Friday, July 28th 2017

AMD Ryzen Threadripper Delidded - It's EPYC

Overclocker extraordinaire der8auer, who was one of the most vocal enthusiasts calling out for better VRM designs on Intel's X299 platform (and who worked with ASUS on redesigning the VRM cooling in its motherboards) has gone and done it: he delidded a Ryzen Threadripper CPU. And this delidding went on to deliver the goods: Ryzen Threadripper delidded is EPYC (pun intended.)

Instead of the expected MCM composed of two dies (with two CCXs of four cores per die, delivering the 16 cores we were expecting), Ryzen Threadripper is actually a much more interesting chip: it seems to be a full fledged EPYC chip, with four dies of eight cores. According to der8auer, when questioned, AMD confirmed that 16-core Threadripper 1950X CPUs are configured with two working eight core dies (four CCXs of four cores each), while the other eight-core dies are disabled by AMD.
A very, very interesting question here is whether these are actually defective or just disabled by AMD, which is something the company is naturally mum about. With AMD's history of core unlocking, and the fact that Zen-based CPUs have high yields, we can certainly hope (maybe even expect) that not all disabled cores are non-functional, and that AMD had to disable some of them so as to achieve their SKU core-count. Whether or not those will be unlockable, though, is anybody's guess. Der8auer also found that AMD Threadripper dies are gold plated on the inside, so as to improve conditions and adherence of the indium solder AMD used (some users might say that while Intel uses cheap TIM on their HEDT, X299 CPUs, AMD even sells these with gold inside.) This fact naturally also opens the question of future platform scalability - AMD can certainly decide to just up available maximum core-count on their Threadripper line of CPUs (we still have at least 4 model numbers above the 1950X, an potentially more.)
Check Der8auer's delidding video on YouTube below.

Source: Der8auer's YouTube Channel
Add your own comment

55 Comments on AMD Ryzen Threadripper Delidded - It's EPYC

#26
bug
chaosmassive, post: 3700515, member: 159641"
this is the reason why Intel so butthurt and go all out insulting Epyc CPU in their slide decks

think about it, one slice silicon to rule them all market segment from server, HEDT, performance, mainstream, all the way down to potato-grade CPU.
meanwhile Intel need to spend R&D and new design of silicons for each segmentation.
Intel's cores have scaled from ultrabooks to HEDT for years, nothing new for them in this area.
Posted on Reply
#27
bug
ShurikN, post: 3700524, member: 140585"
Thanks for the link.
There are some 10yo CPUs there and I have a E8400 at home still going strong, Buddy of mine was using a Q8400 up until a year ago without any issues (16h of gaming every day).
I just wanted to confirm for myself that solder has no issues in the long run. Because I'm more of a facts kinda guy than empty science.
I don't imagine that being a problem for mainstream CPUs. It's probably the HEDT class which are pushed more (i.e. run warmer) that may exhibit some symptoms after some years.
Posted on Reply
#28
chaosmassive
bug, post: 3700525, member: 157434"
Intel's cores have scaled from ultrabooks to HEDT for years, nothing new for them in this area.
my point is the physical silicon, AMD simply repurpose their silicon as they see fit
Intel have been using same uarch across all market segmentation, yes
but they need to create brand new silicon to be able address market demand
you dont just take Xeon 12 core cpu and disable it, sold as G4560
Posted on Reply
#29
iO
pcworld says those two additional dies is unprocessed sillicon used as spacers for an even pressure and pin contact.
So much for unlocking cores...:(
Posted on Reply
#30
r9
iO, post: 3700533, member: 107457"
pcworld says those two additional dies is unprocessed sillicon used as spacers for an even pressure and pin contact.
So much for unlocking cores...:(
And people say silicon is expensive.
Posted on Reply
#31
XiGMAKiD
R0H1T, post: 3700514, member: 131092"
Yes, here's a good list of CPU's to know if they were soldered or glued together with the IHS ~ https://www.overclock.net/t/305443/ihs-removals-how-to-do-it-should-i-do-it-and-the-facts

