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

#101
Mussels
Moderprator
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:
Oh its you again, why do you post the same troll on every AMD news post?
Posted on Reply
#102
theGryphon
I wonder if there can be some innovation with the LGA implementation, in place of the leaf springs. For example, why not tiny bead heads that slot into tiny holes on the CPU side? It would be extremely safe.
Posted on Reply
#103
Arc1t3ct
Both sockets work just fine. I've never had a problem with either...

LGAs "look" cooler though...
Posted on Reply
#104
Lord_Soth
JismThe PR rating was based on AMD's own CPU's, not Intel's counterpart. If you'd buy a 1800+ or so it was guaranteed to be equal like a 1800Mhz chip.

Netburst was horrible. The A64 beat it to it while being clocked lower.
Yes, it was based on the AMD Athlon Thunderbird chip, not the Pentium 4.
AthlonXp IPC>Thunderbird IPC>Pentium III IPC>Pentium 4 IPC
Posted on Reply
#105
RealKGB
theGryphonI wonder if there can be some innovation with the LGA implementation, in place of the leaf springs. For example, why not tiny bead heads that slot into tiny holes on the CPU side? It would be extremely safe.
Kind of like a reverse PGA socket?
That could work. It will also help prevent idiots from trying to put a 12900K in an X670-E.
dicktracyZen3 and its premium prices should’ve rocked LGA socket.
They couldn't do that. They had promised that Zen 3 support AM4 and you can't change a socket midway through its life.
I also like PGA. The motherboards are hard to screw up and if you bend a pin or two just bend it back!
With Intel's LGA style the CPUs are practically invincible, but if you accidentally bend a pin in there, or drop something, etc., that socket's as good as dead.
Posted on Reply
#106
Panther_Seraphin
People miss one of the big reasons as to why Intel went to LGA

Its pretty hard to bend a pin/start a warranty claim on something that cannot be bent.

Also people who say LGA is superior cause you can get more pins into the same space. Well you realise you have to dedicate more of those pins to power planes due to the thinner contact area meaning if you tried to push the same amount of current through an LGA pad as what is possible through a PGA pin you end up with pretty torched sockets.
Posted on Reply
#107
Isaac`
Panther_SeraphinIts pretty hard to bend a pin/start a warranty claim on something that cannot be bent.
so just shuffle all of that over to the motherboard manufactures
that's big brain
Posted on Reply
#108
TechLurker
What are the odds that AMD adopts their EPYC/TR installation and secure mounting system to Socket AM5? It's pretty hard to screw that one up, and it's generally been praised in early TR reviews.
Posted on Reply
#109
Super XP
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:
Quite the Contrary, AMD has been a leader in Innovation & Technological Advancements for over 30+ years. :toast:
_FlareI never bend any socket or CPU pin. How can anybody be so untalented?
Maybe there are people out there poking nose and eyes eveytime washing their face,
or cutting themselves everytime they use a knife.
Maybe some people do not have the ability to get in the state of being calm and just don´t break anything.
I believe the main culprit is when people try to seat the CPU into the socket the wrong way despite the clear arrow indicators telling you how to plop the CPU into the socket with ease.
I've never had a problem with pins bending. The only issue I had once is when I purchased a used AMD Athlon CPU way back and 2 pins weren't exactly perfectly straight. Yes the eBay seller lied and claimed no bent pins, unless it was a shipping situation that caused them to bend. So I gently pushed the pins back until they were straight. Then it went in like warm butter, umm sort of.
MusselsOh its you again, why do you post the same troll on every AMD news post?
Insecurity perhaps? These companies are major multi billion dollar entities, not sure why @voltage has to troll AMD, but even lie about it too in his posts? :nutkick:
Posted on Reply
#110
theGryphon
RealKGBKind of like a reverse PGA socket?
That could work. It will also help prevent idiots from trying to put a 12900K in an X670-E.
Kinda but not exactly... I was thinking more about tiny bumps, or pin heads, but not pins. Something not bendable. Smooth semi-circular tiny bumps instead of pinheads is better. You could even wipe-clean it. I mean seriously, why does something so expensive need to be so fragile? Both LGA and PGA sucks in this. A little TIM goop and it's done for :banghead: I don't need to see the innards of the CPU socket. Just give me the contact interface.

