• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

90c+ CPUs

Unlikely

@tabascosauz talked about IF degradation which related to DRAM OC and IF v/f.
Has nothing to do with false CPU telemetry to force it exceed stock package power.
Fair enough. But what if it was using so much power because the board unnecessarily overvolted the CPU and IF as @Zach_01 mentioned? That overvoltage could have led to degradation maybe?

Just trying to learn how things work. :ohwell:
 
Last edited:
Fair enough. But what if it was using so much power because of the board unnecessarily overvolting the CPU and IF as @Zach_01 mentioned?
Board cannot overvolt or overclock IF by it self. Its different from Core v/f.
Boards are using false current(A) telemetry data to tell the CPU that it draws less power than it does and the CPU asks VRMs for more. But that is on Core VRM rails only.
SoC VRMs are separate
 
I remember my 3DFX Voodoo 3 2000 had a dinky passive heat sink on it, and if you touched it after a gaming session it would burn your finger. No clue what the temps were since GPUs back then didn’t report them, but it was scalding.
 
I remember my 3DFX Voodoo 3 2000 had a dinky passive heat sink on it, and if you touched it after a gaming session it would burn your finger. No clue what the temps were since GPUs back then didn’t report them, but it was scalding.
Fingers are poor thermometers though - even 60°C can give serious burns on skin.
 
It takes quite a lot of effort to scratch through a nickel plating. It's not tin foil.
You're right, it's not like tin foil, it's way thinner than that.

The maximum thickness of electroless nickel plating is limited to approximately 0.1 mm.
 
You're right, it's not like tin foil, it's way thinner than that.
But also far more tightly bonded, and a much, much stronger metal than aluminium. So, scratching it off takes effort. Sure, a single gouge with a sharp piece of steel will do it - but that gouge must then have some force behind it. Light abrasion won't take it off. Heck, you can still see the laser etched markings on them - that really isn't a lot of wear. Nickel is also harder than copper, making it quite wear resistant even in thin coatings, especially when contacting cold plates made from either nickel plated copper, copper, or aluminium.

Also, according to this, standard duty aluminium foil in the US is between 0.0004-0.0007" thick, or ~0.01016-0.01778mm thick. That's an order of magnitude thinner than 0.1mm.
 
But also far more tightly bonded, and a much, much stronger metal than aluminium. So, scratching it off takes effort. Sure, a single gouge with a sharp piece of steel will do it - but that gouge must then have some force behind it. Light abrasion won't take it off. Heck, you can still see the laser etched markings on them - that really isn't a lot of wear. Nickel is also harder than copper, making it quite wear resistant even in thin coatings, especially when contacting cold plates made from either nickel plated copper, copper, or aluminium.

I've circled a torn off piece of the Athlon's IHS, you can clearly see that there is no copper under there.
 

Attachments

  • 20220928_113asd321.jpg
    20220928_113asd321.jpg
    2.2 MB · Views: 85
I've circled a torn off piece of the Athlon's IHS, you can clearly see that there is no copper under there.
Torn off? That's dented. How on earth would you tear a chunk out of an IHS like that?
 
Was it PRD 76% along with 88W PPT? Because you do know PRD value alone doesn't mean anything.
I've seen some CPUs on screenshots here at TPU that had a awfully low PRD but also their PPT value was too low at the same time during all core loads/tests.

I mean if you had 88W PPT, 76% PRD this is:
88 / 0.76 = 115~116W

But if PPT was also lower that 88W (lets say 75) then true wattage is something else
75 / 0.76 = 98~99W

This was Zen2, it consistently maxed out PPT in all-core benches, 87-88W usually

Usually 76.1-76.5%, other boards were around the 95% mark with the TUF at 101-103%
 
This was Zen2, it consistently maxed out PPT in all-core benches, 87-88W usually

Usually 76.1-76.5%, other boards were around the 95% mark with the TUF at 101-103%
Wow, that's pretty bad. Was B450 worse than B550 for this? Both my ASRock B550 boards never leave 100% under load.
 
Wow, that's pretty bad. Was B450 worse than B550 for this? Both my ASRock B550 boards never leave 100% under load.

They definitely wisened up a bit after HWInfo had the feature. I can't recall any specific B550 or X570 boards that bad, but I can remember quite a few B450 and earlier boards being that bad.

I didn't have the opportunity to get past AGESA 1100D I think, to see whether Gigabyte eventually patched it. But I don't think any of my boards have changed their power reporting deviation over time
 
This was Zen2, it consistently maxed out PPT in all-core benches, 87-88W usually

Usually 76.1-76.5%, other boards were around the 95% mark with the TUF at 101-103%
True...
Same board I have now worked the R5 3600 at 87~88W PPT and PRD 91~92% And that was across all BIOSs even the latests (test it up to F35 v1.2.0.5)

Now the 5900X is what ever PPT I have set it and PRD is varying between 101~110% depending the load.
Usually around 102~104%.
 
Last edited:
Torn off? That's dented. How on earth would you tear a chunk out of an IHS like that?
Those dents are from the pair of vice grips used to rip the plate off the chip. Not literally, but a great handle for a delid. I heat the plate with a soldering torch for removal.

