News Intel Lunar Lake CPU benchmarks reveal good power efficiency and GPU performance — Core Ultra 7 268V results show regressions in multi-core perform...

[JamesJones44]

It's a little surprising sites seem to keep coming back to comparing to the M3. The M4 has been out for 6 months now, seems like it would be more relevant to compare up and coming silicon to the current generation of processors rather than previous generation processors. It has kind of gotten to be like comparing 9xxx to 5xxx AMD processors instead of 9xxx to 7xxx.

 

[thestryker]

It's a little surprising sites seem to keep coming back to comparing to the M3. The M4 has been out for 6 months now, seems like it would be more relevant to compare up and coming silicon to the current generation of processors rather than previous generation processors. It has kind of gotten to be like comparing 9xxx to 5xxx AMD processors instead of 9xxx to 7xxx.

What laptops are the M4 SoCs available in?

 

[JamesJones44]

What laptops are the M4 SoCs available in?

It's only in the iPad Pro line at the moment, but if rumors are to be believed it most of Apple's laptops will get a version of it in October. However, the article also makes reference to the Z1 Extreme which can only be had in handhelds at the moment. The article has expanded the comparison beyond laptops by including the Z1 Extreme.

 

[thestryker]

It's only in the iPad Pro line at the moment, but if rumors are to be believed it most of Apple's laptops will get a version of it in October.

Right so there's nothing to compare it to (ios vs macos being the bigger problem than form factor).

However, the article also makes reference to the Z1 Extreme which can only be had in handhelds at the moment. The article has expanded the comparison beyond laptops by including the Z1 Extreme.

Not really because the Z1 Extreme is the same thing as the 7840u minus the NPU.

 

[JamesJones44]

Right so there's nothing to compare it to (ios vs macos being the bigger problem than form factor).

That doesn't really matter. Scores from iPadOS have pretty much matched what ends up being in Apple's laptop lines. This was true for both the M1 and M2 iPads and Macbook Air (there isn't an iPad with an M3 to compare). macOS and iOS share the same base kernels.

https://browser.geekbench.com/macs/macbook-air-2022

https://browser.geekbench.com/ios_devices/ipad14-5

Not really because the Z1 Extreme is the same thing as the 7840u minus the NPU.

They are close, but not the same, the Z1 has optimizations for lower wattage operations which does have an affect on overall performance, marginally, but an affect none the less. Also, it's still comparing a laptop SoC to a handheld SoC, if they didn't want to make that comparison they could have simply referenced the 7840u

 

[thestryker]

That doesn't really matter. Scores from iPadOS have pretty much matched what ends up being in Apple's laptop lines. This was true for both the M1 and M2 iPads and Macbook Air (there isn't an iPad with an M3 to compare). macOS and iOS share the same base kernels.

https://browser.geekbench.com/macs/macbook-air-2022

https://browser.geekbench.com/ios_devices/ipad14-5

You're citing geek bench which wasn't mentioned at all in this article. Cinebench which was the used benchmarking software isn't available on ios. The OS limits comparisons a lot more than the form factor does.

They are close, but not the same, the Z1 has optimizations for lower wattage operations which does have an affect on overall performance, marginally, but an affect none the less. Also, it's still comparing a laptop SoC to a handheld SoC, if they didn't want to make that comparison they could have simply referenced the 7840u

They're running everything directly compared at the same power profiles so there's no effective difference. While this isn't a particularly great way of doing it as absolute power consumption is different using the same profile when comparing Intel to AMD (and Apple as well) it's appropriate for AMD to AMD.

 

[JamesJones44]

You're citing geek bench which wasn't mentioned at all in this article. Cinebench which was the used benchmarking software isn't available on ios. The OS limits comparisons a lot more than the form factor does.

That's a fair point. I've seen others do it with Geekbench and/or SPEC scores, in this case there is at least an argument to be made on why they are doing it against M3 vs M4.

 

[TheHerald]

As I've said in the other thread, it's not hitting anywhere close to the 24 hours some early leaks suggested (the reviewer mentioned it as well) but it still beats amd handily in battery life (while having a smaller battery as well, just wow). Not bad.

View: https://www.youtube.com/watch?v=CxAMD6i5dVc

 

[systemBuilder_49]

4 cores 8 threads/eCores in 2024. Wow. Just wow. I thought we moved beyond that in 2014. ..

Its like Intel started chopping off arms and legs until they hit a power target, like the M-series and Y-series throttlebooks of 2015-2018 ...

 

[TheHerald]

4 cores 8 threads/eCores in 2024. Wow. Just wow. I thought we moved beyond that in 2014. ..

Its like Intel started chopping off arms and legs until they hit a power target, like the M-series and Y-series throttlebooks of 2015-2018 ...

Cores don't matter, since noone cares that Intel offers 14 and 20 cores on the desktop when their competitor offers 6 and 8. Now suddenly it's important again on the ultralight laptop category? Yeah, right

 

[rm12]

Cores don't matter, since noone cares that Intel offers 14 and 20 cores on the desktop when their competitor offers 6 and 8. Now suddenly it's important again on the ultralight laptop category? Yeah, right

Correct that 'cores' don't matter per se. But how many processes and/or threads can be run in a performant way matters. At work 4 cores in my laptop from which at least 1 is continuously fully used by 'mandatory' software, e.g. anti virus ... so yes headroom to have a usable laptop in e.g. two years time seems advisable.