Apple Says New MacBook Air Up to 23x Faster Than Intel-Based Model, But Read the Fine Print

[noobinator]

How much faster than M2 and M3?

 

[nt5672]

Odd, because my MacBook M2 Pro beat my old MacBook Pro 16" i9 in absolutely every respect (except thickness).

I'm talking about my daily use which is primarily code development with Xcode, LLVM, and clang. The i9 having at least a 1Ghz advantage in single core processor speed over the M4 is huge in my workflows. In spite of Apple marketing the simpler instruction set of the M4 does not make up the difference.

 

[ossifer]

Sure, but I can natively boot Windows anytime I want on my Intel MBP

I’m sorry to hear that.

“Sorry boss, I have an M3 mac and can’t boot into windows, so you’ll have to find someone else to do that mindless soul-numbing raks on MS Windows…”

 

[iObama]

I'll take that bet and raise you the original CoreDuo, non-64 bit Intel macs. Or the base 2014 mini.

Ohhhh no no no. I've had both of those. The original Core Duo machines were lightning fast compared to their PowerPC counterparts. The base 2014 mini wasn't fast, but the 2018-2020 Intel Airs are painfully slow and they were out of the box.

 

[hans1972]

and safari still sucks and freezes constantly.

Not everyone can handle Safari like my 71-year old mother does without any problems. For those who fail Safari and have special needs, there are more complex browsers out there.

 

[hans1972]

comparing it to a 6yo chip is peak dishonesty

It's a marketing message on their website trying to convince people to buy the new MBA. That's easier when you have an older Mac and thus the marketing message is tailored to those.

If you look at the website you can even compare it to newer MBAs:

 

[Manzanito]

It's a marketing message on their website trying to convince people to buy the new MBA. That's easier when you have an older Mac and thus the marketing message is tailored to those.

If you look at the website you can even compare it to newer MBAs:
View attachment 2489501

Why stop there, then? Compare them to the Core 2 Duo. Even better, to PowerPC. They used to compare them with the previous generation, at some point they should move on from comparing with intel.

 

[mxrider88]

And 11,397,829 times buggier than the g4 PowerBook titanium.

 

[dandeco]

My M1 MacBook Air with 8-core graphics and 16 GB of RAM can render a video project in just one third of the time it'd take my 2.3 GHz quad-core i7 Mac Mini with 16 GB of RAM (2012) to render the same project! I remember that blowing me away when I bought the Air nearly four years ago.

 

[!!!]

At least they let us know exactly what device they are comparing against. Other brands compare their best products against the weakest of the current Macs, and they call it a win against Macs.

Also, I think this comparison is valid. Current owners of Intel Macs and M1 MacBook Airs are the ones who will buy these new models, not the owners of M2/M3 MBAs.

No, Apple is using the intentionally ambiguous Intel "core i" naming scheme and avoiding saying the year:

Intel Core i7-based MacBook Air systems

 

[PegLegDreg]

Are you using an Intel Mac, or are you making this up?

I usually keep about a hundred Safari tabs open, sometimes several hundreds of them, and never had any issues in any of my M-series Macs.

On the other hand, this may be because my devices are usually loaded with RAM.

Why would you have several hundreds on safari tabs open?

 

[PegLegDreg]

And older Intel Macs can still be used for most tasks that an average consumer would run

I still have a base model Mac mini bought in 2020 and it’s an i3 with 8 GB RAM.

Still running great for productivity and web browsing to this day.

Exactly. That is why they are making these comparisons. Most people do the very basic with computers, including me.

 

[SlaveToSwift]

I'm talking about my daily use which is primarily code development with Xcode, LLVM, and clang. The i9 having at least a 1Ghz advantage in single core processor speed over the M4 is huge in my workflows. In spite of Apple marketing the simpler instruction set of the M4 does not make up the difference.

So Clang doesn’t take advantage of multi-threading?

Cool story bro, yet another of the infinite reasons why Swift replaced it. As my swift builds in my M4 Pro take 20% of the time they took on an i9.

 

[chucker23n1]

So Clang doesn’t take advantage of multi-threading?

Cool story bro, yet another of the infinite reasons why Swift replaced it. As my swift builds in my M4 Pro take 20% of the time they took on an i9.

Nah. The M4 Pro runs circles around the 2019 MBP’s i9, including at Clang.

 

[seek3r]

I'm talking about my daily use which is primarily code development with Xcode, LLVM, and clang. The i9 having at least a 1Ghz advantage in single core processor speed over the M4 is huge in my workflows. In spite of Apple marketing the simpler instruction set of the M4 does not make up the difference.

Clock speed cant be compared directly between architectures, saying “a 1ghz advantage” is literally meaningless, even if it were accurate, the m4 pro can do a lot more per cycle than that i9. What’s particularly funny is it’s not even accurate, the M4 pro’s cores run at a higher clock than that i9, the high performance cores at nearly twice the clock and even the efficiency cores at about 500mhz higher

 

[ossifer]

Ohhhh no no no. I've had both of those. The original Core Duo machines were lightning fast compared to their PowerPC counterparts. The base 2014 mini wasn't fast, but the 2018-2020 Intel Airs are painfully slow and they were out of the box.

And the original PowerPC machines were 30-pin fast compared their Motorola 68k predecessors…

 

[ct1211]

That's fine.

It's this to be aware of: "Apple said its test consisted of using the Super Resolution feature on a 4.4MB image in image editing app Pixelmator Pro."

Very niche, specific task to achieve a mind-blowing 23x-faster marketing claim.

The other tasks that are in the 3-4x faster range are better examples of performance gains to expect when upgrading from an Intel-based MacBook Air, in my opinion.

Still very impressive, as my article concludes.

Impressive, who cares, the M3 was fast enough, how about now that you've pulled ahead in this centuries version of "the Wattage Wars" from way back in the 70's when steroo technology got all the press, how about you deliver a file management system that works? How about now that the M4 iPad Pro can handle a moon launch you give it some productive features more akin to a computer. How about you guys get Google on the phone and sign on with thier voice assistant and dump SIRI, it has to hold the record for "Vital technology that's been broken for over a decade - A decade of clueless responses! I'm trying a Surface Pro 11 with a SD Elite X, 32gb RAM and 1TB SSD I just bought on sale for $1,499, about half of a comparable MBP and so far its been a dream to use.

 

[raccoontail]

Also, it's 860,000x faster than iBook

And it runs hypercard decks 1,630,423 times faster than an Macintosh SE!

 

[WRONG]

OMG Snow Leopard!! Literally my favorite OS of all time! I thought I was the only one who had a special place in my heart for it.

You're not the only one.
Boy do I miss Snow Leopard, my very first MacOSX on my very first MacBook Pro.. ah memories

 

[wyrdness]

I'm talking about my daily use which is primarily code development with Xcode, LLVM, and clang. The i9 having at least a 1Ghz advantage in single core processor speed over the M4 is huge in my workflows. In spite of Apple marketing the simpler instruction set of the M4 does not make up the difference.

Not so different from my workflow of CLion, LLVM & Clang. And my M2 Pro is an order of magnitude faster for Clang than my 2019 Intel MacBook Pro.

Geekbench's multi-core Clang benchmark suggests that both the M2 Pro and the M4 compile over twice as fast as the 2019 i9 MacBook.
If you're not seeing an improvement, then it's not the CPU that's the problem.