You are cherry-picking a yet unreleased gaming laptop that is basically designed to fit a desktop-class CPU. We don’t know how large and heavy blade 18 is going to be. It will probably offer over 150W of thermal dissipation to the CPU alone.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
M2 Pro and M2 Max will be disappointing?
- Thread starter Zest28
- Start date
- Sort by reaction score
I'm just here to read something while I eat my popcorn...
On topic: this topic is moot until the Intel chip (and M2 Pro/Max) are in the hands of users.
On topic: this topic is moot until the Intel chip (and M2 Pro/Max) are in the hands of users.
Most on this thread realize this, but it bears repeating:
When Apple designed their M-series chips they considered these two key points:
1) Laptop sales are much larger than desktop sales, and their most popular Mac, overall, is a small fanless laptop—the Air.
2) Most programs are single-threaded, and people rarely run more than a few programs at a time. Thus, for the overwhelming majority of use cases, when it comes to CPU performance, what matters is SC rather than MC speed.
Thus they designed a chip that allowed their laptops generally, and their most popular and least expensive laptop in particular, to have a CPU that effectively outperformed anything found in any laptop PC, for the overwhelming majority of applications...while being quiet and offering extraordinary battery life.
This was brilliant. The two areas where they fall short are top-end GPU performance, and top-end SC desktop speed. Their SC performance is close enough to the fastest x86 machines that it's probably not worth giving up their massive efficiency advantage to equalize that. But with the GPU there's a bigger gap, and it will be interesting to see what they do with the 3 nm M2's in Spring 2023.
When Apple designed their M-series chips they considered these two key points:
1) Laptop sales are much larger than desktop sales, and their most popular Mac, overall, is a small fanless laptop—the Air.
2) Most programs are single-threaded, and people rarely run more than a few programs at a time. Thus, for the overwhelming majority of use cases, when it comes to CPU performance, what matters is SC rather than MC speed.
Thus they designed a chip that allowed their laptops generally, and their most popular and least expensive laptop in particular, to have a CPU that effectively outperformed anything found in any laptop PC, for the overwhelming majority of applications...while being quiet and offering extraordinary battery life.
This was brilliant. The two areas where they fall short are top-end GPU performance, and top-end SC desktop speed. Their SC performance is close enough to the fastest x86 machines that it's probably not worth giving up their massive efficiency advantage to equalize that. But with the GPU there's a bigger gap, and it will be interesting to see what they do with the 3 nm M2's in Spring 2023.
Well I have an M2 MBA and without a fan and on battery power it is very close to the high end Intel gaming chip. I think that says enough.
I have been saying this for a while. A fan less laptop with serious performance as the M2 is a game changer. Until you actually use such a machine you don't know what you are missing. On top of that performance is super battery life and the same performance plugged in or not.
I like Intel. I am about to get a laptop with 12th Gen chip and Arc A370 4gb graphics. I am excited for Intel because they have a long way to go with process node refinement and the new big little architecture with tons of cores is already paying off. They still have room to make significant gains just on the process refinement alone.
So there is no need for dumping on Apple. Intel needs to get their process down to 4nm. Right now process 7 is still very close to 10nm. People are going crazy for m series to go from a advanced 5nm which is really close to 4nm down to 3nm. So if Intel just goes from 10nm to 5nm in one shot the performance lift and increased efficiency should be seriously impressive specially coupled with big/little.
It is pretty amazing that Intel has been able to catch up in raw performance with the current process even if it does draw a lot more power. 12th Gen chips are seriously good compared to 11th Gen but in terms of thermal output and efficiency they still have a long way to go.
My take on Intel when it comes to Apple is who cares?? Honestly Intel doesn't power Apple devices any more. And MacOS is so much more optimized for the hardware than windows is. Most Windows PC's have terrible driver support and the OS is not custom made for any device so a lot of these chips power is wasted. I think some windows devices like Surface devices show how much optimized devices matter. They are so snappy even with much poorer hardware than the competition.
So I think in the next 2-4 years if Intel can improve it's process are going to be very interesting.
I have been saying this for a while. A fan less laptop with serious performance as the M2 is a game changer. Until you actually use such a machine you don't know what you are missing. On top of that performance is super battery life and the same performance plugged in or not.
