News Intel's flagship Lunar Lake Ultra 9 is slower than an Ultra 7 in the single laptop where it's currently for sale

[bit_user]

Intel created the E cores to get AMD out of the laptop/low end desktop market in california.

I doubt this, as their E-cores date back to 2008 (see Intel Atom) and primarily factored into their attempts to penetrate the cell phone & tablet SoC market. Even after they aborted that effort, they've continued to develop them for embedded applications and low-cost Chromebook-class laptops. They've offered them in the cheapest NUCs going back at least as far as Apollo Lake (Goldmont).

They pushed a bill through the california state senate demanding certain cpus become more "power efficient" and effectively banned all AMD cpu sales in california started in 2025 i think. The "E" cores are supposed to qualify intel chips for this energy efficiency, the hilarious part in all of this, is ever since intel deployed their E cores their cpus have gotten less and less efficiency, meaning the purpose of the california law is being unfulfilled because intel wrote it in such a way as to not hold intel to any energy usage standards, just a core which volts down to a certain number.

As we can see here, intel isn't even trying to pretend E cores are energy efficient anymore. they don't have to, the california law is on the books and supposedly they are in compliance. that is why you see AMD rushing to market their own efficiency core chip. they don't want to be locked out of the cali market.

What are your sources on this?

FWIW, AMD had hybrid APUs in laptops for more than a year and have offered them in AM5 packages since the beginning of this year. See the Ryzen 8000G series, one of which is reviewed here:

https://www.tomshardware.com/pc-components/cpus/amd-ryzen-7-8700g-cpu-review
​

I'm pretty sure the main reason AMD pursued their C-core strategy was for the density-optimized cloud market, where they're mainly going up against ARM. The APU market for them seems secondary.

P.S. Let's not forget that AMD once had the "cat" line of E-cores, which were prominently featured in the PS4 and XBox One.

Jaguar (microarchitecture) - Wikipedia

en.wikipedia.org

 

[thestryker]

There is a whole paragraph in your link testing everyday use case on these cpu's. Lunar Lake absolutely nails strix point on the cross. It's not competitive. Nobody cares about cinebench on a loop on thin and light laptops but about baterry life on normal everyday workloads. Strix points needs 60% more power.

You can also keep in mind that reviewer did power limited tests using TDP limits which aren't even close to accurate on AMD. Due to this none of the power limited tests represent accurate information. I imagine AMD would still be ahead simply due to the massive increase in thread count, but the question is by how much which is certainly not answered there.

The sad part is you can see this reflected in their numbers because they measured overall power. This means they could have correctly normalized for actual power consumption and chose not to.

 

[abufrejoval]

Where to start?

These "i9" and "i7" chips were produced in the very same process perhaps even on a single wafer which aims for "perfect" chips all the time.

Binning for defects may be ever less of a thing these days and binning for voltage/frequency is never a production target per se.

But just how probable is it that those binnings just happen to align with the planned Intel product spread and expected sales numbers? Say Intel aims for 15% sales of i9 vs. 50% i7 and the rest going into various other submodels, just what are they supposed to do when in fact 80% of all non-defective chips qualify for i9?

Putting them into the "the i7" bin or the one labelled "i9" won't be a precise art any more, but a percentage bias on randomly selected chips, which still retain some individual spread yet still well within i9 operating characteristics.

And if then peak frequencies aren't ever really hit because various power limits are reached first, the fused-in frequency limits for i7/i9 differentiation never present a noticeable barrier to performance and you're left with the slight random spread between individual chips or even machines, which might have some slightly production differences e.g. with regards to heat pipes: you can't eliminate those differences in laptops with soldered CPUs.

And it's quite obvious that E-cores that are so much improved in terms of efficiency can occasionally extract more compute power in a workload that is entirely constrained by a shared Wattage pool limit than the same number of P-cores at that very same load point: if you increase the Wattage ceiling Watt-by-Watt you may be seeing all kinds of cross overs, but as you introduce slight waits/delays for inputs or game actions, P-cores might be able to sprint faster with left-over Watts in other situations, providing perhaps better reactivity or lower maximum latencies etc.

Complex architectures won't give simple results and the number of variables are only increasing.

I totally expect you to give up sooner or later, throw up your hands and just admit that meaningfull testing of laptops becomes near imposssible, because the ranking will depend entirely on use cases and individual priorities and settings.

That's perhaps why I also decided to just go cheap on my latest laptop purchases and went with older generations and far more reasonable prices: the increments in terms of whichever performance metric I was looking at was far too insignificant compared to the financial benefit of picking machines from inventory being cleared.

I'm really glad I don't depend on having to make a living from selling the leading edge laptops these days.

 

[Darkbreeze]

E-cores also run faster than P-cores during heavily threaded work, but that’s by design.

This literally makes no sense to me. At all.