Apple's comparisons are to an i3 Intel

[jrlcopy]

Why couldn't they have given us proper specs....

Go here and scroll all the way down:

Mac

The most powerful Mac laptops and desktops ever. Supercharged by Apple silicon. MacBook Air, MacBook Pro, iMac, Mac mini, Mac Studio, and Mac Pro.

www.apple.com

Looks like MOST of their test benches were:

3.6GHz quad-core Intel Core i3-based (for all the other claims)
&
a production 1.2GHz quad-core Intel Core i7-based MacBook Air systems (RE: 3.9X faster video processing claim)

Yes most users, including myself for most cases an i3 is fine and the tradeoff for battery life, etc... def is intriguing...

Like I wonder if they just compared it horizontally M1 to i5 (which is mostly what they were selling, then the gains would have been even?) Is 2x an i3 = i5 power but less power draw?

Having said that, I need to upgrade the ol 2014 13" MBP and thinking about going with an air as my choice. I'm Mac Mini eGPU pro now, so don't need crazy portable power.

 

[mr_roboto]

As far as I can tell, in every case they've taken the approach of comparing the new M1 Mac to a specific Intel Mac which they discontinued.

For the 13" MBP, they replaced the low end i3 Intel models (the ones which had two Thunderbolt/USB ports). The 4-port i5/i7 configs of the Intel 13" MBP are still available for purchase. When Apple replaces them, expect the benchmarks to compare to them instead.

 

[taldo]

Apple did a really good marketing show yesterday, and their claims were very vague in terms of actual numbers.

I think that the most important thing that us users should refer to, is what product do i get for the same price.

it seems like M1 is Apple's version of the low-end of Intel chips, and getting a 8 Core CPU and 7/8 Core GPU which supposedly give 3.5X/6X more performance, for the same price, is probably a very good deal for us consumers.

the question is,
is it that big of a deal getting this type of performance upgrades from a 8 core chip instead of a 2 Core chip? i'm not so sure, it very much depends on what u check - if the 3.5X performance upgrade is on a single-core test, then this is a giant leap forward like Apple is trying to claim.
The reasonable scenario is that the performance upgrade is more due to the Multi-Core test, and when u look at the Core count comparison, it makes sense.

having said that,
i think that as a consumer, what we really need to look at is the "value-for-money" test, and although Apple is a high-end company that sells her products for big money, this time they actually nailed it in terms of price.
getting a much better product (much much better, no matter the reasons...) for the same price is a great deal and an important step for Apple in terms of making us consumers prefer Macbooks instead of PC rivals.

does that mean Apple did something revolutionary in a technological-chip-design level? not so sure...

 

[bobmans]

Because the i3’s are the Macbooks they replaced?
This is their low-end entry level chipset which replaced their low-end entry level Intel equivalents which they compared it to.

People seem to not understand that this is Apple’s low end chipset, there will be a whole family of chipsets covering low-mid-high-ultra high end.

 

[leman]

is it that big of a deal getting this type of performance upgrades from a 8 core chip instead of a 2 Core chip? i'm not so sure, it very much depends on what u check - if the 3.5X performance upgrade is on a single-core test, then this is a giant leap forward like Apple is trying to claim.

Apple marketing is vague, but it doesn't matter much. Given the nature of the performance claims they make, the i3 and i5 are not that far off. Consider this: the i3-8100B (Mac Mini CPU) scores around 980/3500 in GB5. The i9-9980HK scores 1200/7000 — that's up to 2 times faster in multi-core. Apple claims 3.5x performance increase from the i3, which would put it on par with fastest currently available desktop x86 chips.

Few more things to consider:

- Anandtech just did tests of A14 and the 5 Watt iPhone CPU is outperforming (single-core) the latest Intel desktop while being within the striking distance of new AMD Zen 3 CPUs. M1 is basically A14 but with more CPU cores and more thermal headroom

- Only 4 cores of M1 are high-performance cores, the other 4 are much slower cores optimized for power efficiency. It is essentially a quad-core CPU and should be compared to other quad-core CPUs. The additional efficiency cores will have impact comparable to hyper-threading on other CPUs

- Apple doesn't do hyper-threading because they don't need to — they have solved the problem by massively increasing instruction level parallelism of their machines

- This is just the low-end chip... just think about it, we have here a 15W low-end Mac chip that is basically on par in some performance metrics with 100+ watt desktop CPUs... while having GPU performance of a 25Watts discrete Nvidia unit. This is absolutely mind boggling

 

[Icelus]

Apple marketing is vague, but it doesn't matter much. Given the nature of the performance claims they make, the i3 and i5 are not that far off. Consider this: the i3-8100B (Mac Mini CPU) scores around 980/3500 in GB5. The i9-9980HK scores 1200/7000 — that's up to 2 times faster in multi-core. Apple claims 3.5x performance increase from the i3, which would put it on par with fastest currently available desktop x86 chips.