I like Intel. I am about to get a laptop with 12th Gen chip and Arc A370 4gb graphics. I am excited for Intel because they have a long way to go with process node refinement and the new big little architecture with tons of cores is already paying off. They still have room to make significant gains just on the process refinement alone.
So there is no need for dumping on Apple. Intel needs to get their process down to 4nm. Right now process 7 is still very close to 10nm. People are going crazy for m series to go from a advanced 5nm which is really close to 4nm down to 3nm. So if Intel just goes from 10nm to 5nm in one shot the performance lift and increased efficiency should be seriously impressive specially coupled with big/little.
It is pretty amazing that Intel has been able to catch up in raw performance with the current process even if it does draw a lot more power. 12th Gen chips are seriously good compared to 11th Gen but in terms of thermal output and efficiency they still have a long way to go.
My take on Intel when it comes to Apple is who cares?? Honestly Intel doesn't power Apple devices any more. And MacOS is so much more optimized for the hardware than windows is. Most Windows PC's have terrible driver support and the OS is not custom made for any device so a lot of these chips power is wasted. I think some windows devices like Surface devices show how much optimized devices matter. They are so snappy even with much poorer hardware than the competition.
So I think in the next 2-4 years if Intel can improve it's process are going to be very interesting.
Many are placing most of the performance advantage of Apple's AS SoC down to TSMC's process superiority. But with the recent release of AMD's Zen4 CPUs which is on TSMC' 5nm process still cannot catch up to Apple M1, let alone the M2, when it comes to performance per watt. So I doubt Intel will catch up to Apple in the mobile space unless they dramatically improve their u-arch to not burn so much energy doing work.
I didn't say they will ever catch up to Apple in terms of performance per watt because x86 vs. Arm pretty much means they can't.
However they just need to best AMD which seems entirely possible with the new architecture they are using.
Let's not go from one extreme (Apple sucks) to another extreme (Intel sucks ppw).
When it comes down to it as long as Intel can improve battery life and thermals significantly enough while continuing to deliver performance they will do fine and the major amount of space they have to move in terms of process refinement means they should be able to without too much difficulty.
Also more people care about maximum performance vs power per watt at the end of the day. Not saying I agree but that is the populous reality as it is now.
Agree with most of what you wrote except this part. Apple vs. Intel is important, because we seem to be in a period where (especially in the enterprise space) people are switching from PC's to Macs, and how Macs perform relative to PC's affects that. Macs having a higher market share in turn helps us as Mac users, since that affects the fraction of apps that run natively on Macs, and also the attention the developers give to optimizing their Mac ports (including vectorized instruction sets, etc.). Of course, I want Windows (and Intel/AMD) to remain robust as well, since we as consumers benefit when Apple has strong competitors.
Strongly disagree here. Most consumers are thin-and-light laptop buyers, which means they are looking for a combination of performance, quiet operation and long battery life (at the right price).
I personally, by contrast, am principally a desktop user who wants the maximum SC speed attainable with quiet operation. Then again, I recognize that I'm not most users.
Actually, IMHO, the group that cares about maximal performance are in the minority, but they are vocal. Most just want a capable, cool running mobile computer that last a long time on battery power. This is Apple's target market for Macs.
Not sure why you think what I said is going to extreme. In fact, my post is saying that many are looking at extreme only (i.e. pointing to Apple's use of TSMC latest manufacturing process advantage as the main source of performance per watt superiority.) Intel may have a chance to catch up to Apple, but they have to drastically pivot away from their existing design philosophy and solve their process issue. Engineering is always about balance and tradeoffs.
As I've been saying and thinking all along, Apple IMHO, is never interested to be top dog. They have an internal target to reach and they try their best to achieve it. If the result is they are top dog, they they will boast about it. Otherwise they will just point out their product's strength.
All I was saying is that most people want a fast computer and if they can only choose one metric they will choose performance.
Battery life and quiet operation are always important but if people have to choose most will choose speed unless they travel or use a Chrome book or tablet as their primary device.
Of course performance per watt is actually a really big deal but average consumers really only understand cores and frequency.
