Why is my i5 from 5+ years ago not as good as the ones out today?
It has basically the same clock speed and number of processors as the ones coming out today.
>>54518003
Put it on the microwave to recharge the GHz.
IPC
Think of it like improving how powerful each Ghz is.
Higher IPC but lower clockspeed can overcome lower IPC and high clockspeed.
Intel has been pretty consistent in giving ~5-10% IPC improvements per generation. So by now even at the same clock speed, you will likely be a good 20-30% slower, depending on what you're doing.
>>54518003
Newer ones do more per clock -- same reason as why a 3 GHz i5 would beat the crap out of a 3 GHz Pentium 4 even if limited to using one core.
>>54518003
Newer i5 chips from Intel have newer instructional sets, faster/more cache along with DDR4 support. This does translate into 5-15% IPC gain depending if the application supports new instructional sets found in Haswell and Skylake.
>>54518161
>5+ years
>10% improvement if and only if the program has been compiled with certian flags enabled
>>54518022
Fucking this. I got my 2nd gen i5 to 3.9 GHz with my microwave, I'm afraid to go any higher. Cards still doing just fine.
>>54518569
What was it originally? Microwaves are about 2.4 GHz, so that's what you can add to the stock frequency with them.
Why did all technological process stop?
>>54518481
I looked at your post and it sucked
>>54518710
Because we're running up against the limits of silicon. So enjoy very small progress until a viable alternative is found.
>>54518003
>Why is my i5 from 5+ years ago not as good as the ones out today?
Because Intel wants your money.
>>54518751
A viable silicon alternative has been found
Just as blue leds have been discovered too
Like, the time it's taking for this tech to reach the masses makes you think...
>>54518944
Yeah. When things are first invented they work properly under precisely controlled laboratory conditions, for relatively simple circuits, some of the time, and take many bad production runs to get one that works at all. It takes time to go from that to something that works reliably, under real life conditions, and can be mass produced cheaply with high enough yield to be economical. Until then we're stuck with the previous perfected technology.
>>54518003
this is bait
but it's all about architecture
clock speed and cores have never mattered for shit when comparing across even microarchitectures
in the modern day however it doesn't matter as much unless you're a gaymer fag, advancement has been slowing for over a decade and new ISA extensions are generally useless for the majority of software you run in a home environment
>>54518035
I wish people would stop using this synthetic buzzword like it has any real meaning for real world work
>>54522152
>complains about buzzwords
>uses the word "synthetic"
