Thursday, August 14th 2025

Intel "Lunar Lake" Updated PL2 Setting Can Yield Up to 30% Higher Gaming Performance

We recently published our MSI Claw 8 AI+ A2VM review, which is the only handheld gaming device to utilize the Intel Core Ultra 7 258V "Lunar Lake" CPU, featuring four E-cores and four P-cores, paired with the Intel Arc 140V Xe2 GPU. During our testing, we manually set different TDP values via the Manual power profile slider to see what the results are in gaming and how it compares to other, mostly AMD-based, handhelds. However, Intel notified us about performance levels used in testing, and that the official specification for Lunar Lake states that PL2 should always be at least 1 W higher than PL1, or always set at 37 W (the maximum TDP value for Lunar Lake chips). This means that whenever a CPU can extract an additional boost, it will, yielding up to 30% more performance, as we have seen in our own testing.

Initially, we used the Manual power profile, which had allowed owners to match PL2 to PL1. However, Intel noted that when PL2 is set to at least a single W higher than PL1, there is an expected 10% performance increase across the suite of games, but depending on the scenario, it can lead to a 30% increase in some instances. An update has arrived for the MSI Center M app, which locks PL2 to always be at least 1 W higher than PL1. As a result, you can no longer match the two. It was only possible before the update. In our retest, Cyberpunk 2077 and Space Marine 2 achieved nearly a 30% performance uplift with this setting. The Lunar Lake-based MSI Claw AI+ A2VM handheld is the highest-performing device we have tested across the board, even surpassing AMD-based devices in our testing suite. In other games, baseline improvement is set at 10%, just as Intel had notified. Our testing confirmed that Lunar Lake is now the most powerful gaming CPU for handheld devices that we own and have tested.

15:57 UTC: We updated some of the claims to better reflect the PL2 setting and that we manually set TDP values, too.
Below you can find the updated performance graphs.

For more details, 1080p testing, and 1% lows, check out the MSI Claw 8 AI+ A2VM review.
Add your own comment

38 Comments on Intel "Lunar Lake" Updated PL2 Setting Can Yield Up to 30% Higher Gaming Performance

#26
phints
30% faster in a few special cases by jacking up the TDP to an absurd for mobile 37W.
Posted on Reply
#27
JustBenching
DavidC1Did you even read the article? What part of it is overvolting and overclocking?
Of course not, it's an Intel Thread, they don't need to read the article, theyll just downplay intel anyway they can. Case in point the very next post.
phints30% faster in a few special cases by jacking up the TDP to an absurd for mobile 37W.
Apparently another one that didn't read the article. You don't need to jack up the tdp to 37w, all you need to do is set PL2 1w higher than PL1.

Why can't people act like normal human being in intel / nvidia threads? What the actual hell..
Posted on Reply
#28
Cheeseball
Not a Potato
Hecate91Apparently not, and it takes 30W for it to be faster than a 25W Zotac Zone.
Every watt matters on these handhelds, battery life and heat needs to be retested with the CPU allowed to draw up to 37W.


Coincides with my testing.
KarathActual difference in power usage is minor, and it helps a lot with iGPU work. MTL had the same behavior, you could set PL2 to something like 100w with TAU at something silly short like 1-4s and it would perform much better. And average battery life would be roughly the same.

Actually, both MTL and LNL work best with PL1 and PL2 maxed out. It does not change battery life for the worse at all. You can use FPS caps and quality settings to control CPU/iGPU power.
This is true, but the reason why you want PL1 and PL2 set properly is for errant apps that request for more power even though it doesn't need it. An example of this is Fallout 3 where you can absolute run it at 60 FPS at 13W TDP with no frame drops (except for the usual engine jank) and it will only use 16W total system draw, but if you have PL1/PL2 set to the max 30W-37W, it will absolutely keep using more power even though it doesn't need despite locking frame rate to 60 FPS.
bgxBut the most interesting test will be the same plateform, MSI Claw AMD vs Lunar Lake.
This is also a test I wish to try as I'm waiting for the MSI Claw A8 to be available here US-side. I can only simulate the Z2 Extreme for now with my HX 370 devices, which is not going to as accurate as I would like.

The Z2 Extreme is 3C+5c/16T/16CU so it should be match-for-match with Lunar Lake in core configuration and it should be able to beat the 258V in multi-thread performance (16T vs 8T, duh), but is beaten in single-thread performance because the IPC is technically better. I have confirmed this with multiple HX 370 devices and the two Lunar Lake devices (258V and 288V) I have access to, but they are close to each other in performance.

iGPU performance trade blows with each other closely in various games. For example Elden Ring runs faster on the 890M by 10 FPS but Apex Legends runs 12 FPS faster on the 140V, both devices at 1200p, same game settings, and locked to PL1/SPL 17W.
Posted on Reply
#29
DavidC1
phints30% faster in a few special cases by jacking up the TDP to an absurd for mobile 37W.
It's PL2. Do you understand what that means or not? Maybe you read about it but you glanced through it? Or got confused by desktops that set PL2 timer as infinite?

READ this time and actually try to understand. Mobiles CANNOT have infinite PL2 timer. PL2 is temporary.

The power management of modern computers are FAR more advanced than you think they are. We're not in 2002 where Battery = 800MHz and Plug = 1GHz and stayed there forever.
Posted on Reply
#30
watzupken
I think will be good to have a view if there’s any impact to power draw after the update. Such big performance improvement normally don’t come without cost, unless it is a problem with optimization.