You are creating a false dichotomy here. Sure, if you give people a choice between a very fast (but power hungry) system and a very efficient( but slow) system, many will pick the fast power hungry one. But offer them a system that has 90-95% of real-world performance of the fast one while using 5-6 times less power and you'll get a very different result.
The simple fact that in the mobile space M1 still reigns supreme. It's not only as fast or faster than any other mobile chip currently on the market for the majority of workloads, but it also offers consistent high performance on and off the battery, a huge advantage over x86 based laptops that will instantly slow down the moment you remove them from the AC.
Of course, in the desktop space the situation is different.
I think they are interested in real world usability. Which means performance you can actually access in different situations, the battery life to carry you though the day without having to pack a charger and dedicated hardware to accelerate common energy-sipping workloads. Apple's approach is pragmatical and practical.
Yes and no. We're talking about an unreleased gaming laptop's benchmark against an unreleased productivity laptop. So in that vein its not cherry picking.
On the other hand, we're talking about a gaming laptop that cannot achieve those numbers while on battery. Using the laptop to its fullest potential, also means the fans will be like a 747 sitting on your desk, where as the MBP is whisper quiet.
Bonus usage is that I can place the MBP directly on my lap, where these gaming laptops cannot.
Overall, I'm a huge fan of Razer, they make great products that are very Mac like in design language. Yet with that said, they're not designed to be used on the go, they are certainly tied to the power cord.
Just recently, I took my razer to the dealership without the power cord so I good do a little work. This was supposed to be 20 minutes, but things happen. I had the laptop on the most aggressive power savings that the Razer Synapse app and windows would allow but after 3 hours, I was running our juice. Had I brought my MBP, I would not be scrambling to get my work done before the laptop died - and yes I regretted bringing the razer over the MBP.
One more point regrading the unreleased 18" Razer. Look at the price points of the 17" version, and you can that you're going to be paying 3,000 to 4,000 dollars and given its probably a high end premium product 4k is not unrealistic. The MBP offers more value for less
I don't think anyone made authoratiive claims about the performance of these upcoming MacBook Pros. The scores of the 18" Blade were being compared to currently available M2 machines (like the Air).
I know, and I never claimed it, but this very thread is about how poor the M2 Pro/Max is going to be, and so this very topic is about unreleased products.
Fairly apparent Apple was hoping they would be getting 3nm parts from TSMC by now, but things beyond Apple's control held that up...
Because of this, M2 on 5nm (4nm) is most likely a stop-gap; the OG M1 products were getting long-in-the-tooth, the iPad Pro tablets were due for a refresh, and DaVinci Resolve was coming to iPadOS so the Media Engine was needed on the low end...
The question will be, once 3nm products become a thing, will Apple do a refresh of the current M2 products with a 3nm variant of the 5nm (4nm) M2 SoC...?
I can 'hear' the complaints from current M2 product owners now...! ;^p
Because of this, M2 on 5nm (4nm) is most likely a stop-gap; the OG M1 products were getting long-in-the-tooth, the iPad Pro tablets were due for a refresh, and DaVinci Resolve was coming to iPadOS so the Media Engine was needed on the low end...
The question will be, once 3nm products become a thing, will Apple do a refresh of the current M2 products with a 3nm variant of the 5nm (4nm) M2 SoC...?
I can 'hear' the complaints from current M2 product owners now...! ;^p
That's a fine way to put it. My M1 MacBook Air is already my fastest and most powerful Mac on hand, probably because it's my only Apple Silicon Mac I currently have. Even with 8-core graphics and 16 GB of RAM, processing power and speed is equivalent to a multicore Intel Core i7 or even an i9, except it uses less energy and thus doesn't get as hot.
Actually, he was comparing the SC GB5 scores of the unreleased Raptor Lake mobile to those for the current 5 nm (N4P) M2, rather than the upcoming M2 MBP's, which are expected to be on 3 nm (N3). That change by itself should increase processing speed by 10%, giving an estimated 2100 GB5 SC, which exceeds what he presented for the i9-13900HX.
