-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
Apple M1 CPU
- Thread starter benparker
- Start date
lolwut... so your argument is proven to be false and all of a sudden the comparison *you* made isn't the best example? That's a new level of special.
Last edited:
could only be 4k8kW10
The Lenovo you linked to uses a Ryzen 4300U, has a TN 1080p screen (45% NTSC), that consumes much less than the P3 HiDPI IPS screen of Macbook Air. Like, not even close as on a similar resolution and brightness IPS screens use 25% more power on average, and the MBA is both higer res and brighter. The 0.7 inch screen size difference won't matter in this context. The 10% smaller battery also wouldn't explain the 50% difference in battery life (assuming your numbers are correct, I haven't checked them).
So basically an underpowered laptop with a much worse screen and only 10% smaller battery and the M1 MBA gets 50% more battery life than this? Doesn't that kind of prove the opposite of the point you're trying to make?
LinusTechTips tested M1 MBA and MBP against HP Envy 360 (Ryzen 4500U, 13.3-inch 1080p IPS screen, 51Wh battery). M1 MBA narrowly edged it out in their test, although their test was very simple. Just to give some context.
Last edited:
Caporegime
- Joined
- 9 Nov 2009
- Posts
- 25,353
- Location
- Planet Earth
Also the M1 is made on TSMC 5NM and the Ryzen is made on an earlier version of TSMC 7NM. It's a full process node ahead. It's one thing comparing the products but until AMD gets into 5NM with Zen4,then it will be a better comparison at the design level. It's like saying Zen3 has better power consumption than Cometlake with it is obvious as the process node plays a big part in it.
Agreed a bigger screen will consume more power as the backlight is what is the major factor here. A bigger screen requires a larger backlight array.
Last edited:
I don't think we'll even get a design-level comparison out of those, e.g. laptop Zen 4 is likely due in 2023, by then Apple will likely be another node ahead given their ability to outbid everyone else.
Not even remotely close. Don't want to drag this on but they're roughly equal size.
Macbook screen is 16:10 13.3-inches, so roughly 11.3 x 7.1 inches = ~ 80 square inches
Lenovo screen is 16:9 14-inches, so roughly 12.1 x 6.8 inches = ~ 83 square inches
Also:
Macbook: IPS, P3, 400nit, 2560x1600 (16:10)
Lenovo: TN, 45% NTSC, 250nit, 1920x1080 (16:9)
Like 2x pixels, almost 2x brighter, IPS and P3, and roughly equal surface area. The MBA screen is significantly more power hungry.
Apple A15 is out with new performance and efficiency cores. Last year, A14 and M1 shared their cores, just slightly different configurations (clocks, cache, etc). Very likely that we'll see the same thing again with M2 (if we see M2 this fall).
A15's performance cores are ~10% faster than A14's firestorm cores, but they run at 8% higher clocks, i.e. very little IPC gains. Some back of the envelope calculations (by me, not a professional reviewer) suggests they're also about ~20% more efficient at the same time. A15 likely uses TSMC's N5P node, which gives either the option of 7% higher clocks at the same power consumption, or 15% less power consumption at the same clocks compared to A14/M1's N5 node. Apparently Apple went for higher clocks, and their 20% improvement in efficiency comes from design improvements.
So the focus has been on efficiency this year, with the goal of either going for higher clocks, or packing more cores, or a combination of both. They went for higher clocks in A15, I'm hoping for more cores in M2.
Efficiency cores are also about ~20% faster than last year's IceStorm cores.
GPUs got a ~50% gain compared to last year's A14, which means M2 can get within the ranges of GTX 1080 in performance, with a bigger M2X being closer to GTX 1080 Ti, that's where most of the extra transistors went. But that's for the GPU section.
https://creativestrategies.com/iphone-13-and-apple-silicon/
A15's performance cores are ~10% faster than A14's firestorm cores, but they run at 8% higher clocks, i.e. very little IPC gains. Some back of the envelope calculations (by me, not a professional reviewer) suggests they're also about ~20% more efficient at the same time. A15 likely uses TSMC's N5P node, which gives either the option of 7% higher clocks at the same power consumption, or 15% less power consumption at the same clocks compared to A14/M1's N5 node. Apparently Apple went for higher clocks, and their 20% improvement in efficiency comes from design improvements.
So the focus has been on efficiency this year, with the goal of either going for higher clocks, or packing more cores, or a combination of both. They went for higher clocks in A15, I'm hoping for more cores in M2.
Efficiency cores are also about ~20% faster than last year's IceStorm cores.
GPUs got a ~50% gain compared to last year's A14, which means M2 can get within the ranges of GTX 1080 in performance, with a bigger M2X being closer to GTX 1080 Ti, that's where most of the extra transistors went. But that's for the GPU section.
https://creativestrategies.com/iphone-13-and-apple-silicon/
Well, not exactly. A14 was a cut-down M1, A15 is next gen uarch, which will be used in M2. We won't see 3nm until at least M3.
Yep it's pretty clear why Apple spanks everyone in battery life, single core power draw, which for most people is all they will use day to day, is orders of magnitude lower than any x86 processor on the market. The M1 pro/max pulls just 7 watts with single core at full load, while x86 laptops are pulling a 30 watts on average and some as high as 40 watts.
And even if you pull out the multithreaded workloads, the M1 is pro/max peaks at 43 watts while Intels laptops are going over 100 watts and AMD 70 watts.
And if you really want a crazy comparison, put the M1 max/pro against a top Intel gaming laptop and the Apple laptop pulls 40w in Cinebench multithread while the Intel pulls over 40w in single thread. These top gaming laptops can only last 1 or 2 hours on battery while the MacBook does 10
Last edited:
Interesting to see the difference between 7z compression (M1 Pro ~40% faster than 5800HX) and decompression (M1 Pro ~30% slower than 5800HX).
It is a well-known fact that compression works very well with out of offer execution, and decompression doesn't, and that's why the SMT in 5800HX makes a big difference for decompression, but didn't expect it to be this big (was expecting ~10% slower performance than 5800HX). I'm curious to see this on M1 Max to see if M1 Pro is being bottlenecked by memory bandwidth.