Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.
the batch file for linpack doesn't take that long to execute. if you ran linpack manually setting it to loop a problem size of say, 25000 or 30000, for hours it would probably fail too.
i've found the quickest way to get the heat up is to not use the batch file but instead run the .exe directly in interactive mode.
double click .exe, skip the first prompt (press enter), problem size -> 25000, leading dimension -> 25000, number of trials -> 1, data alignment -> 4
one trial...
same here, always an error after every test (if you uncheck 'stop on error'). tested on multiple systems, some that aren't even overclocked.
probably a bug between linx and the latest version of linpack.
haswell does this by design, even if you're running at stock clocks and stock voltage. the only way to prevent the extra voltage juicing is to set a fixed manual voltage.
is your motherboard running all cores at maximum turbo? remember, intel specifies that this cpu should be running at 3.7 with all cores loaded, but ultimately this is up to the motherboard manufacturers to implement. a lot of them 'cheat' and run all cores at 3.9 instead. this in conjunction...
it looks like edge to edge glass, but it's really not. none of these slim 'bezelless' ips panels use actual glass in front of the panel, it's thin plastic (probably polycarbonate) and because of the way these panels are constructed they superficially look like edge to edge glass in photos...
it depends on how the voltage is set. if you set voltage using a fixed manual voltage, the cpu will always be running at that voltage regardless of load or p-state. similarly, using a voltage offset will increase voltage by the specified amount everywhere. adaptive voltage is the only one...
this is possible, since when using adaptive voltage, the cpu will even more voltage when stress testing with a program that uses avx instructions. that means that if you run regular apps that don't use avx, the cpu is using the adaptive voltage you specified, which may not be stable at your...
aida64's fpu stress test is no joke when run by itself (you'll need to untick the other stress test boxes) - on my i5-4670k, it easily drove up temps 10°c higher over prime95 and linx. with the other stress test boxes ticked in addition to fpu, the temps don't get quite as high.
on one run...
keep in mind though that on current intel cpu's, the socket clamp pushes down on the ihs/lid to force the cpu down against the cpu socket contact pins. if you remove the lid, you'll need to put some sort of shim in its place.
offset isn't going to help. setting a positive offset will just increase voltage everywhere. the only way to prevent that .090v jump when stress testing is to use manual voltage.
it's possible you may have hit a voltage wall at the multiplier you're trying to run. try dropping the multiplier a step and see if that gets you stable at 1.2 volts or lower.
on my cpu, i had to push 1.255 volts (adaptive) to get it stable at 4.4, but it will run all day at 4.3 with only...