Ivy was the first mainstream CPU, in a long time, that wasn't soldered, SKLX being the first HEDT which went the Ivy way. Basically there's nothing out there that suggests that solder goes kaput first, as compared to TIM, having said that we have TR & SKLX which will prove this theory one way or another.
Well my Q9450 still going strong and relatively cool (65'C gaming temp) but with the stream of demanding new games I think this year is it's final year of services
Posted on Reply
#32
theoneandonlymrk
iO, post: 3700533, member: 107457"
pcworld says those two additional dies is unprocessed sillicon used as spacers for an even pressure and pin contact.
So much for unlocking cores...:(
They would say that though would they not, pure bs imho , they're binned prior to packaging.
They would use a shim over that or dead chips.
Posted on Reply
#33
bug
chaosmassive, post: 3700529, member: 159641"
my point is the physical silicon, AMD simply repurpose their silicon as they see fit
Intel have been using same uarch across all market segmentation, yes
but they need to create brand new silicon to be able address market demand
you dont just take Xeon 12 core cpu and disable it, sold as G4560
Silicon is paid for by the sq mm, so repurposing doesn't work like you think it does.
Posted on Reply
#34
OSdevr
r9, post: 3700534, member: 56898"
And people say silicon is expensive.
Relatively speaking silicon isn't that expensive, search on Ebay and you can find low quality silicon wafers for $20 American. Processing silicon into microchips is also not horribly expensive either. However, designing and making the mask sets takes many millions. If the company isn't fabless then they must also acquire the equipment to produce the chips.

The NRE cost of chips is the big expense.
Posted on Reply
#35
ShurikN
XiGMAKiD, post: 3700548, member: 161561"
Well my Q9450 still going strong and relatively cool (65'C gaming temp) but with the stream of demanding new games I think this year is it's final year of services
This is another one of my points. It's more likely your CPU will become obsolete than the solder becoming defective.
Posted on Reply
#36
Pap1er
cadaveca, post: 3700440, member: 25138"
The problem with solder is slowly the indium will get sucked onto the large patch of gold (used because of how indium bonds to it), and this will create holes in the solder that then lead to hot spots. This is why I like Intel using paste-based TIM... if it dries out, you can de-lid and replace, but when that solder wicks out, and the CPU runs hot only, you're forced to buy new gear.

This behavior of indium-based solders is properly documented, too, so while some see it as a boon, I see it a weakness in AMD's products, because, you know, I like science, not hype.
:) How many Intel users will de-lid their CPU? Your argument is pointless
Posted on Reply
#37
bug
Pap1er, post: 3700599, member: 86238"
:) How many Intel users will de-lid their CPU? Your argument is pointless
Not many, probably, but the bulk of those that do are HEDT users.
Posted on Reply
#38
Pap1er
bug, post: 3700602, member: 157434"
Not many, probably, but the bulk of those that do are HEDT users.
Honestly do you believe that HEDT user will pay over 1000$ for their CPU which is about to be used to make daily bread and then he'll decide to de-lid it? I am not perfectly convinced by this idea :)
Posted on Reply
#39
bug
Pap1er, post: 3700606, member: 86238"
Honestly do you believe that HEDT user will pay over 1000$ for their CPU which is about to be used to make daily bread and then he'll decide to de-lid it? I am not perfectly convinced by this idea :)
So do you think enthusiasts buy Celerons and try to run them at 5GHz? Atoms maybe?
Posted on Reply
#40
Pap1er
bug, post: 3700607, member: 157434"
So do you think enthusiasts buy Celerons and try to run them at 5GHz? Atoms maybe?
Not at all, I'm just wondering why would anyone even consider deliding their CPU, that's all.
Posted on Reply
#41
deu
cadaveca, post: 3700440, member: 25138"
The problem with solder is slowly the indium will get sucked onto the large patch of gold (used because of how indium bonds to it), and this will create holes in the solder that then lead to hot spots. This is why I like Intel using paste-based TIM... if it dries out, you can de-lid and replace, but when that solder wicks out, and the CPU runs hot only, you're forced to buy new gear.

This behavior of indium-based solders is properly documented, too, so while some see it as a boon, I see it a weakness in AMD's products, because, you know, I like science, not hype.
Or delid it like they just did? :D
Posted on Reply
#42
dozenfury
Pap1er, post: 3700608, member: 86238"
Not at all, I'm just wondering why would anyone even consider deliding their CPU, that's all.
I wouldn't do it with the silly vice/hammer methods, but with the delidding tools out there for $25 or so you can easily and very safely delid. And on the Intel side it definitely makes a big difference to replace it. That's what I did.