It would mean the M1 is as fast as the 10 core iMac Pro -- and I still see reactions where people don't believe this "phone CPU" (as they refer to it) is faster than their desktop CPU.

 

[taldo]

Apple marketing is vague, but it doesn't matter much. Given the nature of the performance claims they make, the i3 and i5 are not that far off. Consider this: the i3-8100B (Mac Mini CPU) scores around 980/3500 in GB5. The i9-9980HK scores 1200/7000 — that's up to 2 times faster in multi-core. Apple claims 3.5x performance increase from the i3, which would put it on par with fastest currently available desktop x86 chips.

Few more things to consider:

- Anandtech just did tests of A14 and the 5 Watt iPhone CPU is outperforming (single-core) the latest Intel desktop while being within the striking distance of new AMD Zen 3 CPUs. M1 is basically A14 but with more CPU cores and more thermal headroom

- Only 4 cores of M1 are high-performance cores, the other 4 are much slower cores optimized for power efficiency. It is essentially a quad-core CPU and should be compared to other quad-core CPUs. The additional efficiency cores will have impact comparable to hyper-threading on other CPUs

- Apple doesn't do hyper-threading because they don't need to — they have solved the problem by massively increasing instruction level parallelism of their machines

- This is just the low-end chip... just think about it, we have here a 15W low-end Mac chip that is basically on par in some performance metrics with 100+ watt desktop CPUs... while having GPU performance of a 25Watts discrete Nvidia unit. This is absolutely mind boggling

First things first - When comparing the M1, we should also compare it in terms of power envelope, which means that we need to compare the M1 mac mini, to the baseline M1, the i3 found at mackbook air to the m1 fanless macbook air, etc.
The reason is that every version of a chip is optimized for a different Voltage level, impacting performance abilities significantly, not to mention cooling effects, etc.

Moreover, the Mac mini's processor is an 8th generation Intel processor - Apple didn't update the mac mini with the latest Intel processor before making this comparison, so comparing it to the prev baseline Mac mini is a good marketing trick.

Also, if u really want to get into the Engineering feat Apple is claiming for itself, lets not forget that the manufacturing technology they used is at least 1.5 generations ahead of the manufacturing technology used to create the 8th generation i3 that the prev baseline mac mini had. the manufacturing technology alone gives u a lot more power efficiency products not to mention the increase in performance.

When u look at Apple's Mac mini page for example, in the CPU section, the numbers they are showing (comparing to the baseline Intel Mac mini..) is up to 3.4x performance upgrade (on final cut pro - faster ProRes transcode for example..) but in Affinity Pro the numbers are 2.3X faster multicore in vector performance, so is it a simple 3.5X on all tasks/apps/stuff u do with a desktop computer? probably not, and lets not forget the fact that they control all the computing process - chip, OS, app - does all the performance upgrade come from the chip itself? app or OS optimization..? don't know and it's probably a really hard thing to say, and that's why the direct comparison between the chips is not that straightforward.

regarding Desktop chips or real Pro chips from Apple -
i think that the major step as already been done, basically the M1 is the same as the A14. i don't expect Apple to do another this big of a jump also at the high-end desktop, especially when u consider all of the above regarding manufacturing technology and the fact that High-end Macs are a bit more updated with the last Intel generations.

I do think that the biggest news from this Apple event was that the sleepy chip industry has awaken itself from Intel reign of status quo, and now us consumers will have a lot more options and hopefully better devices for less money.
I love the fact Apple kept the pricing the same, as we get a much better product as a whole for the same money which is great. Also, I love Apple products and their approach to what consumer products should be, and excited to see what this Optimization of Hardware and Software could do to a laptop. hopefully the Mac will be the iPhone of the Laptop/Desktop industry.

 

[taldo]

It would mean the M1 is as fast as the 10 core iMac Pro -- and I still see reactions where people don't believe this "phone CPU" (as they refer to it) is faster than their desktop CPU.