You guys have really trouble reading? The comparison is against the M2 Pro and M2 Max, it even says so in the thread title. And there is not a chance the M2 Pro and M2 Max will be on 3nm, as the iPhone chip will get 3nm first, not the Mac. Apple makes most of their money with the iPhone.
And people wondered why I didn't include the GPU in this thread as I knew how messed up this thread would become.
And people wondered why I didn't include the GPU in this thread as I knew how messed up this thread would become.
What M2 Pro and M2 Max? These chips literally don't exist.
Says who? You have industry sources?
Well sure it'll messed up if someone uses leaked benchmarks of unreleased hardware to argue that it's going to be faster than non-existing hardware...
Unlike Apple, Intel and AMD make big bucks selling powerful server chips. While performance per watt is important for all chips, for some applications (like server chips) absolute performance is at least as critical. Thus Intel and AMD design choices will inevitably be different from Apple choices. On a flips side, this may explain why Apple is having problems with releasing M-based Mac Pro.
I don't think this explains anything at all, with regards to the AS Mac Pro.
IMHO, Apple do not need the N3 process for a Mac Pro SoC solution. They already have the technological means (in the form of UltraFusion) to stitch together many GPU dies, each with their own memory controller and a gigantic CPU die, and that could probably rival anything on the market today for consumer/workstation solution.
The Mac Pro likely is not something that Apple feels warrant more attention that it already have been given.
On the contrary, perf/W is much more important in the datacenter than it is on the desktop. Intel's highest profit margin server chips run at much lower clock speeds than the desktop products they share core designs with, because this lets Intel run the cores in a much more efficient region of the power/frequency curve. Big Xeons typically run no faster than 3.5 GHz.
As for Apple's delays shipping the AS Mac Pro, it's a big stretch to attempt diagnosing specific engineering issues from the mere fact that there is a delay. The elephant in the room here is COVID; it's hard to imagine that it hasn't affected their timelines.
I wonder if there will not be a M2 Pro, M2 Max, and M2 Ultra.
Just a guess, but I think that the M2 might be a stop-gap until the 3-Nanometer chips come out, and the Pro, Max, and Ultra chips will be updated with what Apple will call the M3 using that 3-Nanometer process.
Maybe the even number M-chips will be reserved for a few entry-level updates, just a slight modification of the M-chip that came before it. The odd number M-chips will be a significant change in design and/or manufacturing process, and used for entry-level chips and for the Prosumer devices as well.
So, we might see a launch of the new redesign M3 Mac Mini and the AS Mac Pro with the successor of the M1 Ultra for the base model (M3 Ultra), and double the Ultra for the high tier, not sure of the name (M3 Max Pro Mac Pro Ultra is my guess).
After that, a 14" and 16" MBP and 27" iMac, with a M3 Pro and M3 Max, and a M3 for the 24" iMac.
Later on, a M3 Max and M3 Ultra for the Studio.
Probably the fall of 2024, the M4 will launch and will just be a minor update over the M3. The MBA and 24" iMac will get them.
Rinse and repeat.
I benchmarked a few of my Macs with 2nd, 3rd, and 4th gen i7 chips using Geekbench5 and compared it to my M1 Mac Mini. The M1 was definitely faster for multicore performance, over 100% faster for the 4th gen in my Late 2013 iMac.
The issue was when it came to real world SW encoding using Handbrake, the M1 didn't perform as well as it does in the Geekbench results. It was still faster than these old Intel Macs, but as the encode times would increase, the performance gap would close more and more.
Basically, I tested all the Macs listed above with the same MakeMKV rip, and all the same settings on Handbrake. I tried to keep everything similar the best I could to keep the results fair. I changed the encode settings on Handbrake to see how different settings would impact the performance difference between the Macs.
For example, I would encode H.264 on Medium on the Preset encoder option, and then ran it in Placebo (except the 4th gen i7). I would record the time and convert it into fps. I would also test H.265 on Medium, but not Placebo.
With H.264 on Medium, the M1 was exactly 100% faster than gen 3 i7 and 83% faster than the gen 4 i7 Intel.
With H.264 on Placebo, the M1 dropped to only 50% faster than gen 3 i7, (I didn't test the gen 4 i7 on Placebo).
With H.265 on Medium, the M1 was 86% faster than gen 3 i7 and only 26% faster than the gen 4 i7 Intel.
