The last test I embarked on has finished! And the results are very interesting. See the results.txt file:

[CODE]
[Sat Jan 2 18:39:23 2016]
M14942437 is not prime. Res64: 683A0DFFC5827CD8. We8: E57106A7,7379210,00000000
M14942267 is not prime. Res64: C35562BC4F3511F3. We8: D8A74C7B,2423514,00000000
M14942209 is not prime. Res64: 8587C9937E3BED22. We8: CDAD4A41,7713418,00000000
M14942293 is not prime. Res64: 035EFC95F88CFC27. We8: 36084309,4746867,00000000
[Sun Jan 3 22:02:18 2016]
Iteration: 14329935/14942209, ERROR: FFT data has been zeroed!
Possible hardware failure, consult the readme.txt file.
Continuing from last save file.
[Sun Jan 3 22:26:02 2016]
M14942267 is not prime. Res64: D20C84656405F3FB. We8: FCFDD819,14910347,00000000
M14942539 is not prime. Res64: 0A930E56A9284971. We8: 7FE55A2A,1188977,00000000
M14942437 is not prime. Res64: 136153185F4D524F. We8: B81CE272,9576909,00000000
[Sun Jan 3 22:37:01 2016]
M14942497 is not prime. Res64: 80BD5A064693F1C0. We8: 0CAD30A7,2607443,00000000
M14942293 is not prime. Res64: 035EFC95F88CFC27. We8: 36502AEF,8394253,00000000
M14942567 is not prime. Res64: D233F12AC3781E04. We8: 59875C25,3894081,00000000
[Sun Jan 3 22:42:28 2016]
M14942563 is not prime. Res64: 6815BC39FCD7650F. We8: A94AFB88,2473090,00000000
[Sun Jan 3 22:55:12 2016]
M14942209 is not prime. Res64: 0AA69D2EA9100E22. We8: 7D077832,14397436,00010000
[/CODE]

The first four results belong to the first test I did with v27.9 and was done by two threads on each exponent. It matches the three tests I did on this machine using v28.7 and the data in the GIMPS database. These three other tests are described in an earlier post of mine.

The last test consists of eight threads (4 cores, 6700K processor) working on eight exponents. Throughout the 28 hours the testrun lasted no errors were reported except for one on thread 1 when 96% of the run was completed.

Notice that the results of this run do not match the previous runs except for one test. Of the four exponents that were tested for the first time on this machine concurrently with the other four, also one test did not fail whereas the other three did.

The two successfully completed tests matching with the database were running on the following threads:
[CODE]
[Worker #3 Jan 3 21:54] Iteration: 14460000 / 14942293 ...
[Worker #6 Jan 3 21:53] Iteration: 14620000 / 14942539 ...
[/CODE]

The threads 1 & 5, 2 & 6, 3 & 7 and 4 & 8 are the four pairs that each share one of the four physical cores.

Small complication is that due to glazed frost on the high tension power lines in the northern parts of The Netherlands there were some noticeable power cuts lasting milliseconds throughout the last 20% of the test run. This did not stop the machine running, and the web browser that was running on this machine did not fail either because of it.

I will now restart this exact test and finish it in an expected 28 more hours, I will probably be asleep when that test finishes but will report a few hours later.

If someone has a reasonable fast pre Skylake four physical cores with hyperthreading Intel machine available and feels like it than feel free to run the worktodo file from [URL="http://www.mersenneforum.org/showpost.php?p=420995&postcount=262"]post 262[/URL].

Just for reference, the outcome will be predictable, 8 correct residues, which is what I am looking four. The only such machine I have is about eight years old and would take too much time to run this test. Please post here if you embark on it.

If someone with another Skylake wants to run this test, than of course, feel free to do so.

Before I started the test for a second time on my Skylake machine I rebooted it.

[QUOTE=tha;421143]Before I started the test for a second time on my Skylake machine I rebooted it.[/QUOTE]

Did that enter more or less entropy into the system?

[QUOTE=chalsall;421146]Did that enter more or less entropy into the system?[/QUOTE]

I am assuming it is a joke.

But a serious answer to the question is that I don't think it made any difference. Just a safeguard.

[QUOTE=tha;421148]I am assuming it is a joke.

But a serious answer to the question is that I don't think it made any difference. Just a safeguard.[/QUOTE]

It was kind of a joke, but also a serious question...

We still don't understand what is happening. So, restarting might make sense. Then again, it might not.

In a perfect universe, we could capture the quantum state of the computing devices we use, and run many tests based on their initial states.

We humans are not that powerful, but we still have the ability to try.... :smile:

[QUOTE=chalsall;421146]Did that enter more or less entropy into the system?[/QUOTE]

Yes. (couldn't resist... LOL)

[QUOTE=Madpoo;421153]Yes. (couldn't resist... LOL)[/QUOTE]

Cool....

Haha, I was also reading it like "Did that, more or less, enter entropy into the system?" but you were faster with the answer...

Edit: Reading all this I feel sorry I don't own a Skylake... Itching hands to try some tests by myself.

[QUOTE=Madpoo;421153]Yes. (couldn't resist... LOL)[/QUOTE]

:tu::tu:

Some times contradictory discussions yield best results. One thing is sure, time will tell how this issue sorts out. I would not rush in.

[QUOTE=megabit8;421178]I would not rush in.[/QUOTE]

I was interested to see if this had been picked up by any mainstream media yet. I ran a few Google queries, and it appears it hasn't.

This is a good thing in my mind -- we want Intel and the motherboard manufacturers to have as much lead time as possible to solve what appears to be a very subtle bug without the hysteria which mainstream reporting often brings.

I did find [URL="http://www.anandtech.com/show/9607/skylake-discrete-graphics-performance-pcie-optimizations"]this article from AnandTech[/URL] on a tangential bug interesting. Also, that the [URL="https://communities.intel.com/message/361811"]Intel Forum thread on this matter hasn't been posted to[/URL] since December 31st.

It's only a matter of time before this is "out there". I do hope that Intel are taking this seriously....