If it was as simple as that, don't u think that Apple would have said that the fanless Macbook Air can outperform Intel's high end Desktop computer?..

There's a reason u compare a notebook processor to a notebook processor and a desktop processor to a desktop processor - they are totally different products even if the architecture is the same.

Edit:
in its event yesterday, Apple showed and talked about the performance per watt test, and that is the most important thing we need to discuss when comparing Intel and Apple chips, especially in the low power area of ultra-thin notebooks like the macbook air.

Apple claims for a 3X performance upgrade in the performance per watt test, which sounds awesome, but it's hard to say how did they reach this number, this is a footnote from Apple's website:
"Testing conducted by Apple in October 2020 using preproduction 13‑inch MacBook Pro systems with Apple M1 chip and 16GB of RAM, as well as previous‑generation Mac notebooks. Performance measured using select industry‑standard benchmarks. Performance tests are conducted using specific computer systems and reflect the approximate performance of MacBook Pro."

very unclear which processors they compared it too and on which configurations...

another example of Apple's great marketing - this is what Apple had to say when they released the new baseline Macbook pro 13 inch regarding the processor

"Better Performance​

The 13-inch MacBook Pro lineup now offers up to 10th-generation quad-core Intel Core processors with Turbo Boost speeds of up to 4.1GHz. Customers who are upgrading from a 13-inch MacBook Pro with a dual-core processor will see up to 2.8 times faster performance."

a 2.8X performance jump is quite easy when u double the Core count - that's why vague marketing declarations can be confusing and misleading (it's not misleading in terms of performance, but more about the reason for it or engineering abilities that caused this major performance boost)

The only thing that really matters in this debate is the value for money aspect - and i think that the fact that Apple is creating an In-House chip will finally give the product we pay for when we buy a mac.
999$ is not cheap, most baseline laptops cost almost half, and so far the air didn't really justify its price.
Finally today it's far superior from its PC competition and the 999$ price all of a sudden seems justified.

 

[Hexley]

Why couldn't they have given us proper specs....

Go here and scroll all the way down:

Mac

The most powerful Mac laptops and desktops ever. Supercharged by Apple silicon. MacBook Air, MacBook Pro, iMac, Mac mini, Mac Studio, and Mac Pro.

www.apple.com

Looks like MOST of their test benches were:

3.6GHz quad-core Intel Core i3-based (for all the other claims)
&
a production 1.2GHz quad-core Intel Core i7-based MacBook Air systems (RE: 3.9X faster video processing claim)

Yes most users, including myself for most cases an i3 is fine and the tradeoff for battery life, etc... def is intriguing...

Like I wonder if they just compared it horizontally M1 to i5 (which is mostly what they were selling, then the gains would have been even?) Is 2x an i3 = i5 power but less power draw?

Having said that, I need to upgrade the ol 2014 13" MBP and thinking about going with an air as my choice. I'm Mac Mini eGPU pro now, so don't need crazy portable power.

It probably has more to do with the price points where Core i3 occupied then was replaced by M1.

After seeing all the reviews of M1 Macs... do you have any complaints about the product?

eGPu support may occur on a later update to macOS.

 

[Luposian]

It would mean the M1 is as fast as the 10 core iMac Pro -- and I still see reactions where people don't believe this "phone CPU" (as they refer to it) is faster than their desktop CPU.

When I first found out that Apple was making ARM Macs, I was like, “Yeah... we’ll see how that works out for them.” I couldn’t fathom how a chip used in iPhones and iPads could play on par with (much less exceed) Intel Macs, even though my friend kept saying how later iOS chips were outperforming even other PC’s! So this is definitely a pleasant wake up call to me!

 

[Icelus]

When I first found out that Apple was making ARM Macs, I was like, “Yeah... we’ll see how that works out for them.” I couldn’t fathom how a chip used in iPhones and iPads could play on par with (much less exceed) Intel Macs, even though my friend kept saying how later iOS chips were outperforming even other PC’s! So this is definitely a pleasant wake up call to me!

Yes, and not strange at all as it was a gradual improvement over the years. The CPU used in the iPad 2 was a very efficient CPU at the time and with the A7 Apple released the first mobile 64-bit CPU. This was something the industry didn't expect at the time and a year earlier than other manufacturers (even a year before ARM!). With the A10 it was clear they approached desktop speeds but most people (still) didn't believe the benchmarks.