On the Threadripper I agree I wouldn't be too concerned about aging. Since it's got quality solder you don't have to worry about delidding (which shouldn't be necessary if Intel would use decent compound themselves), and it will be obsolete before needing replaced. If I get 4 or 5 overclocked years out of a CPU I'm happy and relegate it to live out it's days in a secondary PC.
Posted on Reply
#43
Zerofool
Is it only me or the IHS is made of copper? The last image shows four holes which they drilled for some reason, and the color of the material resembles copper. If so, this is yet another thing done right (along with the soldering). Man, if it turns out the extra cores are unlockable... Intel is in huge trouble :) although I doubt it, as it would cannibalize their EPYC sales.
Posted on Reply
#44
bug
Pap1er, post: 3700608, member: 86238"
Not at all, I'm just wondering why would anyone even consider deliding their CPU, that's all.
Well, if you're serious about overclocking, you're going to delid and make sure what's under there makes the best heat transfer possible.
You'd probably never do it and I certainly won't, but apparently there's enough people that do it out there, that this exists: https://rockitcool.myshopify.com/
Posted on Reply
#45
T4C Fantasy
CPU & GPU DB Maintainer
cadaveca, post: 3700440, member: 25138"
The problem with solder is slowly the indium will get sucked onto the large patch of gold (used because of how indium bonds to it), and this will create holes in the solder that then lead to hot spots. This is why I like Intel using paste-based TIM... if it dries out, you can de-lid and replace, but when that solder wicks out, and the CPU runs hot only, you're forced to buy new gear.

This behavior of indium-based solders is properly documented, too, so while some see it as a boon, I see it a weakness in AMD's products, because, you know, I like science, not hype.
i still have a working 2600k, it may happen but by the time it does you will replace the threadripper anyways, so invalid argument/fact to me anyways.
Posted on Reply
#46
cadaveca
My name is Dave
T4C Fantasy, post: 3700652, member: 105373"
i still have a working 2600k, it may happen but by the time it does you will replace the threadripper anyways, so invalid argument/fact to me anyways.
I have an AMD 4400+ that is pasted and still works fine too. I mean, we can say the same things about both options. Personally, I prefer paste.

Pap1er, post: 3700608, member: 86238"
Not at all, I'm just wondering why would anyone even consider deliding their CPU, that's all.
I feel the way you do, honestly, but we do have quite a few members here that would say "why wouldn't you, the paste is inferior!"

Anyway, it's awesome to see under the heatspreader anyway, so while I may not approve of de8hauer at all times, I still gotta say thanks. :p

Boosnie, post: 3700482, member: 170766"
I wonder how many cpu in the world perform a complete thermal cycle in theyr lives.
Every single one of mine, and there are a few members on here that do the LN2 stuff as well as air/water and all that too. Guys like macci and K1ngp1n drew me deep into hardware, and that led to me doing reviews. I just don't post anything about my LN2 adventures, because I'm not a competitive person. It's more about seeing how these CPUs work, and since I do tend to get quite a few free CPUs, I'm not going to have any losses from doing so. I mean, a decade ago there were many many of us buying $300 CPUs/GPUs every two weeks with our pay cheques and clocking the crap out of them. It was loads of fun. Then HWBot came, people went into teams, and that whole world became quite closed and elitist.
Posted on Reply
#47
Basard
bug, post: 3700607, member: 157434"
So do you think enthusiasts buy Celerons and try to run them at 5GHz? Atoms maybe?
They used to. You don't have to be rich to be an enthusiast. Remember the old 300mhz Celeron? No L2 cache... you could OC to 500Mhz easy. Now they are 10x the Ghz and 10x the cores.... holy shit!

Edit: Now that I think of it, it was either Celeron or Pentium back then, or K6 series.... And I remember the prices on even the Celerons being too high--for me anyways.
Posted on Reply
#49
TheLostSwede
Sir Alex Ice, post: 3700916, member: 73456"
Video is private.
AMD made him take it down.
Posted on Reply
#50
OSdevr
Zerofool, post: 3700626, member: 57662"
Is it only me or the IHS is made of copper? The last image shows four holes which they drilled for some reason, and the color of the material resembles copper. If so, this is yet another thing done right (along with the soldering). Man, if it turns out the extra cores are unlockable... Intel is in huge trouble :) although I doubt it, as it would cannibalize their EPYC sales.
I'm pretty sure both Intel and AMD have made the IHS out of nickel plated copper since they started using them.
Posted on Reply
Add your own comment