Saturday, May 22nd 2021

AMD Socket AM5 an LGA of 1,718 Pins with DDR5 and PCIe Gen 4

A reliable source with AMD and NVIDIA leaks, ExecutableFix has shared some interesting bits of early information on AMD's next-generation Socket AM5. Apparently this will be AMD's first mainstream-desktop socket that does away with pins on the processor package, shifting them to the motherboard, in a Land Grid Array (LGA) format. This won't be AMD's first client LGA, though, as it was the Quad FX platform from 2006, which used a pair of Socket F LGAs. Socket AM5 will have a pin-count of 1,718 pins, 18 more than Intel's upcoming Socket LGA1700, on which its 12th Gen Core "Alder Lake-S" is expected to be based.

AMD will give the I/O of its client desktop platform a major update, with the introduction of DDR5 memory. Socket AM5 processors are expected to feature a dual-channel DDR5 memory interface. With Intel "Alder Lake-S" implementing DDR5, too, you now know why every major memory manufacturer is unveiling their first DDR5 U-DIMM product development. Interestingly, the PCI-Express interface on Socket AM5 will remain PCI-Express 4.0, even though PCI-Express 5.0 is being rumored for "Alder Lake-S." The switch to PCI-Express 5.0 may not be significant from a graphics cards perspective immediately, but paves the way for next-gen M.2 NVMe SSDs with double the transfer-rates of current drives that use PCI-Express 4.0. AMD is developing the new 600-series chipset to do with its next-generation Socket AM5 processors.
Source: ExecutableFix (Twitter)
Add your own comment

138 Comments on AMD Socket AM5 an LGA of 1,718 Pins with DDR5 and PCIe Gen 4

#77
R0H1T
"Rumored" i.e. take it with ocean full of salts.
Posted on Reply
#78
docnorth
I think PCI-e gen 5 means CXL protocol ready. The question is how important (or not) will this be after 2-4 years.

Posted on Reply
#79
TheinsanegamerN
Finally, AMD goes LGA. I'm fine with PGA, but AMD seems utterly unable to figure out how to make a retention bracket for PGA. Finally no more CPUs stuck to the bottom of heatsinks with bent pins.
Posted on Reply
#80
Sithaer
TheinsanegamerNFinally, AMD goes LGA. I'm fine with PGA, but AMD seems utterly unable to figure out how to make a retention bracket for PGA. Finally no more CPUs stuck to the bottom of heatsinks with bent pins.
Never happened to me and I mainly owned AMD systems in the past ~10+ years. 'only execption was a i 3 4160 for 2+ years'
Not a single bent pin or any CPU related damage, when I'm about to remove the cooler for whatever reason I just let some CPU stress run for 5-10 mins like Aida64 or similar and let it heat up before turning off the PC.

After that I give the cooler a gentle twist and it never got stuck nor damaged anything, heck I'm far from an expert and even I can manage this so idk why this is always brought up as an AMD issue or an issue in general.:confused:

To be honest I'm more concerned about Mobo socket pins than having them on the CPU but thats just me I guess.

This is coming from someone who ruined a 1TB NVMe M2 SSD with an aftermarket heatsink install+removal, yet I never had issues with CPU pins.:laugh:
Posted on Reply
#81
Unregistered
lol. maybe they'll actually have a good retention mechanism this time.

also the heatsink problem isn't an issue if you have half a brain and know to stress test CPU for 5 mins & then twist the cooler before taking it off...
Posted on Edit | Reply
#82
Haile Selassie
I'm pretty sure that someone mentioned this already, but the photo shows SP3 socket, not AM5.
Posted on Reply
#83
thesmokingman
Panther_Seraphinwww.techpowerup.com/281550/amd-zen-5-strix-point-processors-rumored-to-feature-big-little-core-design
lmao, use a dictionary to define what rumor means?
TheinsanegamerNFinally, AMD goes LGA. I'm fine with PGA, but AMD seems utterly unable to figure out how to make a retention bracket for PGA. Finally no more CPUs stuck to the bottom of heatsinks with bent pins.
They've been using LGA on Threadripper and Epyc from the beginning of those product stacks. And your issue stuck cpus is an end user problem with your paste hardening.
Posted on Reply
#84
TheinsanegamerN
SithaerNever happened to me and I mainly owned AMD systems in the past ~10+ years. 'only execption was a i 3 4160 for 2+ years'
Not a single bent pin or any CPU related damage, when I'm about to remove the cooler for whatever reason I just let some CPU stress run for 5-10 mins like Aida64 or similar and let it heat up before turning off the PC.

After that I give the cooler a gentle twist and it never got stuck nor damaged anything, heck I'm far from an expert and even I can manage this so idk why this is always brought up as an AMD issue or an issue in general.:confused:

To be honest I'm more concerned about Mobo socket pins than having them on the CPU but thats just me I guess.

This is coming from someone who ruined a 1TB NVMe M2 SSD with an aftermarket heatsink install+removal, yet I never had issues with CPU pins.:laugh:
Why does everyone think ABS is a good thing? I've never had my wheels lock up in the past 10+ years, not sure why everyone wants it on cars nowadays.

