be useful in conjunction with the server feature that sends an email to
you if the computer is more than one hour late contacting the server. To turn on that
server feature, go to the CPUs web page, and click on the CPU, there you can turn on a
couple of email features.
2) AVX2 support for trial factoring. About a 50% speedup over the previous version. However,
all trial factoring should still be done on a GPU. A GPU is on the order of 100 times more
efficient at trial factoring than a CPU!!!
 
Prime95v.28.9 Build 2

ftp://mersenne.org/gimps/p95v289.win32.zip
ftp://mersenne.org/gimps/p95v289.win64.zip

New features in Version 28.9 of prime95.exe
-------------------------------------------

1) Since GPUs are so much better at trial factoring than CPUs, benchmarking no longer times
prime95's trial factoring by default. Two new benchmarking options are available:
OnlyBenchThroughput and OnlyBenchMaxCPUs. See undoc.txt for details.
2) Slightly reduced the memory bandwidth requirements for several large FFTs. May lead to
a very small speed increase for users testing 100 million digit numbers.
3) If running more than worker, prime95 looks for any sin/cos data that it can share among
the workers. Depending on the FFT sizes you are running, this could lead to a very slight
reduction in needed memory bandwidth.
4) Method for choosing the best FFT implementation changed. In previous versions, the FFT
implementation that resulted in the fastest single worker timing was used. In this version
the FFT implementation that had the best throughput was selected. For FMA3 FFTs I used a
4-core Skylake to measure best throughput. For AVX FFTs I used a 4-core Sandy Bridge
to measure best throughput. Not many FFTs were affected, but you may see a few percent
variation in throughput with this version.
5) Improved AVX2 trial factoring in 64-bit executable. Trial factoring should still be done
on a GPU. A GPU is on the order of 100 times more efficient at trial factoring than a CPU!!!
6) Trial factoring now defines one "iteration" as processing 128KB of sieve, or 1M possible
factors. In previous versions an iteration was defined as 16KB of sieve in 32-bit executables
and 48KB in 64-bit executables. The trial factoring benchmark still times processing 16KB of sieve.
7) Trial factoring in 64-bit executables is now multi-threaded.
8) On initial install, the default settings for number of worker windows will be set to
the number of cores / 4 with multithreading turned on.
9) The worker windows dialog box now enforces a minimum number of multi-threaded cores for some
work types to ensure timely completion of assignments. Also, the worker windows menu choice
no longer allows assigning work to hyperthreads (they are rarely beneficial in prime95).
This behavior can be overridden with the ConfigureHyperthreads undoc.txt feature.


 
Download links don't work :(
 
You can ftp this directory and find them: ftp://mersenne.org/gimps/

For some reason the link button, in the editor, adds ""http://" in front of the ftp address.
 
It would have been better if this stasio guy (who seems to be employed fulltime to post this sort of stuff on every single IT-related forum on the internet) posted a link to the appropriate forum instead of direct links that might change in future.
 
Prime95v.28.10 Build 1

ftp://mersenne.org/gimps/p95v2810.win32.zip

ftp://mersenne.org/gimps/p95v2810.win64.zip


I took down the links to 28.9 because of recently discovered torture test issues on SSE2 machines.
There is little reason to upgrade from 28.7. Users have reported speeds plus-or-minus a couple of percent.
 
Last edited:
Prime95 v.29.7 Build 1

ftp://mersenne.org/gimps/p95v297b1.win32.zip
ftp://mersenne.org/gimps/p95v297b1.win64.zip

1) Support added for AVX-512 FFTs.
2) FMA3 FFTs now have slightly higher FFT crossover points. Soft crossovers are
no longer used by default. See undoc.txt.
3) Torture test dialog box options now based on cache sizes. Options for performing
a weaker torture test are available. Torture tests that use all RAM are now more
stressful. In-place vs. not in-place memory accesses now displayed on screen.
On machines with more than 5GB of memory, blend defaults to all but 3GB of RAM.
4) Add & subtract operations for AVX-512 FFTs are now multithreaded. This should
improve performance for P-1 and ECM when using multiple threads.
5) Benchmark results are now written to results.bench.txt.
6) JSON results are now available for all work performed. JSON results are
written to results.json.txt.
7) PRP tests with Gerbicz error checking are more immune to hardware errors.
 
Prime95 v.30.3 Build 2

ftp://mersenne.org/gimps/p95v303b2.win64.zip

Chanelog included in file.
 
Back
Top