Lunar Lake is an impressive chip from Intel for sure, but I think we cannot directly compare it with the likes of the current gen of Ryzen chip because they are built on different TSMC node. Lunar Lake has an advantage being on 3nm instead of 5nm (even in name it’s called 4nm) on the AMD chips. So while LL power draw includes memory, it’s already having an advantage from a node perspective.
Posted on Reply
#31
DavidC1
watzupkenI think will be good to have a view if there’s any impact to power draw after the update. Such big performance improvement normally don’t come without cost, unless it is a problem with optimization.

Lunar Lake is an impressive chip from Intel for sure, but I think we cannot directly compare it with the likes of the current gen of Ryzen chip because they are built on different TSMC node. Lunar Lake has an advantage being on 3nm instead of 5nm (even in name it’s called 4nm) on the AMD chips. So while LL power draw includes memory, it’s already having an advantage from a node perspective.
It doesn't matter. CPU reaches the power limit that is set. Both chips are set to same power. Lunarlake platform is just much more optimized, so it gets it's advantages everywhere else such as having a lower memory power and so on.

Also, N3B on Lunarlake is a minor advancement on the power level over N4P on Ryzen. Something like 6-8%. Also, AMD chose not to use N3. As a customer you are buying a black box.
Posted on Reply
#32
bgx
DavidC1CPU reaches the power limit that is set. Both chips are set to same power.
not really. AMD and Intel does not count power the same. Clearly, AMD allow to draw more power / intel is more conservative in the powerdraw it seems when both are set with the same power.

In particular, the RAM is on the SOC so included in the 25W for intel, and not for AMD (should be 1-2W, i.e. 8% difference)

Again, I hope we can have actual consumption of the plateform Claw AMD vs Claw Lunar Lake (time / battery), not just relying on what different manufacturer claim.
same screen, same almost everything, but the SOC (and the southbridge).
DavidC1Lunarlake platform is just much more optimized, so it gets it's advantages everywhere else such as having a lower memory power and so on.
memory is not lower power, it is just included in the SOC package of intel.
Posted on Reply
#33
Vegemeister
Intel's PL2 throttling algorithm is a lot less efficient than their PL1 throttling algorithm. I have seen this with desktop parts too.

It requests excess voltage margin that it doesn't use, changes the clock frequency too much, and downclocks in big jumps.

Efficiency would probably be even better with PL1 and PL2 maxed out so they're never hit, plus a userspace program controlling clock frequency to maintain constant FPS measured with presentmon.
Posted on Reply
#34
Rightness_1
dyonoctisAre you saying that a +1w PL2 at sub-60w is going to be unstable on a CPU that uses the same Lion Cove architecture used in Aroar Lake?
So 30% more performance comes from magic?
Posted on Reply
#35
colossusrageblack
TumbleGeorgeStrange, resolution of the device display is 1920x1200. I think that results on its natural resolution must be worser than on FHD. The iGPU already is weak for most of games on FHD resolution. This is visible from tables with results.
Handhelds at 1200p are excellent at scaling down to 1000p and 800p without looking blurry. This is mostly because the screen is only 8" and the pixel density is good enough to make it look clean. All of these handhelds are fairly weak for 1080p, but that doesn't mean that having a higher resolution display is bad, it just gives you options for whenever you run something that is light weight enough to run at native resolution.
Posted on Reply
#36
dyonoctis
Rightness_1So 30% more performance comes from magic?
I think that you don't understand how mobile chips work/are made. That update just allowed the chip to use more power in the PL2 settings, which increases performance. (and that an update from MSI, not Intel)
But at wattages that are still very low for what Lion coves can exploit. It's the same architecture used in the +200W desktop chips. Mobile chips are just made with high-quality silicon that can use a lower voltage to be more efficient. And tuned for efficiency. That means lower power limits, shorter bursts of turbo boost, resulting in lower sustained performance.

Overvolting is only necessary when you push the silicon to/beyond its limits. That lunarlake chip is running at fairly low clocks, compared to what lion cove can achieve, there's no need to overvolt. If an overvoltage were necessary, that would have been the worst grade of silicon ever, and would have been a defect. If you read the reviews, you can see that this chip stays below 3Ghz in sustained load, and even under 2Ghz in gaming. Why would you need to overvolt to reach those clocks??? Again, it's the same arch used in the desktop chips, just tuned for low power.
www.notebookcheck.net/HP-OmniBook-X-Flip-16-convertible-review-Luxury-features-for-a-midrange-price.1049258.0.html#c13736824
Posted on Reply
#37
DavidC1
bgxnot really. AMD and Intel does not count power the same. Clearly, AMD allow to draw more power / intel is more conservative in the powerdraw it seems when both are set with the same power.
Set power has to be accurate for laptops or you'll have thermal issues
bgxmemory is not lower power, it is just included in the SOC package of intel.
Memory is lower power for Intel, because nearly every memory power used is up by the controller and Lunarlake has controller power optimizations for 40% lower power. Lunarlake is significantly more efficient than AMD chips because of this regard under lower load conditions because of this and others.
Posted on Reply
#38
randomdude9182
JustBenchingOf course, Intel on top - we have to make some stuff up to not admit it. It's dominating everything at 25w (26 with the fix I guess), obviously that's going to cause degradation :kookoo:
Lol. Its beacuse amd's stuck on 4nm, whilst intel and Apple got to 3nm. This by default makes them a shit ton more efficient.
Posted on Reply
Add your own comment
Nov 13th, 2026 23:56 CST change timezone

New Forum Posts

Popular Reviews

TPU on YouTube

Controversial News Posts