No, it makes it possible to respond to individual statements so that things are at least moderately readable, unlike this wall of spaced-out sentences you've posted here.

Okay, I had a long answer here, but reading to the end of your post has made it abundantly clear to me that we're not discussing the same thing. You've been working off of radically different assumptions for what is being discussed here, and arguing against my statements about a given platform with statements that assume every single variable is up for discussion. It isn't. Yes, there is a relation between die size, IHS size, and IHS thickness. For a given platform, IHS size is fixed, as it's determined by the socket and package dimensions (with maybe a mm or two of leeway for tweaking the design in X and Y directions. So: the area of the IHS for any given platform is fixed, and not up for discussion - at least unless it's brought up at the beginning of the damn discussion. Which you didn't. You responded to me saying the IHS was too thick by arguing that no, a thicker IHS would be better. Which ... no. Just no.

The question I've been trying to discuss with you this whole time has been this: given the size of AM5 heatsinks, assuming a cooler designed for the platform, would it be better with a thinner IHS? This is what I have argued for. Neither more nor less. You've come here dragging in all kinds of absurd examples, which are entirely irrelevant to this discussion. If you have been trying to argue this whole time that the fundamental characteristics of the platform are wrong, then why didn't you say so? It is so utterly and completely frustrating to try have a discussion with someone who doesn't even try to clarify the parameters of what they're discussing. This is also why you've been constantly contradicting yourself in all kinds of ways: because you've been responding to my arguments that presuppose that all other factors are pre-given with statements that assume that everything is in flux, making your logic fundamentally inapplicable to the question at hand.

To sum up: for any given surface area, a thinner IHS is better, assuming it has sufficient thickness to be mechanically stable and isn't so thin as to entirely hinder spreading heat outwards (which generally overlaps with mechanical stability - an IHS so thin as to stop heat spreading outwards sufficiently would also bend under pressure). Which I have come to believe that you agree with.

Whether the AM5 socket should have had a larger IHS is an entirely different discussion.


Why on earth are you linking garbage rumor mill site WCCFtech instead of Der8auer's video for this? You prefer to use a secondary source notorious for terrible analysis rather than the original source? Cool.

Jesus, have you read anything that has been posted in all the various 7000-series threads? CPU core temperature does not equal heat exhausted into your room. Power draw equals heat exhausted into your room. The absolute temperature of your CPU is not directly tied to the amount of heat energy put out by the system - if it was, then you wouldn't be able to lower temperatures with better cooling on any CPU ever.

The absolute temperature of the CPU is a function of power consumption and how quickly the cooling setup (TIM, IHS, TIM, cooler, ambient air temp) is able to move heat away from the core.

The heat output into ambient air is a function of the power consumption of the core. Period. If your CPU is consuming 230W consistently, then 230W of heat energy is being dumped into your room, regardless if the CPU cores are 30°C or 95°C.


That's what Eco mode is for. IMO, that should have been the default setting for these CPUs, but sadly the competitive situation makes things different right now.

It takes quite a lot of effort to scratch through a nickel plating. It's not tin foil.
Yessir, I am completely aware we are not talking about the same thing. lol. I knew that when I told you a second time my reason to quote you in the first place.

Simply can't take Der Bauers word on the plate being to thick, or too small or too large without testing with other plates, It's simply an opinion made statement and nothing more.

My point is that I've tested various sized plates and don't think thinner would suffice given the nature of current gen chips and the very heavily packed transistor area which the die sizes are getting smaller and smaller. This makes less surface area, a thicker or perhaps just larger in general might actually work better.

No, you don't need to use an IHS plate to increase how tall a CPU sticks out from a board. They could force cooler manufacturers to accommodate this minor issue, which isn't AMD's issue, but the end users. I've not read any statement from AMD concerning the design. Again, these are mostly assumptions.

Unless someone can slap a statement here (From AMD themselves) as to WHY the IHS plate is thicker, then it's all fake news. After a quick google search, there's just a bunch of guessing going on there.....

-----

To back up your statements to the other gentleman, (I underlined it)

CPU doesn't consume hardly any power a all. Super inefficient really. Most of the energy is dissipated as a heat, not being used.

An engine for example on average is about 30% efficient. All that fuel used and the rest dissipated by heat and cooled 3 ways. Water, Air and Oil.
 
Why has no one referenced the fact the the base clock on the 7950X is about the All core Boost clock of the 5950X and the single core is 600-800 MHZ higher depending on the board in both scenarios. It is obvious the 7700x is just like the 5800x in terms of thermals. As much as the cooler compatibility seems to be an issue we have to keep in mind that most AM4 coolers also supported 11th and 12th Gen CPUs so they are already designed with the intention of keeping 200+ Watt CPUs cool.
 
Why has no one referenced the fact the the base clock on the 7950X is about the All core Boost clock of the 5950X and the single core is 600-800 MHZ higher depending on the board in both scenarios.
Advertised clock speeds are never to be taken seriously, imo. Base clock is just a number (f**k knows what it really means nowadays), and boost clock is the clock speed you see in light workloads if your temperature, power consumption, motherboard VRM capabilities and the alignment of the planets in the solar system are OK. The clocks you really see are somewhere in between the two.
 
