[Rumor] ZEN to be "soldered"

[Prysin]

apparently, a Chinese overclocker delidded a ZEN sample and found it to be using Liquid metal TIM, or "soldered" TIM. This is just one indicator that overclocking will be supported.

Whether the chips will overclock well is anyones guess at this point. But if the IHS is "soldered", then atleast we will not see bullshit high temps like we did with original Haswell. It is a nice touch from AMDs side and goes to show that they might just care about delivering good products.

Quote

The overclocker Nam Dae Won has Delidded an AMD AM4 CPU, revealing that the CPU uses solder or liquid metal as its thermal interface material. 

 

Over the past few years deliding CPUs has been popular amongst overclockers, removing the IHS (integrated heat spreader) improve heat transfer between the CPU and your chosen cooling solution. This method has been particularly popular on Intel's LGA 115X series CPUs, which use thermal paste to connect their CPUs to the IHS instead of solder or other more thermally conductive materials. 

 

It looks like AMD's AM4 CPUs will be using a high-quality thermal interface material, which means that mainstream users will likely see no benefit from de-lidding their CPUs. 

 

 

 

SOLDERED OR NOT. THIS IS STILL NOT AN INDICATOR WETHER THE ZEN CPUs WILL OVERCLOCK WELL. IT SIMPLY MEANS THAT THE IHS AND TIM WILL NOT BE A ISSUE.

 

The downside to Liquid Metal, or "Solder" is that these forms of alloys or metals should not be heated beyond 60 degrees Celsius. As they will often become more corrosive the hotter they get.

 

Source: https://www.overclock3d.net/news/cpu_mainboard/overclocker_delids_amd_am4_cpu/1

[suicidalfranco]

you know what time it is?

time to hate on AMD

[ItsTheDuckAgain]

seriously: wayne?

[DELTAprime]

I'd rather see benchmarks than a de-lidding, just sayin'.

[Bouzoo]

The 3rd picture looks really nasty. Disgusting in fact. 

[Dash Lambda]

Well, at least you're not saying it'll be $1 and twice as fast as Skylake.

I do like the idea of a soldered TIM... Let's see if it turns up true.

[patrickjp93]

No one realizes that's a SandyBridge E5 Xeon delidded?

[Prysin]

4 minutes ago, patrickjp93 said:

No one realizes that's a SandyBridge E5 Xeon delidded?

last time i checked. XEONs didnt have AMD stamped onto the IHS.... AMD CPUs often does.

[HKZeroFive]

Looks to be a Bristol Ridge APU rather than a Zen-based CPU...

[Dash Lambda]

11 minutes ago, patrickjp93 said:

No one realizes that's a SandyBridge E5 Xeon delidded?

I'm not seeing it. Most of the SandyBridge Xeons I'm finding have different component layouts or different dimensions on the die, and it looks like that chip is sitting above the surface it's on (meaning it looks like it has pins).

[Orcblood]

14 minutes ago, patrickjp93 said:

No one realizes that's a SandyBridge E5 Xeon delidded?

If it looks like a xeon that's because of how big it is to fit those extra cores.

[samcool55]

39 minutes ago, patrickjp93 said:

No one realizes that's a SandyBridge E5 Xeon delidded?

Apart from the AMD logo in the IHS it does look like the CPU is a little bit lifted which would suggest there is something below the cpu.

Which would make sense because afaik AM4 will use a PGA socket so there would be pins on the cpu

[patrickjp93]

14 hours ago, Prysin said:

last time i checked. XEONs didnt have AMD stamped onto the IHS.... AMD CPUs often does.

Photoshopped

 

14 hours ago, Dash Lambda said:

I'm not seeing it. Most of the SandyBridge Xeons I'm finding have different component layouts or different dimensions on the die, and it looks like that chip is sitting above the surface it's on (meaning it looks like it has pins).

Are you out of your mind? E5 2840.

 

14 hours ago, samcool55 said:

Apart from the AMD logo in the IHS it does look like the CPU is a little bit lifted which would suggest there is something below the cpu.

Which would make sense because afaik AM4 will use a PGA socket so there would be pins on the cpu

Every Intel, AMD, and Nvidia silicon die sits on top of the package be it a CPU, GPU, APU, etc.. CPU dies are BGA on top of their package substrates.

[Morgan MLGman]

2 minutes ago, patrickjp93 said:

Photoshopped you ditz.

 

Are you out of your mind? E5 2840.

 

Every Intel, AMD, and Nvidia silicon die sits on top of the package be it a CPU, GPU, APU, etc.. CPU dies are BGA on top of their package substrates.

E5 2840? Wut is that? I work in the server-grade part of IT industry and have never seen nor heard about a CPU like that.

 

Also, I have never seen an Intel CPU that looks like this after being delidded. Looks genuine to me.

[79wjd]

6 minutes ago, patrickjp93 said:

Photoshopped you ditz.

 

Are you out of your mind? E5 2840.

 

Every Intel, AMD, and Nvidia silicon die sits on top of the package be it a CPU, GPU, APU, etc.. CPU dies are BGA on top of their package substrates.

Do you have any proof that it's an E5 2840? What even is a 2840?

 

That still doesn't explain the air gap under neath the CPU. Although, it's hard to tell how elevated it actually is, and it could just be elevated from the LGA pads. 

[Sakkura]

5 minutes ago, patrickjp93 said:

E5 2840.

Doesn't exist.