Driver issues? I've never had a driver issue in 10+ years, so IDK why this is always brought up as an AMD issue or issue in general.

Antivirus? I've never had an infection in 10+ years, so IDK why this is always brought up as an end user problem or a problem in general.

I could go on. You've never had an issue? Congratulations, doesnt mean a problem doesnt exist. A simple look at forums over the last 10+ years would show that a LOT of people have issues with AMD CPUs sticking to heatsinks and getting pulled out of motherboards, bending pins in the process. Multiple tech youtubers like Linux have put out videos on the subject over the years. It happens, it is an implementation issue with PGA sockets, used to happen on laptop CPUs too, which were all PGA and had a nasty habit of doing the exact same thing.
thesmokingmanAnd your issue stuck cpus is an end user problem with your paste hardening.
Hard paste is the cause, but there is a solution. Intel figured that out with the LGA retention bracket, they had the same issue on socket 370/423/478. Third party brackets for some coolers on socket AM4 have a retention bracket that fixes the issue. Whether due to lack of care, cost, or not wanting to break backwards compatibility with previous coolers, AMD's PGA sockets have never fixed this problem, while thanks to the retention bracket I've never had a modern intel desktop CPU stick to a heatsink no matter how hardened the paste was. I look forward to AM5 fixing this issue.
Posted on Reply
#85
Sithaer
TheinsanegamerNWhy does everyone think ABS is a good thing? I've never had my wheels lock up in the past 10+ years, not sure why everyone wants it on cars nowadays.

Driver issues? I've never had a driver issue in 10+ years, so IDK why this is always brought up as an AMD issue or issue in general.

Antivirus? I've never had an infection in 10+ years, so IDK why this is always brought up as an end user problem or a problem in general.

I could go on. You've never had an issue? Congratulations, doesnt mean a problem doesnt exist. A simple look at forums over the last 10+ years would show that a LOT of people have issues with AMD CPUs sticking to heatsinks and getting pulled out of motherboards, bending pins in the process. Multiple tech youtubers like Linux have put out videos on the subject over the years. It happens, it is an implementation issue with PGA sockets, used to happen on laptop CPUs too, which were all PGA and had a nasty habit of doing the exact same thing.

Hard paste is the cause, but there is a solution. Intel figured that out with the LGA retention bracket, they had the same issue on socket 370/423/478. Third party brackets for some coolers on socket AM4 have a retention bracket that fixes the issue. Whether due to lack of care, cost, or not wanting to break backwards compatibility with previous coolers, AMD's PGA sockets have never fixed this problem, while thanks to the retention bracket I've never had a modern intel desktop CPU stick to a heatsink no matter how hardened the paste was. I look forward to AM5 fixing this issue.
I never claimed that the issue doesn't exist, but you can avoid it if you bother to spend 5-10 mins before taking your system apart like I and someone else mentioned with stressing the CPU beforehand and also not brute force the damn thing.

Sure I can see this being a problem if you wan't to replace cooler/CPU in a system that wont boot up but at that point you have bigger issues to worry about anyway.

Edit:
Forgot to mention, you could to the same search about ppl who bent pins/ruined their Mobo socket pins, I can see such mobos on the second hand market every now and then.
Its all the same, both designs have its problems and at the end of the day its mainly user end error/fault.
Posted on Reply
#86
zlobby
R0H1TNeeds multiple reliable citations o_O
Well, at that point it's more or less expected from them, only that it will come to mainstream and portables. You can't really expect a generic workstation or server to have common demand for mix of cores.
Posted on Reply
#87
Totally
TheinsanegamerNWhy does everyone think ABS is a good thing? I've never had my wheels lock up in the past 10+ years, not sure why everyone wants it on cars nowadays.
So you are saying your ABS is working as in intended. Hasn't it been mandatory for over 30 years now, it's optional? people actually have a choice whether they want it or not?
Posted on Reply
#88
GoldenX
I don't understand the point of big.LITTLE. That space could be better used with more big cores.
This isn't Android where the OS is shit and you need 4 cores constantly working on its background tasks, not even Windows is that bloated.
Posted on Reply
#89
sollord
voltageAMD = After other Manufacturers Do it. When will they be first to create their own new technology, Never!

They cant even come up with their own product naming scheme. :roll:
Kinda like Intel copying LGA from MIPS and HP?
Posted on Reply
#90
Isaac`
voltageAMD = After other Manufacturers Do it. When will they be first to create their own new technology, Never!

They cant even come up with their own product naming scheme. :roll:
Ok so your using a 64bit cpu right
that's amd seems your computer would not actually exist right now if amd did not (:
Posted on Reply
#91
windwhirl
Haile SelassieI'm pretty sure that someone mentioned this already, but the photo shows SP3 socket, not AM5.
It's mostly to illustrate LGA-style CPUs.
Posted on Reply
#92
Bones
TheinsanegamerNWhy does everyone think ABS is a good thing? I've never had my wheels lock up in the past 10+ years, not sure why everyone wants it on cars nowadays.

Driver issues? I've never had a driver issue in 10+ years, so IDK why this is always brought up as an AMD issue or issue in general.

