Despite Low-Power Cores, Intel's Alder Lake Mobile Could Push Up to 115W TDP
And we thought 14nm was power-hungry
Respected Linux patch detective Coelacanth’s Dream has penned a new blog post decoding the latest Intel patches, giving us some insight into unreleased information about the upcoming Alder Lake-P laptop processors. The information indicates some of Intel’s Alder Lake CPUs could have a configurable TDP as high as 115W, making Alder Lake the most power-hungry Intel chips ever produced for the laptop market.
The post reveals three core configurations for Alder Lake-P, consisting of a 2+8+2, 4+8+2, and a 6+8+2 configuration. This is a result of Alder Lake using bigger and smaller cores in its architecture, meaning 'big' high-performance cores, and 'small' Atom cores for better efficiency. The first number belongs to the bigger (performance) cores, while the second number is the smaller (power saver) cores. The last number is the integrated graphics.
Each core configuration scales with higher TDPs; the chip with two high-performance cores operates at a max of 55W for the PL2 turbo rating. The quad-core operates at up to 64W, and the hexacore operates at a peak of up to 115W.
While this number may sound insane for a notebook, even for some of the largest notebooks on the market, TDP is a very different animal than it used to be. Intel allows OEMs to adjust the TDPs of its chips significantly, so the CPU is geared more toward the notebook’s design, rather than static TDPs which were once popular.
Intel has also introduced two levels of Turbo Boost behavior to help optimize efficiency in both notebook and desktop form factors. These levels are known as PL1 and PL2 states. PL2 is the most aggressive turbo clock available and is designed to go way beyond the CPU's base power consumption for a short amount of time. This is why Intel's mobile chips in general have seen higher TDP numbers over the past few years, making them look more power-hungry than they actually are.
Even for notebooks that are equipped with a high core count Alder Lake chip that boosts up to 115W, it's doubtful it will boost that high for any long duration of time. And don’t expect many notebooks to be configured with that high of a TDP.
Stay On the Cutting Edge: Get the Tom's Hardware Newsletter
Get Tom's Hardware's best news and in-depth reviews, straight to your inbox.
Aaron Klotz is a contributing writer for Tom’s Hardware, covering news related to computer hardware such as CPUs, and graphics cards.
-
peachpuff If mobile is going 100+watts then desktop should easily reach 300watts, that 10nm sure is paying off... <Borat>Not! </Borat>Reply -
jgraham11 It's even worse than this article makes it out to be. Intel's current 11th gen laptop parts are no where even close to Ryzen for power efficiency. Look at the base clocks for comparison:Reply
Intel 11800h @ 45w runs at base of 2.3GHz vs Amd's 5800h base clock is 3.2GHz at the same 45watts. That's almost 1GHz faster, using the same power.
Intel has no choice but to boots their cores to go as fast as possible just to compete with Amd. This is why the boost clocks are so high and so extreme.
This is finally coming to the surface with Intel's desktop chips on B560 motherboards. Most can't run the high power chips because they can only put out approximately 100watts for a 65watt chip, because their boost behaviour draws more than 125watt for a 65 watt chip. They can't run any of Intel's high power chips within Specifications. And they throttle hard. -
JayNor The Alder Lake mobile SKUs also includeReply
M5, 5W-9W
U9, 9w-15W
... so, what is the problem? -
spongiemaster PL2 for Tiger Lake H is 109W for 8 cores. 115W for a 6+8 configuration isn't that much more. The performance will determine whether it is worth it or not. If it brings a new level of performance to laptops, ehh .. whatever. If it needs that much power just to compete with AMD, that's a problem. Either way, I doubt there are that many people in the market for that many cores in a laptop in the first place.Reply -
Why_Me
Most entry level B560 boards can't run Intel's unlocked cpu's without throttling down due to poor VRM's. Those boards were made for locked cpu's. I've still yet to meet anyone who's purchased a $90 - $100 B560 board and paired it up with a $560 i9 11900K.jgraham11 said:It's even worse than this article makes it out to be. Intel's current 11th gen laptop parts are no where even close to Ryzen for power efficiency. Look at the base clocks for comparison:
Intel 11800h @ 45w runs at base of 2.3GHz vs Amd's 5800h base clock is 3.2GHz at the same 45watts. That's almost 1GHz faster, using the same power.
