Why does the GP106 PCB have so much capacitors? It's usually 2 caps per VRM phase, why does this one have three?
Even the much larger big brother GP104 has two caps per phase.
>>55508421
Probably because the caps have lower capacitance individually?
tbqh graphics cards can't have enough caps with all those massive power consumption peaks all the time
Probably has large spikes.
>>55508447
The PSU does so it's okay.
>>55508421
Am I getting trolled here? Who thought this up? 1 mosfet, 1 VRM, 3 caps and 1/3 driver per phase!?
>>55508475
>he thinks the PSU is ironing out everything
you know what's between the graphics card and the PSU?
yeah, exactly that will feel all those delicious spikes all the time first.
Those last 2 caps don't seem to be even routed anywhere.
Is this a rushjob PCB? Consider I've seen screenshots of this being a A1 revision, it's probably very likely.
>2 memory pads unused
256bit GPU?
>>55508542
No, the PCB is probably gonna be shared with a third bin GP104 if yields are as bad as they seem.
And it will be called 1060Ti, I think.
>>55508513
A cable. The cable isn't feeling anything.
>>55508542
Why is GDDR5 so shit, look at the fucking ungodly amount of pins it needs for a single module.
>>55508584
/g/ everybody
>>55508590
Ask AMD, they developed GDDR5, AMD is a shit memory developer, HBM is another FAIL
HBM2 is the real memory standard improvement
>>55508607
Try making an argument, chump. I know you're tech illiterate, but you could at least try to learn.
>>55508421
>one dvi port
Jimmies rustled
I have two dvi displays and if I would like to update my card I don't want to use an adaptor. Autism is too strong, when will this single dvi meme die?
>>55508896
when people get over their muh displayport fad.
>>55508896
If anything DVI ports are going to be gone completely in a few years. At least on AMD cards they are going poof real quick.