Here is an analogy...
What we have:
+

What we could have:
Posted on Reply
#111
Super XP
theGryphonKinda but not exactly... I was thinking more about tiny bumps, or pin heads, but not pins. Something not bendable. Smooth semi-circular tiny bumps instead of pinheads is better. You could even wipe-clean it. I mean seriously, why does something so expensive need to be so fragile? Both LGA and PGA sucks in this. A little TIM goop and it's done for :banghead: I don't need to see the innards of the CPU socket. Just give me the contact interface.

Here is an analogy...
What we have:
+

What we could have:
I hate those old red-black-black-yellow power connections. Always struggled to get them plugged in LOL.
I am surprised nobody came out with something like that metal housing around it 10-15+ years ago.
Posted on Reply
#112
RealKGB
theGryphonKinda but not exactly... I was thinking more about tiny bumps, or pin heads, but not pins. Something not bendable. Smooth semi-circular tiny bumps instead of pinheads is better.
I wonder if a CGA, or contact-grid array socket could be feasible?
Kind of like if you took the bottom of an LGA CPU and made that both the socket and CPU contact. It would be sensitive to dust though...
Or maybe we could have something similar to how the Compute Module 4 (Raspberry Pi) connects to its daugherboard. CPU is on a PCB that hooks onto 2 slots similar to PCIe slots.
Posted on Reply
#113
NuCore
newtekie1LGA pins are LGA pins. At least they are less likely to get damaged than pins on the CPU. I'm so glad AMD is finally moving to LGA for their mainstream desktop processors.
Well, I wouldn't say. Both solutions can be damaged (pins in the processor and pins in the processor socket). Most often this happens when someone who has no idea about building computers starts to do it.
Posted on Reply
#114
Caring1
theGryphonI wonder if there can be some innovation with the LGA implementation, in place of the leaf springs. For example, why not tiny bead heads that slot into tiny holes on the CPU side? It would be extremely safe.
Like BGA without the soldering.
Posted on Reply
#115
theGryphon
Caring1Like BGA without the soldering.
Thank you, yes, exactly. Micro-BGA without soldering. CPU side would have micro (even smaller than the bumps) holes, so they would line up correctly.

They already have it: eesemi.com/vfbga.htm

As small as 0.4mm pitch. 432 balls in 11x11 mm2 area. Which would give you 1718 balls easily in a 22x22 area. :rockout:
Posted on Reply
#116
Tomorrow
TechLurkerWhat are the odds that AMD adopts their EPYC/TR installation and secure mounting system to Socket AM5? It's pretty hard to screw that one up, and it's generally been praised in early TR reviews.
Haha no. Even people like Der8auer have accidentally dropped Threadrippers on the socket pins and LGA (especially the more pins there are) has a major issue with memory detection etc. So it may take a few retriers to be reseated for everything to work correctly.

As for why more pins? well thats easy to answer. Both Intel and AMD will increase core counts in mainstream sockets. Hence the need to add more pins. Intel moves from 8c/16t to 8c/24t (likely even higher in the future) and AMD will likely go even further with 8 or 12 core CCD. If each chiplet contains two CCD's like it does now we will see 32c/64t models (4x8) with two chiplets or if CCD=Chiplet then we will see 24c/48t models (2x12).
Posted on Reply
#117
Bones
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:
Talk about an Introll strollin and trollin here.... :laugh::roll::laugh:

FACT:
AMD was the first to introduce 64 bit capable chips and Intel coudn't for a long while, that is until they made a deal with.......
AMD to get the knowledge on how to do it.
So goes that BS theory out the window - AMD can and does innovate new technology.