The 2nd gen i7 was also tested, but didn't include the results here in this post to keep things easier to understand.
I ran these tests a while ago, and got distracted, but my original plan was to keep testing the 3rd gen and 4th gen i7 iMacs a long with the M1 and see how long of an encode it would take to have the old Intel Macs out perform the M1. I think if I would run the H.265 Placebo test on the Late 2013 iMac with the 4th gen i7, it would out perform the M1.
With the Studio, I am sure the performance different would be significant from what I did with the M1, but I wonder with SW encoding if the Studio would suffer from the same issue that the M1 does, and have the performance drop with long encodes.
Just a guess, but I think that the M2 might be a stop-gap until the 3-Nanometer chips come out, and the Pro, Max, and Ultra chips will be updated with what Apple will call the M3 using that 3-Nanometer process.
Maybe the even number M-chips will be reserved for a few entry-level updates, just a slight modification of the M-chip that came before it. The odd number M-chips will be a significant change in design and/or manufacturing process, and used for entry-level chips and for the Prosumer devices as well.
So, we might see a launch of the new redesign M3 Mac Mini and the AS Mac Pro with the successor of the M1 Ultra for the base model (M3 Ultra), and double the Ultra for the high tier, not sure of the name (M3 Max Pro Mac Pro Ultra is my guess).
After that, a 14" and 16" MBP and 27" iMac, with a M3 Pro and M3 Max, and a M3 for the 24" iMac.
Later on, a M3 Max and M3 Ultra for the Studio.
Probably the fall of 2024, the M4 will launch and will just be a minor update over the M3. The MBA and 24" iMac will get them.
Rinse and repeat.
This is with the video toolbox, I am curious to how the Studio performs with SW encoding on Handbrake.
I benchmarked a few of my Macs with 2nd, 3rd, and 4th gen i7 chips using Geekbench5 and compared it to my M1 Mac Mini. The M1 was definitely faster for multicore performance, over 100% faster for the 4th gen in my Late 2013 iMac.
The issue was when it came to real world SW encoding using Handbrake, the M1 didn't perform as well as it does in the Geekbench results. It was still faster than these old Intel Macs, but as the encode times would increase, the performance gap would close more and more.
Basically, I tested all the Macs listed above with the same MakeMKV rip, and all the same settings on Handbrake. I tried to keep everything similar the best I could to keep the results fair. I changed the encode settings on Handbrake to see how different settings would impact the performance difference between the Macs.
For example, I would encode H.264 on Medium on the Preset encoder option, and then ran it in Placebo (except the 4th gen i7). I would record the time and convert it into fps. I would also test H.265 on Medium, but not Placebo.
With H.264 on Medium, the M1 was exactly 100% faster than gen 3 i7 and 83% faster than the gen 4 i7 Intel.
With H.264 on Placebo, the M1 dropped to only 50% faster than gen 3 i7, (I didn't test the gen 4 i7 on Placebo).
With H.265 on Medium, the M1 was 86% faster than gen 3 i7 and only 26% faster than the gen 4 i7 Intel.
The 2nd gen i7 was also tested, but didn't include the results here in this post to keep things easier to understand.
I ran these tests a while ago, and got distracted, but my original plan was to keep testing the 3rd gen and 4th gen i7 iMacs a long with the M1 and see how long of an encode it would take to have the old Intel Macs out perform the M1. I think if I would run the H.265 Placebo test on the Late 2013 iMac with the 4th gen i7, it would out perform the M1.
With the Studio, I am sure the performance different would be significant from what I did with the M1, but I wonder with SW encoding if the Studio would suffer from the same issue that the M1 does, and have the performance drop with long encodes.
You got it all backwards. Server chips are precisely about performance per watt. They trade peak performance for sustained multi core performance = many cores that run at lower frequencies to deliver the best aggregated performance for the given socket power. It’s the enthusiast/gamer chips that sacrifice power efficiency to reach higher clocks and this higher peak performance.
In fact, back in the day Apple had Intel to develop a custom series of workstation Xeon chips because regular Xeons had the performance of a wet noodle when it came to single-threaded workloads.