AMD Zen CPU & AM4 Socket Pictured – PGA Design With 1331 Pins Confirmed

[csbin]

http://wccftech.com/amd-am4-socket-zen-bristol-bridge-soc-package-pictured/


AM4 Socket

AMD Zen CPU

 

[Doom2pro]

Watch AM4+ remove three notches for 1337 pins

 

[Justinbaileyman]

Awesome!! Thanks for posting and jeeze hope not Doom...

 

[nerp]

Still pins. Ughhhh.

 

[SpaceBeer]

What's wrong with that?

 

[Shehriazad]

I always liked pins....seeing all that gold makes me feel special. :]

 

[ShintaiDK]

What's wrong with that?

Its inferior plain simple. AMD also uses LGA in servers as a big hint. Its most likely due to motherboard makers pressuring AMD to avoid taking on the problems, should there be any. So you bend/break pins=AMD problems, not motherboard makers.

 

[stuff_me_good]

I always liked pins. If you bend few, easy to bend back, but if you bend LGA pins, it's game over.

 

[Hi-Fi Man]

Reminds me of socket 754 with the space in the middle.

I never liked when Intel switched from socket 478 to LGA 775 because bending CPU pins back into place is far easier then LGA pins. Bare die mounting is also much easier with PGA.

 

[Zor Prime]

Pins for the win.

 

[crashtech]

It's not an inferior design by any means. LGA just shifts the complexity and fragility from the CPU to the motherboard. For consumer products this can be a liability, since many Windows licenses are tied to the motherboard, but in many situations LGA makes sense because the CPU is the more expensive part. I don't mind ZIF pins at all, as pointed out they are generally easy to repair, MUCH easier than a bent LGA pin.

 

[The Stilt]

There is nothing wrong with the pins IMO. Despite they are even thinner now on AM4, they are still much more rigid than any of the Intel LGA sockets. They are also symmetric unlike the LGA contact arms, which means that it is much easier to repair them if you manage to bend them. PGA looks kind of low-tech nowdays, but who cares when it does what it is supposed to do.

 

[VirtualLarry]

Its inferior plain simple. AMD also uses LGA in servers as a big hint. Its most likely due to motherboard makers pressuring AMD to avoid taking on the problems, should there be any. So you bend/break pins=AMD problems, not motherboard makers.

Is it "inferior", because AMD uses it and Intel does not?

Personally, I think Intel adopted LGA as a cost-cutting measure, given their volume of CPU production, and nothing more. AMD doesn't ship nearly the volume as Intel, and likely, wouldn't save much money doing so.

 

[Dresdenboy]

I'm bored by reading the 517th copy of the pin vs. LGA discussion. And usually I see more convincing points to have pins. So why discuss it again?

 

[Doom2pro]

I'm bored by reading the 517th copy of the pin vs. LGA discussion. And usually I see more convincing points to have pins. So why discuss it again?

Simply because there hasn't been a discussion that stokes as much as this now because nobody has released a picture of AM4 or the CPU package until now even though Bristol Ridge and AM4 have been released for over a week! This is the only place that has reported it in a forum (Nobody in SemiAccurate has at least until recently).

Whatever, AMD is getting done what it needs to get done in AM4 be it Bristol Ridge or Summit Ridge or Whatever Ridge, who cares?

AM4 works.

 

[scannall]

I'm bored by reading the 517th copy of the pin vs. LGA discussion. And usually I see more convincing points to have pins. So why discuss it again?

Eh, it's a good way to filter out people and their ability to fix things. If they rage about pins on a CPU, don't hire them to do anything remotely delicate. After I don't know how many CPU's and other assorted chips, I don't think I've bent more than a couple of pins. Just pay attention, and don't be drunk or high, and the pins will be fine. ;-)

 

[Dresdenboy]

Simply because there hasn't been a discussion that stokes as much as this now because nobody has released a picture of AM4 or the CPU package until now even though Bristol Ridge and AM4 have been released for over a week! This is the only place that has reported it in a forum (Nobody in SemiAccurate has at least until recently).

The µPGA leak was in March (supported by a Zauba entry soon afterwards), with a photographic confirmation in June. Each time: "Why no LGA? AMD is doomed." (on Twitter, Yahoo Finance, Reddit, Stock Twits, AT, SA, TPU, WCCF, you name it) OK, looking back, I think that at least this herd behaviour is funny.

Maybe I'm just following that stuff too closely.

Eh, it's a good way to filter out people and their ability to fix things. If they rage about pins on a CPU, don't hire them to do anything remotely delicate. After I don't know how many CPU's and other assorted chips, I don't think I've bent more than a couple of pins. Just pay attention, and don't be drunk or high, and the pins will be fine. ;-)

Most people also won't apply accurate statistical methods to their observations, but just use the episodic story mode, distributing any pin related experience as it fits. Somehow this also remembers me of guys trying to impress others with stupid things, ending in broken bones or even necks. That's today's variant of natural selection (seen the Darwin Awards?).

 

[superstition]

Speaking of pins... when lapping does getting oil on the corner pins result in a CPU requiring higher voltages? Maybe it's just my imagination but it seemed like after lapping my 8370E I needed more CPU NB core voltage than prior to it. I cleaned with 99.9% alcohol but Q-Tips are full of dust.

I guess wearing thin gloves is a good idea for lapping.

 

[DrMrLordX]

Oil? Are you talking skin oil or . . . ?

 

[LTC8K6]

I'm not against pins per se, it's just that these seem like they are going to be much more delicate than before. They have put a lot more pins in the same amount of space. Some articles say this is the most pins ever for PGA.

 

[scannall]

I'm not against pins per se, it's just that these seem like they are going to be much more delicate than before. They have put a lot more pins in the same amount of space. Some articles say this is the most pins ever for PGA.

Intel x58 would disagree. 1366 pins there.

 

[LTC8K6]

Intel x58 would disagree. 1366 pins there.

Well, that was LGA, so not sure what that has to do with it?

 

[Dresdenboy]

I'm not against pins per se, it's just that these seem like they are going to be much more delicate than before. They have put a lot more pins in the same amount of space. Some articles say this is the most pins ever for PGA.

Of course. The pin density per dimension is roughly ~20% higher. If hit with a hammer, there shouldn't be a noticeable delta in the result between a 1331 pin µPGA and one with 906/~940 pins. Falling onto a desk might bend the pins more (density+diameter), while they might be more flexible and the guy dropping it might be somewhat more careful in awe of their sheer number. How are we going to quantify advantages and disadvantages vs. LGA?

 

[LTC8K6]

Of course. The pin density per dimension is roughly ~20% higher. If hit with a hammer, there shouldn't be a noticeable delta in the result between a 1331 pin µPGA and one with 906/~940 pins. Falling onto a desk might bend the pins more (density+diameter), while they might be more flexible and the guy dropping it might be somewhat more careful in awe of their sheer number. How are we going to quantify advantages and disadvantages vs. LGA?

I build AMD systems all the time, even though I am an Intel fan. AMD FX has been my go to budget system for friends and family. So when I say I am concerned about 1,331 tiny little pins in the same space as before, it merely means that I am actually concerned about it. It does not mean whatever your post was trying to imply. Not everyone is like you think they are.

 

[Arachnotronic]

I like LGA chips because I can stuff one in my pocket, forget the chip is in there, run it through the wash, and still have a working chip at the end