da pins

you have to be pretty spastic to bend a pin in the first place.

>>61027347
Some of the pins are redundant. So it might not always work.

Threadripper is LGA

Also if you're not being an idiot, it's unlikely you'll bend them

Don't even think about it. Unless you have a severe disability your CPU pins won't suffer from damage.

>>61027347
>(Intel fanbooy : ( ),
>rig
stopped reading. reddit leave

>>61027347
I've unbent all the pins on a stepped on P4. Took like an hour, but the proc worked, and no pins broke.

File: linus.jpg (20KB, 400x400px) Image search: [Google]

20KB, 400x400px

>>61027347
>Welcome to a very brief and unexpected

I'm pretty sure most people knew exactly what they were expecting when they saw him show up.

The pins aren't fragile. They require a little bit of force to bend. They aren't like copper wires that just bend and break. Think of them more like the solid fins on a radiator or a heat sink. Could you bend the fin easily? Sure. Will you bend it easily? Not unless you're a gorilla and force everything in to place.

Take her easy, but take her.

The only way you'll bend the pins is if you're a moron, and the only way you can't bend them back is if you're a crackhead.

I don't even know about breaking the pins off entirely. Guess that's some sort of special Linus technique born out of gross incompetence.

>>61027347
I went full retard with my old P4 rig and ended up bending some pins.
Cursed at myself and then sat down with a pair of tweezers and bent all the pins back in place, plopped the CPU back into the socket and it booted just fine.

As long as you have patience and don't go full gorilla on it the CPU won't give a shit.

>>61027347
How do you bend a pin, just drop the CPU in don't force it.

If it doesn't just plop in then you done it wrong faggot.

I find LGA pins to be much more fragile and harder to work with than normal CPU pins.

>>61027347
>But How often does this work?
There are a lot of vcc and gnd pins so losing 1 of them doesn't make any changes. Most of the it'll work but you can still break a data pin and render the thing unfunctional.

>>61027347
>But How often does this work?
I heard account of sch case too, it came up a year or so ago. Myself I straightened pins successfully twice (no failures either), both cases were Pentium 4s in socket 478. The socket had bad design that didn't allow you to twist the cooler before taking it off, so it would pull out CPUs.

The installation is safe (as lang as you aren't idiot or drop the CPU on the floor - that leads to some bends but to me it was straightenable). You only have risk of bending when removing the cooler. You have to remember to twist the cooler a bit around after you loosen the screws/clips. When you feel it has gave way, you can pull it off - ideally by nudging it to the side, not straight up. Done like that, the CPU stays in the socket safely.

>>61027732
Indeed, and the problem is that they don't face straight up, so it is complicated to lift them to proper position, even if you are nimble enough with tweezers. You probably need to have watchmaker agility.

File: 61530256_p5.jpg (137KB, 683x800px) Image search: [Google]

137KB, 683x800px

>>61028883
With the pin inserted in the socket and touching, there can actually be contact, so it is not just a matter of dummy pins being expendable.

File: IMG_20170622_204509.jpg (499KB, 2520x1760px) Image search: [Google]

499KB, 2520x1760px

i fixed my cpu before

pic related

>>61029449
Wow you converted it to LGA.

>>61027347
https://youtu.be/61-_UfVXLTc?t=6m10s
Here is another method.

>>61029769
Fuck, his CPU has more RPMs than mine. brb trying this out.

>>61027347
Take 2 credit cards bend pin. Take your time.

>>61027347
I bought an 1151 motherboard from eBay a while back and the jerkoff seller shipped it some way that fucked some of the pins in the socket. Lost like two of them completely.
>Still worked

A lot of the pins are just a ground, this is not really unique to AMD

>>61027367
>>61027388
>>61027392

First off this. Its not like you take the cpu in and out all the time. Just dont fuck it up.


>>61027347
Also isnt there always pins either on the cpu or in the socket?

>>61027388
Wrong

Threadripper is DOA

File: intel ceo JUST.jpg (119KB, 1000x1000px) Image search: [Google]

119KB, 1000x1000px

>>61030936
Go away, Brian.

>>61027347
Holy fuck that actually worked

>>61029827
>>61027347
I like the micropencil way. It is probably less stressing for the pins, less risk they could break at the base.

https://www.youtube.com/watch?v=5bjPSH41FRQ

Why did AMD go with pins on the CPU vs on the motherboard? I'm genuinely honest. Is it for cost savings, or is it for performance gains, or is it Intel vs AMD proprietary stuff?

>>61032264
It is less damage prone and also cheaper to produce on the motherboard side, more or less same on the CPU side, so the whole thing is cheaper.

Threadripper and Epyc use LGA despite being more expensive and finicky, because they have way too much pins (4094) to use PGA. Basically there was no choice, it's not like LGA is inherently better.

>>61027347
wow.
>bump computer
>instantly shuts off and wont turn back on
You would think that they would atleast try to solder it

>>61032727
It's Linus. What did you expect?

>>61027347
cuck tips now have a new name?:

>>61032727
I still dont like pins on cpu, its a whole harder to land the cpu, not that its dificult, but it increase the possibility of bending. On LGA you just land the cpu and its done

>>61032766
Linus started as a sales rep at NCIX.

Relevant: https://www.youtube.com/watch?v=vzKcrwlmREE

My fear is not the PGA installing, but the disassemble. What If I want to reapply thermal paste to the cpu and the cpu sticks to the cooler while I am taking it off. I don't understand why don't they secure the cpu more so It would never happened like with Intel sockets.

>>61033230
give a little twist before removing the cooler, works for me every time

>>61033230

Thermal paste should come apart with almost zero effort.

>>61030936
that's x299 though

File: C5kA5gZXQAAzl_d.jpg large.jpg (57KB, 750x735px) Image search: [Google]

57KB, 750x735px

>>61033230
>>61033563
Exactly - always twist first, you can feel if it still holds or not well.

>>61027347
I mean, it might work if you do this but if this happens to you you should definitely be sweating buckets and shitting bricks.

>>61027434
This was 2009, before he was famous for breaking shit

>>61034458
He was always breaking shit, just back then he wasn't famous for it.