M3 Max chips being tested for future MacBook Pro models
Apple has begun testing the M3 Max processor with a 16-core CPU for 2024 MacBook Pro models.

MacBook Pro will get M3 Max in 2024
Apple revealed the M2 Pro and M2 Max-equipped MacBook Pros in January 2023, so news of the M3 models being tested isn't unexpected. Previous rumors point to a late 2024 release window for M3 Pro and M3 Max chips in updated MacBook Pro models.
According to a new report from Bloomberg's Mark Gurman, the M3 Max chips are being tested in an upcoming MacBook Pro model with 48GB of RAM. The news is sourced from a third-party Mac app developer's test logs.
The report states that the M3 Max will have a 16-core CPU and up to 40 GPU cores. The 16-core CPU breaks down to 12 high-performance cores and 4 efficiency cores.
That's a step up from the 10-core CPU and up to 38-core GPU in the current M2 Max.
The codename for the MacBook Pro showing up in testing with M3 Max is J514. The M3 series is expected to be built on the 3-nanometer process, offering increased power and efficiency.
The news arrives on the heels of M3 chips being tested in a Mac mini. The base M3 would have an 8-core CPU.
M3 could debut in the fall, given Apple's current chip release cycle -- M2 was revealed in June 2022 after M1 Ultra debuted in March 2022. But Gurman has pointed to 2024 for the M3 Mac mini and M3 Max MacBook Pro.
Read on AppleInsider
Comments
https://www.apple.com/shop/buy-mac/macbook-pro/16-inch-space-gray-apple-m2-max-with-12-core-cpu-and-38-core-gpu-1tb
I reckon M3 Max won't top out at 40-core GPU as that's very little improvement over the current 38-core. I think that would be the equivalent of the M2 Max with 30-core GPU:
https://www.apple.com/shop/buy-mac/macbook-pro/16-inch-space-gray-apple-m2-pro-with-12-core-cpu-and-19-core-gpu-1tb
M3 Max would more likely have 2 options: 16-core CPU, 40-core GPU, 16-core CPU, 50-core GPU ($200 upgrade).
This would be 50% more CPU performance cores and 30% more GPU cores, combined with 15% performance gain from N3 so 70% CPU boost, 50% GPU boost over M2. All while consuming less power. This should be a decent upgrade but I definitely don't expect Pro models before March 2024 (January at the earliest) so if people need a Pro M-series upgrade, M2 Pro/Max would be worth getting now. Entry level (Air) would be best to wait for M3 in October.
Fall 2023 - M3 iMac / MBA (13" & 15") / 13" MBP
Late Jan 2024 - M3 Pro / Max MBP / M3 Mac mini / M3 Pro Mac mini
Spring 2024 - M3 iPad Pros
"Apple testing M3 for iMac"
"Apple testing M3 Max for MacbookPro"
But between the M3 die and the M3 Max die that covers all the Mac product and a few iPads. What with Ultra and Pro M3's being based on the Max die.
M1 11/20
(11 months)
M1 Pro/Max 10/21
(5 months)
M1 Ultra 03/22
(3 months)
M2 06/22
(7 months)
M2 Pro/Max 01/23
(5 months)
M2 Ultra 06/23
Following that cycle:
M3, 3 months after M2 Ultra = 09/23, likely 10/23 as Apple usually has the Mac events in October.
M3 Pro/Max, 5-7 months after M3 = 03/24-05/24.
M3 Ultra, 5 months after M3 Pro/Max = 08/24-10/24.
Given TSMC N2 isn't until late 2025/early 2026, it's not clear what Apple will use in 2025, perhaps some 3D stacking or N3P.
M4, 01/25
M4 Pro/Max, 06/25
M4 Ultra, 11/25
M5 (N2), 02/26
...
M3 Max should be roughly double the performance of M1 Max at the same power so when doing things like rendering or video encoding, it will either run twice as fast if it uses all the available hardware or it will run at half the power, which means less chance for the fans to spin up even when doing heavy tasks and longer battery life. It's not worth upgrading a single generation but every second generation is worthwhile.
For 3D rendering on the following site, page 3, M2 Max managed to perform better than M1 Ultra, 90% faster than M1 Max:
https://opendata.blender.org/benchmarks/query/?compute_type=OPTIX&compute_type=CUDA&compute_type=HIP&compute_type=METAL&compute_type=ONEAPI&group_by=device_name&blender_version=3.6.0
That's only on specific tasks though, most improvements per generation run around 20-40%. Usually you'd notice 2x improvement and that needs 4 generations of 20% improvement or 2 generations of 40% improvement.
These are expensive laptops so that has to be factored in. 64GB M2 Max is around $4k. That's a lot to be spending every 3 years, even if the upgrade costs $2k after selling the old one. 5-6 year upgrade cycle would give a noticeable improvement about 3-4x increase in performance. That takes a Macbook Air up to the performance of the top Macbook Pro so someone who owned an M1 Max MBP (2021, 5nm) could get an M5 MBA (2026, 2nm) and get roughly the same performance in a fanless laptop.
For iPhone we expect Apple to limit 3 nm to the iPhone Pro due to expensive technology and limited manufacturing capacity.
For Mac we expect Apple to deliver 3 nm to low-end products. This makes no sense.
Apple needed to do low-end first for the transition. To get the tech ready and applications for Pro users. The same happened from PowerPC to Intel,
Now the transition is in place and Apple can once again introduce tech from the top.
https://www.statista.com/statistics/299153/apple-smartphone-shipments-worldwide/
Apple has to get nearly 80 million iPhone units ready for the Christmas period for iPhone but less than 5 million Macs. The number of high-end iPhones still outnumbers all their Mac shipments.
TSMC has to reduce defects in the manufacturing process for the low-end chips to get the best yield when they produce the Pro chips. The same happened with Intel, the entry i3 chips used the most advanced manufacturing and the Xeons were based on years old processes.
So... introduce iPad Pro with M3 and Air with A17 - but only do so if inventory is available? The iMac is a lost case, and the MBP 13 should die, but I see no urgency for upgrades to the low-end Mac products.
architectural advantage speaks volume.