Antivirus? I've never had an infection in 10+ years, so IDK why this is always brought up as an end user problem or a problem in general.

1: I could go on. You've never had an issue? Congratulations, doesnt mean a problem doesnt exist.
A simple look at forums over the last 10+ years would show that a LOT of people have issues with AMD CPUs sticking to heatsinks and getting pulled out of motherboards, bending pins in the process. Multiple tech youtubers like Linux have put out videos on the subject over the years. It happens, it is an implementation issue with PGA sockets, used to happen on laptop CPUs too, which were all PGA and had a nasty habit of doing the exact same thing.

2: Hard paste is the cause, but there is a solution.
Intel figured that out with the LGA retention bracket, they had the same issue on socket 370/423/478. Third party brackets for some coolers on socket AM4 have a retention bracket that fixes the issue. Whether due to lack of care, cost, or not wanting to break backwards compatibility with previous coolers, AMD's PGA sockets have never fixed this problem, while thanks to the retention bracket I've never had a modern intel desktop CPU stick to a heatsink no matter how hardened the paste was.
3: I look forward to AM5 fixing this issue.
I'll pick out a couple of things here to elaborate on.

1:
YES It doesn't mean there isn't a problem, no denying that but since I own both and have so for years now I've had way less trouble from AMD's current design (PGA).
The thing about cooler and CPUs coming out together I've ran into myself many times over and I can tell you, most of those damaging CPU pins are STILL twisting the cooler as they remove it. The suggestion is always to twist the cooler and many take that literally as what do to before AND while trying to separate them from each other or just to remove the chip from the socket.

And by doing that what do you think IS going to happen to the pins doing that as it comes out of the socket?
The correct way is simply to pull straightup and out of the socket, no bending of pins that way. By doing that you won't damage any pins on the chip and it doesn't hurt the board either.
Remove it now, separate them after it's out of the board.

2:
Thicker TIMs along with the stock "Gum"/TIM are also contributors, thinner TIMs aren't nearly as bad about causing it.
I always remove the factory gum that comes on them and replace it with a good, thinner type TIM such as MX4 or NT-H1.

3:
LGA is probrably going to make things worse overall.
The older sockets such as 775 are good and far more robust than the later 1151 socket for example.
The problem is the fitting all those pins in the socket, a socket of the same basic dimensions as what's been used for years so to pack these extra pins in you gotta make them thinner, which also means weaker and they are more prone to bend/break because of that.

If they change the physical dimensions of the socket allowing for thicker pins there would be far less trouble but ATM I don't see that happening, in fact as the numerical pin count go up in these sockets it's gonna get worse.

There is yet another thing that many don't seem to know about or have mentioned much if at all, that would be the pins themselves shifting around in the socket, because for some reason they get/are loose where they are fixed into the socket.
They don't get bent, they just move around to one side.
All it takes is for a single pin to shift to one side "Just enough" and it will start throwing codes, making one think the board is defective or they just broke a pin. I went through exactly that with my Z77 OCF when it was new, took forever to figure it out but when I did I fixed it and it works BUT it also still tends to do this.

Not everytime but annoying as hell to deal with it, that sometimes happens to my IX Apex board too and had it do just that the other day.

I know each socket type has advantages but from what I've dealt with up until now PGA is still better.
Posted on Reply
#93
Minus Infinity
trparkyWhat's "IPA" stand for?
Indian Pale Ale, always best to have a beer before working on the computer.
Posted on Reply
#95
BackSlash
Let's hope... Holes on the Motherboard matches with Intel's MB, I mean. Coolers should be compatibles each others.
Will be good for Makers, them and us.
I hope.....
Posted on Reply
#96
theGryphon
Minus InfinityYou mean how socket AM4 managed to go unchanged when moving from PCI-E 3.0 to 4.0. Gees already worrying about non-issues when it is Intel that drops a new socket every 2 years.
Most people don't realize what AMD managed to do with keeping AM4 all this time, especially for 5000 series, is a true engineering feat.

If it was INTEL as usual, we'd be on the third socket now since 1000 series, easily. For INTEL, boards and chipsets are part of the business plan. So, it's easy decision for them: invest R&D to keep the previous generation, or invest in marketing to sell the new generation?
Posted on Reply
#97
Caring1
mechtech
Yoda is wrong, the future is static, the present moves forward.
Posted on Reply
#98
Ferrum Master
So much BS in comments.

I've lived on nForce 680i SLI Quadfather setup for a while years ago. AMD has used LGA CPU's since ages. The desktop segment was the only exception, not vice versa.
Posted on Reply
#99
TheLostSwede
BackSlashLet's hope... Holes on the Motherboard matches with Intel's MB, I mean. Coolers should be compatibles each others.
Will be good for Makers, them and us.
I hope.....
Holes? There are no holes where we're going...
Personally I'm hoping for coolers that attach to the socket with screws.
Posted on Reply
#100
dicktracy
PGA has always been crap. It only made sense with Zen 2 and below when AMD was slow and selling at bargain bin prices. Zen3 and its premium prices should’ve rocked LGA socket.
Posted on Reply
Add your own comment