• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Why do most desktop x64 CPUs have lids these days?

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
dullard said:
The problem is people who start and stop high powered turbo use over years of time.
Click to expand...

Unless you are talking about cycling that can occur from booting the machine, there's plenty of temp swing on many of the computers I have used over the years, including the ones with soldered-on IHSes. For example, my current Ryzen rig has idle temps of ~28-30C Tdie and ~48-52C Tctl (this is X370 so keep that in mind). If I hit it with something with y-cruncher - something I've done many times - it can swing as high as 67C Tdie and 87C Tctl on a short 500M run. That's running a "constant" speed of 4 GHz (to the extent that Ryzen CPUs ever really run a constant speed).

Also the x2 I have running in a mining rig is a pull from an office computer, sold on eBay. I do not think it's ever been overclocked.

Now, fact is, Intel DOES permit more temp swing just from having chips that can run as high as 90C without shutting themselves down. The Ryzen temp limits are a bit nebulous (especially on X370), but everything prior had an almost hard limit of 72C.
 
DrMrLordX said:
Unless you are talking about cycling that can occur from booting the machine, there's plenty of temp swing on many of the computers I have used over the years, including the ones with soldered-on IHSes. For example, my current Ryzen rig has idle temps of ~28-30C Tdie and ~48-52C Tctl (this is X370 so keep that in mind). If I hit it with something with y-cruncher - something I've done many times - it can swing as high as 67C Tdie and 87C Tctl on a short 500M run. That's running a "constant" speed of 4 GHz (to the extent that Ryzen CPUs ever really run a constant speed).

Also the x2 I have running in a mining rig is a pull from an office computer, sold on eBay. I do not think it's ever been overclocked.

Now, fact is, Intel DOES permit more temp swing just from having chips that can run as high as 90C without shutting themselves down. The Ryzen temp limits are a bit nebulous (especially on X370), but everything prior had an almost hard limit of 72C.
Click to expand...

I looked up the temperature differences for some popular chips :
Intel :

I7 6700 Tcase (max temp for IHS) = 71C.
I7 7700K Tjunction(No Tcase mentioning at the ark site) = 100C.
I5 8600K Tjunction(No Tcase mentioning at the ark site) = 100C.
I7 8700 Tjunction(No Tcase mentioning at the ark site) = 100C.
All have a TIM material that is not soldered.
E7 - 8855 V4. Tcase (max temp for IHS) = 73C.
This one i do not know if it is soldered or not.

AMD:
This is a bit confusion because for the ryzen chips a max temp is mentioned but not what that temperature is.
It only says : Max temps.
I assume this is the maximum allowed junction temperature.
1600 = 95C
1700 = 95C
1700X = 95C
1800X = 95C

The chips do not differ much in max temperature.

Perhaps because the ryzen 1xxx on the 14nmLPP process was pushed to the limits, AMD had no choice to use the best thermal conductive TIM available and that is why they went for solder.
The ryzen+vega apu do not use soldered TIM. It is cheaper to manufacture.
It still makes me wonder if the PGA package is more rigid and warps less because of the pins in comparison to LGA.
Will ryzen 2xxx also use soldered TIM ?
 
chrisjames61 said:
The APUs are considered "budget" parts. They used TIM to keep the price point down.
Click to expand...
I expect its also to prevent cracking with hot spots. As far as I know with Ryzen it can probably keep both CCX and the uncore evenly cooled. It's possible with the split in cores/gcn modules, that one of the two units would be vastly hotter than the other, playing a game the cpu's clock down and the GPU clocks up, while compiling the GPU takes a break and the CPU's go full bore. It's these dramatic shifts that lead to the solder microfracturing potential issue (which on a standard CPU is more of an LN2 issue).
 
Can someone provide good reference reading material for this "soldered" lid? I have never seen a soldered lid and it makes no sense to me. Is it possible people are referring to a metal loaded thermal adhesive as solder?
 
mindless1 said:
Can someone provide good reference reading material for this "soldered" lid? I have never seen a soldered lid and it makes no sense to me. Is it possible people are referring to a metal loaded thermal adhesive as solder?
Click to expand...

Well, here is a quote: "It seems that AMD is using two small pads of Indium solder along with some gold plating on the inside of the IHS to facilitate heat transfer and allow the solder to mate with the IHS. Because AMD is using what seems to be high quality solder TIM"

Source: https://www.pcper.com/news/Processo...onfirms-AMD-Using-Solder-IHS-Ryzen-Processors
 
^ Thanks. Still skeptical that this should be called solder but I suppose it doesn't matter what it's called.
 
I guess I get what you are saying. But it is a metal layer heated to the point of bonding with two different otherwise disattached components that solidfies after being cooled. If that isn't soldering than what is soldering?
 
^ I would like to read about this metal layer heated to the point of bonding. So far all I have read is that someone tried heating one to remove it, which is not the same thing. For example I can heat paint to remove it but that doesn't mean that's how it was applied, including back in the day when paint had lead in it. That wouldn't make that paint solder.

Back in the day when AMD and Intel were using a metal-less (silicone?) junction, that too came loose by heating it.
 
^ I do believe that now, but the links were not evidence, just someone saying it was. For more info anyone can research Reactive Nanofoil, which does involve solder but it's a bit more than just that.
 
mindless1 said:
^ I would like to read about this metal layer heated to the point of bonding. So far all I have read is that someone tried heating one to remove it, which is not the same thing. For example I can heat paint to remove it but that doesn't mean that's how it was applied, including back in the day when paint had lead in it. That wouldn't make that paint solder.