[patrickjp93]

4 minutes ago, Morgan MLGman said:

E5 2840? Wut is that? I work in the server-grade part of IT industry and have never seen nor heard about a CPU like that.

 

Also, I have never seen an Intel CPU that looks like this after being delidded. Looks genuine to me.

Westmere, Nehalem, and Sandy Bridge Xeons often came in two dies. You're not nearly as well informed as you believe.

 

3 minutes ago, djdwosk97 said:

Proof that it's an E5 2840. What even is a 2840

 

That still doesn't explain the air gap under neath the CPU. Although, it's hard to tell how elevated it actually is, and it could just be elevated from the LGA pads. 

A standard 10-core Xeon from Sandy Bridge coming in a 2-die package with 1 core disabled on each.

 

It's not like the die is PGA on top of the package substrate! The die is BGA on the package which is PGA on the socket.

[79wjd]

2 minutes ago, patrickjp93 said:

Westmere, Nehalem, and Sandy Bridge Xeons often came in two dies. You're not nearly as well informed as you believe.

 

A standard 10-core Xeon from Sandy Bridge coming in a 2-core dies with 1 core disabled on each.

 

It's not like the die is PGA on top of the package substrate! The die is BGA on the package which is PGA on the socket.

I can't find a single reference to an E5 2840 anywhere. Why don't you actually provide a link for once in your life.

 

I'm not talking about the elevation of the die itself. The entire CPU is elevated off the table. 

[Morgan MLGman]

1 minute ago, djdwosk97 said:

Do you have any proof that it's an E5 2840? What even is a 2840?

 

That still doesn't explain the air gap under neath the CPU. Although, it's hard to tell how elevated it actually is, and it could just be elevated from the LGA pads. 

 

1 minute ago, Sakkura said:

Doesn't exist.

It's AMD, got some confirmation, here's how Sandy/Ivy bridge CPUs look like:

 

And now that delidded Zen:

 

 

And now an AMD FX-8350:

 

See how much closer the IHS is to the borders of the CPU, a typical "AMD" style with pins underneath (PGA)

 

1 minute ago, patrickjp93 said:

Westmere, Nehalem, and Sandy Bridge Xeons often came in two dies. You're not nearly as well informed as you believe.

Apparently better informed than you as you called BS on a news related to AMD without posting any proof but your own word. Everything (when looking at the pictures) indicates that it's a genuine Zen and the Xeon model you listed does not exist. Sorry to burst the bubble.

[patrickjp93]

1 minute ago, djdwosk97 said:

I can't find a single reference to an E5 2840 anywhere. Why don't you actually provide a link for once in your life.

 

I'm not talking about the elevation of the die itself. The entire CPU is elevated off the table. 

 

1 minute ago, Morgan MLGman said:

 

It's AMD, got some confirmation, here's how Sandy/Ivy bridge CPUs look like:

 

And now that delidded Zen:

 

 

 

And now an AMD FX-8350:

 

 

See how much closer the IHS is to the borders of the CPU, a typical "AMD" style with pins underneath (PGA)

 

Apparently better informed than you as you called BS on a news related to AMD without posting any proof but your own word. Everything (when looking at the pictures) indicates that it's a genuine Zen and the Xeon model you listed does not exist. Sorry to burst the bubble.

Nope, light source is ~15* above the horizontal. It's sitting on top of the capacitors. Nothing PGA about it.

[79wjd]

2 minutes ago, patrickjp93 said:

 

Nope, light source is ~15* above the horizontal. It's sitting on top of the capacitors. Nothing PGA about it.

Maybe, maybe not. It could be elevated from LGA pads or PGA pins, it's hard to tell for certain with the picture. But even that aside, it looks a lot more like an AMD cpu than an Intel one. And again, I'm still waiting for proof that an E5 2840 even exists. Not to mention proof that a delidded 2840 looks like the leaked pictures. 

[patrickjp93]

Just now, djdwosk97 said:

Maybe, maybe not. It could be elevated from LGA pads or PGA pins, it's hard to tell for certain with the picture. But even that aside, it looks a lot more like an AMD cpu than an Intel one. And again, I'm still waiting for proof that a 2840 even exists.

Forgive me, 2830, but GFY. You know I'm right.

[Morgan MLGman]

Just now, patrickjp93 said:

 

Nope, light source is ~15* above the horizontal. It's sitting on top of the capacitors. Nothing PGA about it.

What about this?

A delidded FX-8320:

 

A delidded rumoured "Zen":

 

Notice the similarities. Not sure if it's Zen for 100% as that's not possible to confirm, but it's definitely AMD's work.

[79wjd]

3 minutes ago, patrickjp93 said:

Forgive me, 2830, but GFY. You know I'm right.

You mean an E7 2830? And it still doesn't look like the pictured CPU. 

[patrickjp93]

1 minute ago, Morgan MLGman said:

What about this?

A delidded FX-8320:

 

A delidded rumoured "Zen":

 

Notice the similarities. Not sure if it's Zen for 100% as that's not possible to confirm, but it's definitely AMD's work.

Similarities my right foot! Different lenses, different lighting angles, different surfaces, different distances to sensors!

 

The "Zen" picture is a photoshopped E7 2830 with a lightsource at 15* to the horizontal which any halfwit using basic knowledge of optics and refraction equations could tell you. The shadow is produced by the elevation of sitting on top of the capacitors on the bottom of the package.