Advertised clock speeds are never to be taken seriously, imo. Base clock is just a number (f**k knows what it really means nowadays), and boost clock is the clock speed you see in light workloads if your temperature, power consumption, motherboard VRM capabilities and the alignment of the planets in the solar system are OK. The clocks you really see are somewhere in between the two.
I thought base clock was the guaranteed clock that you get all-core at the rated TDP. You can get way more, but that’s what the manufacturer guarantees under their recommended minimum setup.
 
I thought base clock was the guaranteed clock that you get all-core at the rated TDP. You can get way more, but that’s what the manufacturer guarantees under their recommended minimum setup.
That may be true for Intel, but AMD has to give a proper definition of "TDP" first (I know their stupid formula - I just don't believe that it has any connection with reality).
 
That may be true for Intel, but AMD has to give a proper definition of "TDP" first (I know their stupid formula - I just don't believe that it has any connection with reality).
And you say Intel has one? TDP is just a number now.
 
And you say Intel has one? TDP is just a number now.

Intel TDP is always defined as the max load power at base clock. At boost this goes out the window, but that's what their Turbo power metric is for, to better approximate realistic max power. Intel also renamed TDP to "Base Power", so now you get a number for base clock and a number for boost all-core.

AMD's TDP is utterly meaningless to power consumption. PPT is more or less the equivalent to Turbo power (barring obvious big differences in how Turbo vs. PB works), but AMD still advertises mostly "TDP" as if it meant something useful, and ignores PPT in most of their advertising materials. For some CPUs they get a number close to TDP at base clock all-core, but it's not defined or intended in that way.

There are more references to it now in 7000 slides but I've only really seen it in reference to AM5's design power (230W), the weird power efficiency slide about the "Ryzen 7970X", and the EDC and TDC changes (from 88/60/90, 142/95/140 to 88/75/150, 142/110/170 and 230/160/225).

Could argue it either way I guess, blue didn't even have a measurement for actual power for years, and red keep advertising this absolutely pointless number.

the 7970X slide:

7970x.png
 
Last edited:
I thought base clock was the guaranteed clock that you get all-core at the rated TDP. You can get way more, but that’s what the manufacturer guarantees under their recommended minimum setup.
That is the Intel definition of TDP up to 11th. On 12 gen they specify a PL1 and PL2 TDP.

AMD’s is way more complicated than this and has no real value to the user as it’s formulated in a way to tell something to cooler manufacturers mostly …only?

Advertised clock speeds are never to be taken seriously, imo. Base clock is just a number (f**k knows what it really means nowadays), and boost clock is the clock speed you see in light workloads if your temperature, power consumption, motherboard VRM capabilities and the alignment of the planets in the solar system are OK. The clocks you really see are somewhere in between the two.
At least for AMD if you disable PB, performance boost, the upper clock is the base clock on any situation (ST or MT).
Also with PB on and under minimum cooling conditions (TDP formula) you will get base clock at least on any given all core workload, even on the heaviest one.

So because 99.99% of users are not meting these situations the base clock is a speed that a CPU will not stop and has no meaning. But it’s not a random number.
 
Mate my 5950x hits 80c tops with PBO/curve tweaked, 4.4 all core with a 4.95 single, as the chip is intended. Your cooler is just not enough for that chip if it can't cool 130w.
If you don't mind me asking what settings are you using for your PBO setup? I'm playing around with PBO on my 5950x and looking for ideas that will likely work well.
In limited limited testing I've had time for PPT=170, TDC=130, and EDC=136 will get me CB23=27814 (better than my stock CB23=25067) under 70c and that is without any curves applied.
 
If you don't mind me asking what settings are you using for your PBO setup? I'm playing around with PBO on my 5950x and looking for ideas that will likely work well.
In limited limited testing I've had time for PPT=170, TDC=130, and EDC=136 will get me CB23=27814 (better than my stock CB23=25067) under 70c and that is without any curves applied.
The truth is that the higher the SKU the smaller the room for negative curve shifting. Almost all or most 5600X can do -30 steps on all 6 cores, depends on boost override too.
My CPU (5900X) on couple of cores can't do past -5 and I have the rest so far at -9 up to -19. Also currently the boost override is +75MHz, PPT 148, TDC 101, EDC 138.
Peak temp for all core boost is under 70C but on every other stuff wants to go to 80C or even more. I have a limit of 76C so cant pass that.

Avg clock on a couple of cores on games is 4.7~4.85GHz depending the type of game with a peak of 4.95~5.0+GHz
 
Last edited:
If you don't mind me asking what settings are you using for your PBO setup? I'm playing around with PBO on my 5950x and looking for ideas that will likely work well.
In limited limited testing I've had time for PPT=170, TDC=130, and EDC=136 will get me CB23=27814 (better than my stock CB23=25067) under 70c and that is without any curves applied.
I just did -10 curve on all core and left everything else auto. I hit around 28.4k in R23

Oh and +200 on the uh, override? Thing. 4.4ghz all core and 4950mhz single.
 
Back
Top