Back in the day when AMD and Intel were using a metal-less (silicone?) junction, that too came loose by heating it.
Click to expand...
The trick to bonding whether it is solder or something else to the die of the CPU is that you have to avoid heating the CPU to much that it might destroy the CPU. The advantage of Indium is that it has a realitively low melting point (~160C) and can wet itself to just about anything including glass. So all AMD or Intel has to do is press the lid down while having the press heat the lid to over 160c. The Indium wets itself to the the CPU and then the heat is removed and the the Indium soldifies.

The link I gave you specifically covered the process of bonding the Indium to the die. That should have alieviated any question on whether it was "soldered" or not.
 
I think, as stated already, physical protection and maximising contact area with the aftermarket heatsinks installed by newbs (vs trusted die to lid connection) are the two largest reasons.

But the lids also act as a heatsink themselves in emergency situations. I think this is an important safeguard, giving even a few more fractions of a second for on board sensors to detect an overtemp and allow the system to throttle down and avoid damage. The days without any thermal protection are long gone, so not many here may have the experience of watching a CPU smoke in front of your eyes, but I can vouch it happened very quickly without a heatsink installed. Plus even basic on-die thermal sensors may miss hot spots and allow CPU damage if it happens too fast. So a lid, with multiple times the surface area, can act to spread and slow down (to an extent) thermal changes even without a heatsink. This is obviously less of an issue as thermal protection advances and multiple on-die temp sensors but it still comes into play.
 
  • Like
Reactions: OTG
Modern bga packages have a metal frame around the die to prevent uneven pressure on the edges. I think it would be near impossible to damage the die by installing a heatsink.

Cooling would be better an manufacturing cheaper. Sounds like a win-win situation to me.
 
Topweasel said:
The trick to bonding whether it is solder or something else to the die of the CPU is that you have to avoid heating the CPU to much that it might destroy the CPU. The advantage of Indium is that it has a realitively low melting point (~160C) and can wet itself to just about anything including glass. So all AMD or Intel has to do is press the lid down while having the press heat the lid to over 160c. The Indium wets itself to the the CPU and then the heat is removed and the the Indium soldifies.

The link I gave you specifically covered the process of bonding the Indium to the die. That should have alieviated any question on whether it was "soldered" or not.
Click to expand...

Except that's not how it's done. Search reactive nanofoil .
 
Last edited:
mindless1 said:
Except that's not how it's done. Search reactive nanofoil.
Click to expand...
Well I imagined they did it the easy way since Indium has such a low melting temp. But I don't see how the usuage of a nano foil to super heat the Indium pad and only the Indium pad, makes it not Soldering.
 
DrMrLordX said:
Bollocks, I've got a soldered Phenom downstairs and a soldered Athlon II x2 upstairs, both in working condition. Years from now, you will probably have soldered Ryzen chips plugging away without complication, and I would suspect the same of Haswell-E.
Click to expand...

Not to be rude, but this is anecdotal, whereas dullard's post is based upon scientific evidence.
 
Topweasel said:
Well I imagined they did it the easy way since Indium has such a low melting temp. But I don't see how the usuage of a nano foil to super heat the Indium pad and only the Indium pad, makes it not Soldering.
Click to expand...
and as I wrote, "I do believe that now", that in some aspects it could be considered soldered for lack of a better term lay people will be familiar with for this manufacturing method, but that it's not "soldered on" in the conventional sense. You can't just buy that solder paste, apply heat to the lid, and DIY.
 
Last edited:
Phynaz said:
Not to be rude, but this is anecdotal, whereas dullard's post is based upon scientific evidence.
Click to expand...
Ridiculous.

Extrapolating from the paper quoted and saying that the reason overclocks result in unstable chips years down the road is due to solder, is the antithesis of scientific evidence. It might be a contributing factor, but for certain, insufficient data was presented. He made a sweeping generalization on a single viewpoint, as in "I feel this is the reason, because.........".

You appear to be supportive of his post as solder has become a differentiator between manufacturers and you need to establish a defense. Intel good, AMD bad.
 
mindless1 said:
and as I wrote, "I do believe that now", that in some aspects it could be considered soldered for lack of a better term lay people will be familiar with for this manufacturing method, but that it's not "soldered on" in the conventional sense. You can't just buy that solder paste, apply heat to the lid, and DIY.
Click to expand...

There are exotic welding techniques that are used in all sorts of situations (like underwater), and you can't just pop down to the hardware store and grab supplies.
They still fuse two pieces of metal together, so they're still welding.

Soldering using nanofoil to melt the solder, is soldering, even if I can't do it at home to enjoy any toxic fumes that the factory is selfishly keeping to themselves.
It's even in the article:
"Temperatures at the joining surface are raised above the temperature necessary to melt the solder or braze, and thus create a bond."
 
^ I asked for info and received bs. Why did I have to ask and then come back as the source of the info? Seems like egos matter more than science.
 
Last edited:
^ All you have done is explain by your own personal feeling the process for melting the solder. I don't believe that you have provided any proof that it is actually the case. But he and myself gave you that you could and probably are correct. The point is how does that isn't that soldering? While you concede that it might be, it seems to be under protest. Whereas many people including myself post links to exactly what material and why it was chosen for the Soldering. Your theory on how the metal is melted doesn't change that and doesn't make you the source.

If anyone is having Ego problems it is you.
 
Ironic that by asking for info I ended up having to research and provide it myself and yet still there is resistance based on ego.
 
Last edited:
mindless1 said:
Ironic that by asking for info I ended up having to research and provide it myself and yet still there is resistance based on ego.
Click to expand...
It is a marketing term that has stuck. The name of the product is "Solder Thermal Interface Material". http://www.indium.com/thermal-interface-materials/solder-tim/ The whole industry and the entire enthusiast group calls it solder for short. It is like you complaining that someone calls a tissue a Kleenex, or asks for an Aspirin, or says to Google something, or has a Videotape of a show, or goes to a Laundromat.
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top