M2 smokes 13th gen Intel Raptor Lake P mobile CPU in single core benchmarks?

[Wizec]

I thought maybe I would give Intel a try again with Raptor Lake, but at least one site is seeing single core scores of only 1655 for the Core i7-1370P

Intel 13th Gen Core i7-1370P 'Raptor Lake-P' Laptop CPU Leaks Out, 14 Cores & 5 GHz

Intel's 13th Gen Core i7-1370P 'Raptor Lake-P' CPU which replaces the laptop 12th Gen Core i7-1270P 'Alder Lake-P' has leaked out.

wccftech.com

That’s 12% slower in single core than the MacBook Air M2 where we see 1889

MacBook Air (2022) Benchmarks - Geekbench

 

[falainber]

I thought maybe I would give Intel a try again with Raptor Lake, but at least one site is seeing single core scores of only 1655 for the Core i7-1370P

Intel 13th Gen Core i7-1370P 'Raptor Lake-P' Laptop CPU Leaks Out, 14 Cores & 5 GHz

Intel's 13th Gen Core i7-1370P 'Raptor Lake-P' CPU which replaces the laptop 12th Gen Core i7-1270P 'Alder Lake-P' has leaked out.

wccftech.com

That’s 12% slower in single core than the MacBook Air M2 where we see 1889

MacBook Air (2022) Benchmarks - Geekbench

Core i7-1370P is the least powerful model of mobile Raptor Lake lineup. The full lineup is described here.

 

[theorist9]

When the M1 came out it had a higher GB5 SC score than any mobile chip from Intel or AMD. That continued with the M2 (GB5 SC ~ 1900, released June 2022), which beat Intel's highest-end ultra-enthusiast mobile chip, the i9-12900HX (GB5 SC ~1850, released May 2022). I expect Apple's mobile superiority will continue if and when we get to compare the 3 nm M2's and the i9-13900HX (both will probably be released in Spring 2023).

By using high clocks and power budgets, Intel can drive their highest-end desktop chips to faster SC speeds than contemporary Apple M-series chips. But they don't have that option in mobile devices and thus Apple, with their high efficiencies, has been able to retain the mobile SC crown.

It's of course notable that this gives Apple a fanless thin-and-light laptop (the Air) with higher SC speeds than Intel's and AMD's highest-end 7 lb. gaming laptops.

EDIT: There are a few laptops that do use desktop CPU's, like this 17" beast from Origin. But while it's a current model, it's limited to 11th-gen Intel CPU's (max option is i9-11900K, GB SC = 1811); apparently it can't handle the thermals from an Alder Lake desktop. So even it has a lower GB5 SC score than an M2 MacBook Air:

 

[leman]

Those Intel scores are a bit on the low side, but it would surprise me if Raptor Lake is much faster in single core than its predecessor. I’m the end, it’s the same CPU with tweaked clocks and more E-cores.

 

[Gudi]

1655 is a very respectable single core score, with 1687 the M1 MacBook Air shocked everybody. It's all the other problems of the Intel platform and PC industry, which make a Windows laptop a horrible choice.

 

[MayaUser]

Core i7-1370P is the least powerful model of mobile Raptor Lake lineup. The full lineup is described here.

is least powerful even than i5-13420H ?!

 

[theorist9]

Those Intel scores are a bit on the low side, but it would surprise me if Raptor Lake is much faster in single core than its predecessor. I’m the end, it’s the same CPU with tweaked clocks and more E-cores.

The i9-11900K and M2 scores were taken from GB's list of vetted average scores (https://browser.geekbench.com/processor-benchmarks for the Intel). Curiously, they don't have vetted scores for any 12th-gen Intel mobile processor (in fact, nothing beyond the Intel Core i7-11800H), so I got the i9-12900HX's score from a different source--forget where. But if I instead use notebook checks's median scores, I get these, which says the M2 is faster than the i9-12900H, but about the same as the i9-12900HX:

i9-11900K: 1788
i9-12900H: 1848
i9-12900HX: 1907
M2: 1914

Intel Core i9-12900HX Processor - Benchmarks and Specs

Benchmarks, information, and specifications for the Intel Core i9-12900HX

www.notebookcheck.net

 

[senttoschool]

The i9-11900K and M2 scores were taken from GB's list of vetted average scores (https://browser.geekbench.com/processor-benchmarks for the Intel). Curiously, they don't have vetted scores any 12th-gen Intel mobile processor (in fact, nothing beyond the Intel Core i7-11800H), so I got the i9-12900HX's score from a different source--forget where. But if I instead use notebook checks's median scores, I get these, which says the M2 is faster than the i9-12900H, but about the same as the i9-12900HX:

i9-11900K: 1788
i9-12900H: 1848
i9-12900HX: 1907
M2: 1914


View attachment 2120967
View attachment 2120968

Intel Core i9-12900HX Processor - Benchmarks and Specs

Benchmarks, information, and specifications for the Intel Core i9-12900HX

www.notebookcheck.net

Also note that the type of laptop the 12900HX goes into looks like this:

It also boosts up to 155w, on a laptop.

Conversely, an M1/M2 chip will sip power between 0.3w - 5w during a Geekbench5 ST test in a fanless Macbook.

 

[Pressure]

Core i7-1370P is the least powerful model of mobile Raptor Lake lineup. The full lineup is described here.

The M2 is also the least powerful chip in the M2 line-up 🤷🏼‍♂️

 

[theorist9]

Also note that the type of laptop the 12900HX goes into looks like this:

It also boosts up to 155w, on a laptop.