I still remember wondering where all the 64 bit Intel chips were but there was exactly 0 around because they didn't exist with AMD having them listed all over the place and that went on for at least a full 6 months if not longer.

Of course Intel then ran off with the crown for awhile but since not only have they lost it but things have come out as fact along the way in what they did (Or didn't do) to get it - None of it to any benefit to their customers and that would include you.
Next time get your BS right so your post(s) won't turn into the latest running joke of the thread. :nutkick:
Posted on Reply
#118
Wirko
theGryphonThank you, yes, exactly. Micro-BGA without soldering. CPU side would have micro (even smaller than the bumps) holes, so they would line up correctly.

They already have it: eesemi.com/vfbga.htm

As small as 0.4mm pitch. 432 balls in 11x11 mm2 area. Which would give you 1718 balls easily in a 22x22 area. :rockout:
I think you're looking for spring-loaded pins, or "pogo pins". These, for example, are very, very thin. Here you can see how they are made.

Non-bendable, though ... well, that's a lot to ask.
Posted on Reply
#119
BackSlash
TheLostSwedeHoles? There are no holes where we're going...
Personally I'm hoping for coolers that attach to the socket with screws.
Yeah you're right. I mean... same Heatsink for Intel and AMD. less mess
Posted on Reply
#120
Panther_Seraphin
AMD have actually had a lot of "firsts" in the tech arean for quite a few years

64Bit? AMD
Multi Core Consumer Processors? AMD
First CPU to 1Ghz? AMD

I cant see something like BGA ever working in a consumer environment due to tolerances. There is already enough issues with LGA sockets and mounting pressures with effectively spring loaded contact points giving you far more tolerance. Take that away and I doubt any lay person would be able to get something like a TR working with a BGA arrangement correctly without a ritual sacrifice to the tech gods every time they went for a remount or God forbid a CPU cooler remount.

Pogo pins definately look interesting but I wonder how hard it would be to design the traces from all those pins out from the socket to the rest of the mobo would look.
Posted on Reply
#121
cyberloner
i just upgrade fx8350 to 5800x ... :/ new socket new ram now.......
Posted on Reply
#122
cheapcomputers
I hope they don't make the pins thinner. I've had my cpu stick to the heatsink plenty of times. I didn't know or think about heating the cpu first. Thanks for the tip. I've been building computers for 20 years. Luckily I've only bent the pins once. I dropped the cpu. I was able to straighten them and had no problems afterwards. My 2400G came with two bent pins. It wouldn't fit in the socket. I didn't want to return it so I straightened those pins and it's been fine for 3 years now. Buying a new cpu with bent pins did make me angry at the time but if I can get at least another few years out of it I'll be happy.
Posted on Reply
#123
candle_86
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:
AMD64
Onboard Memory Controller
Single Die multi core
Infinity Fabric
Posted on Reply
#124
The red spirit
repman244If you are somehow unable to not damage pins on CPU/MOBO leave it to a shop or someone qualified to do it for you then (that goes for untalented people as well).

I really don't get all the moaning about bent pads/pins on LGA or PGA...I've removed/installed CPUs countless times without damaging any of it.
I managed to damage socket pins by just mounting Scythe Choten cooler too tightly. I never had any problems with AMDs before. You could probably smack them with hammer and they will be fine (as long as they are in socket). Intel just had inferior mounting system with LGA1200. Also if CPU pins were bent then fixing that is easy, but Intel's motherboard pins just straight up fell out. I don't wan to see any other LGA socket in my life.
Posted on Reply
#125
Mussels
Moderprator
candle_86AMD64
Onboard Memory Controller
Single Die multi core
Infinity Fabric
he got upset people gave him facts and told him he was wrong, inboxed me some harassment and left the forum. He gone.
Posted on Reply
Add your own comment