Intel has no choice but to boots their cores to go as fast as possible just to compete with Amd. This is why the boost clocks are so high and so extreme.
This is finally coming to the surface with Intel's desktop chips on B560 motherboards. Most can't run the high power chips because they can only put out approximately 100watts for a 65watt chip, because their boost behaviour draws more than 125watt for a 65 watt chip. They can't run any of Intel's high power chips within Specifications. And they throttle hard. -
Phaaze88
Because it's more likely that the cost trimming was done on the cpu cooler...Why_Me said:I've still yet to meet anyone who's purchased a $90 - $100 B560 board and paired it up with a $560 i9 11900K. -
TerryLaze
Because intel has to account for AVX 512 power draw while AMD uses an inferior and much less power heavy version of AVX.jgraham11 said:Intel 11800h @ 45w runs at base of 2.3GHz vs Amd's 5800h base clock is 3.2GHz at the same 45watts. That's almost 1GHz faster, using the same power.
Also AMD hides a big amount of power draw by making it an mobo thing.
As steve found out 'Applications with high thread counts, and/or “heavy” threads, can encounter PPT limits' so AMD throttles as well if not being provided with enough power from the mobo.
https://www.gamersnexus.net/guides/3491-explaining-precision-boost-overdrive-benchmarks-auto-ocPackage Power Tracking (“PPT”):
The PPT threshold is the allowed socket power consumption permitted across the voltage rails supplying the socket. Applications with high thread counts, and/or “heavy” threads, can encounter PPT limits that can be alleviated with a raised PPT limit.
Default for Socket AM4 is at least 142W on motherboards rated for 105W TDP processors.
Default for Socket AM4 is at least 88W on motherboards rated for 65W TDP processors. -
Dragonking_1 Up to 115W TDP (Intel) on mobile chips. Wow. Those chips will be a big challenge to cool - esp. with a GPU on the side. Could make thin laptops even more expensive with a need to put in so much more effort into cooling. Hope then that the 2+8+2/4+8+2 are great performers.Reply
Given the atom cores will run cooler - what matters is the wattage in the end right?
Imagine the desktop versions in Desktop replacement laptops - gonna be behemoths. I know there isn't a big market for desktop replacement laptops, but I am one of those who is in that category, requiring a desktop and a desktop replacement, sadly. -
Dragonking_1 TerryLaze said:Because intel has to account for AVX 512 power draw while AMD uses an inferior and much less power heavy version of AVX.
Also AMD hides a big amount of power draw by making it an mobo thing.
As steve found out 'Applications with high thread counts, and/or “heavy” threads, can encounter PPT limits' so AMD throttles as well if not being provided with enough power from the mobo.
https://www.gamersnexus.net/guides/3491-explaining-precision-boost-overdrive-benchmarks-auto-oc
Yea, there should be a standard that sets how a CPU's TDP should be shown to the consumer. AMD uses 'heat' to measure its TDP which is stupid for real.
But yea, AMD doesn't outright say that 105 or 65W is its max. draw, and neither does Intel. I though prefer the fact that AMD only jumps to 142W at PBO oc., for the performance it gives; compared to >190W (upto what 230W for the 11900K) for Intel 10th and 11th gen while showing a 125W TDP. That's a huge gap for Intel. Though Intel running power hungry or hot is not new info.
I don't get why both don't just publish the max power they can draw. Its gonna be published by 3rd party and discussed by consumers anyway. -
TerryLaze
You have to completely disable every safety feature for the mobo to even give you that much power and then you need to run a special software that has the sole purpose of drawing as much power as possible, it has nothing to do with how any CPU is supposed to be used and that's why they are not giving you a completely crazy and useless number, while the media is trying their hardest to make you believe that somehow this is relevant to you.Dragonking_1 said:I though prefer the fact that AMD only jumps to 142W at PBO oc., for the performance it gives; compared to >190W (upto what 230W for the 11900K) for Intel 10th and 11th gen while showing a 125W TDP. That's a huge gap for Intel. Though Intel running power hungry or hot is not new info.
Because that's just a hardware spec that would be, and maybe even is, buried somewhere next to the humidity it can work in an the min max temp.Dragonking_1 said:I don't get why both don't just publish the max power they can draw. Its gonna be published by 3rd party and discussed by consumers anyway.