Conversely, an M1/M2 chip will sip power between 0.3w - 5w during a Geekbench5 ST test in a fanless Macbook.

Agreed, I made that point here:

It's of course notable that this gives Apple a fanless thin-and-light laptop (the Air) with higher SC speeds than Intel's and AMD's highest-end 7 lb. gaming laptops.

 

[1@@@]

-

 

[Kazgarth]

That's because it's still using Intel's ancient 10nm+++ process. If they clock it any higher its gonna drain battery like no tomorrow.

AMD however should be more interesting once they release their 5nm Ryzen 7000 mobile CPUs. And I'd love to see how they compare to M2 in performance and efficiency.

 

[exoticSpice]

That's because it's still using Intel's ancient 10nm+++ process. If they clock it any higher its gonna drain battery like no tomorrow.

AMD however should be more interesting once they release their 5nm Ryzen 7000 mobile CPUs. And I'd love to see how they compare to M2 in performance and efficiency.

AMDs mobile CPUs will be on 4nm.

 

[Love-hate 🍏 relationship]

1655 is a very respectable single core score, with 1687 the M1 MacBook Air shocked everybody. It's all the other problems of the Intel platform and PC industry, which make a Windows laptop a horrible choice.

M1 is actually around 1770

 

[Gudi]

M1 is actually around 1770

Geekbench scores may vary.

 

[apparatchik]

The M2 is also the least powerful chip in the M2 line-up 🤷🏼‍♂️

Yep, but unlike Intel, Apple Silicon shares the same core design so single core performance is about the same from M1 to M1 Ultra, so for single threaded tasks you get the same perf from MBA to Mac Studio, which is great.

 

[klasma]

is least powerful even than i5-13420H ?!

Yes, because it’s specced for much lower power consumption. See the table in article link you replied to.

The lower power consumption is also why it makes sense to compare it to the M2.

 

[slooksterPSV]

Just ran GB on mine and um... wow I'm shocked.

 

[RedTheReader]

None of this matters. You buy an Intel machine if you need Windows, or if you need x86 support, or if you need higher multi-core performance, or you need a competent GPU. Yes, that brings heat, and yes, you lose the niceness of macOS, but it's often about what you need to have rather then what you'd like to have.

 

[apparatchik]

None of this matters. You buy an Intel machine if you need Windows, or if you need x86 support, or if you need higher multi-core performance, or you need a competent GPU. Yes, that brings heat, and yes, you lose the niceness of macOS, but it's often about what you need to have rather then what you'd like to have.

A lot of people can work on either platform, cross-shop, and actually jumped ship from Windows with the release of the M1 MBA.

There’s a segment were they do compete and things like battery life and performance related to the chip are central.

 

[Wizec]

None of this matters. You buy an Intel machine if you need Windows, or if you need x86 support, or if you need higher multi-core performance, or you need a competent GPU. Yes, that brings heat, and yes, you lose the niceness of macOS, but it's often about what you need to have rather then what you'd like to have.

I wouldn’t say none of it matters.

I run a MacBook Air M1 with Parallels to run Windows 11. macOS, for me personally, is really unfriendly. Finder especially is an abomination. Constant menu at the top? Sorry, don’t want to turn this into an OS argument, I just really, really hate macOS, but that’s a subjective, personal thing.

Regardless, single thread performance is SO GOOD on the M1, and the device is fanless, cool and quiet, long battery life, lightweigh, etc that for me it is the absolute best silent Windows device money can buy, barring the M2 Air.

Now that Visual Studio supports ARM natively, I can do everything I need to in Windows 11 ARM, without any emulation.

I actually can’t wait for the MacBook Air M3. I plan to get the 24GB model so I can allocate 16GB to my windows VM 😆 Hopefully the 3nm design gets us longer battery life and a couple more E cores or perhaps 300Mhz faster, who knows… maybe 2 of those 3 items…

 

[Gudi]

None of this matters. You buy an Intel machine if you need Windows, or if you need x86 support, or if you need higher multi-core performance, or you need a competent GPU. Yes, that brings heat, and yes, you lose the niceness of macOS, but it's often about what you need to have rather then what you'd like to have.

No, it isn't. Humans are great at ignoring necessities in favor of desires. Guess why gaming and social vanity drive so much of the computing industry? We actually need to bring energy consumption down to survive, but oh-no triple-A gaming is more important than the environment says my bitcoin miner's instagram.

 

[senttoschool]

Geekbench scores may vary.

View attachment 2121111

That's because Geekbench compiles all scores and either does an average or a median. A lot of amateurs run the tests under suboptimal conditions such as having other apps running at the same time or running it in a hot environment which will cause the SoC to throttle.

It's best to compare chips running in optimal conditions instead of an average of random tests. Look for professional reviews. The M1 scores in the 1750s.

 

[praetorian909]

I wouldn’t say none of it matters.

I run a MacBook Air M1 with Parallels to run Windows 11. macOS, for me personally, is really unfriendly. Finder especially is an abomination. Constant menu at the top? Sorry, don’t want to turn this into an OS argument, I just really, really hate macOS, but that’s a subjective, personal thing.

Well I would take constant (aka consistent) menu at the top over Windows 11’s moving Start Menu that now looks like an app icon. Oh and instead of having it in the place where it’s been—you know since 1995—they put the weather widget icon that pops up when you hover.

For me Windows 11 really wins for being most unfriendly...

 

[Ashbash75]

Another windows release and another application to bring the old start menu back.

StartAllBack to fix all Windows 11 deal-breaking UI issues

StartAllBack restores taskbar, start menu, context menus, explorer, sanity, style and order!

www.startallback.com