Apple M3 Max 40-Core GPU vs Nvidia RTX 1000 Ada Generation Laptop GPU vs Apple M3 Pro 14-Core GPU

Apple M3 Max 40-Core GPU

The Apple M3 Max 40-Core GPU is a self-designed graphics card in the Apple M3 SoC and according to Apple with forty cores. According to Apple, the GPU uses a new architecture and now offers dynamic caching, mesh shading and ray tracing acceleration.

The GPU can access up to 128 GB LPDDR5-6400 unified memory via a 512 bit memory bus (max 400 GBit/s). Thanks to dynamic caching, the GPU should only reserve as much memory as is required.

Since the SoC is manufactured in the current 3nm (probably N3B) at TSMC. According to powermetrics, the GPU requires up to 60 watts (chip incl. CPU total 78 watts).

Nvidia RTX 1000 Ada Generation Laptop GPU

The Nvidia RTX 1000 Ada Generation Laptop GPU, not to be confused with the A1000, P1000 or T1000, is a lower-end professional graphics card for use in laptops that sports 2,560 CUDA cores and 6 GB of GDDR6 VRAM. It would be fair to say that this is a GeForce RTX 4050 (Laptop) in disguise; consequently, both are powered by the AD107 chip and are fast enough to handle most games at 1080p with quality set to High. The product was launched in February 2024; it leverages TSMC's 5 nm process and the Ada Lovelace architecture. The Nvidia-recommended TGP range for the card is very wide at 35 W to 140 W leading to bizarre performance differences between different systems powered by what is supposed to be the same product.

Quadro series graphics cards ship with much different BIOS and drivers than GeForce cards and are targeted at professional users rather than gamers. Commercial product design, large-scale calculations, simulation, data mining, 24 x 7 operation, certified drivers - if any of this sounds familiar, then a Quadro card will make you happy.

Architecture and Features

Ada Lovelace brings a range of improvements over older graphics cards utilizing the outgoing Ampere architecture. It's not just a better manufacturing process and a higher number of CUDA cores that we have here; under-the-hood refinements are plentiful, including an immensely larger L2 cache, an optimized ray tracing routine (a different way to determine what is transparent and what isn't is used), and other changes. Naturally, these graphics cards can both encode and decode some of the most widely used video codecs, AVC, HEVC and AV1 included; they also support a host of proprietary Nvidia technologies, including Optimus and DLSS 3, and they can certainly be used for various AI applications.

The RTX 1000 Ada features 20 RT cores of the 3rd generation, 80 Tensor cores of the 4th generation and 2,560 CUDA cores. Increase those numbers by 20%, and you get the RTX 2000 Ada - as long as we pay no attention to clock speed differences, of course. Unlike costlier Ada Generation professional laptop graphics cards, the RTX 1000 comes with just 6 GB of non-ECC VRAM; the lack of error correction makes this card less suitable for super-important tasks and round-the-clock operation. The VRAM is just 96-bit wide, delivering a not-so-impressive bandwidth of ~192 GB/s.

The RTX 1000 Ada Generation makes use of the PCI-Express 4 protocol, just like Ampere-based cards did. 8K SUHD monitors are supported, however, DP 1.4a video outputs may prove to be a bottleneck down the line.

Performance

While we are yet to test a single laptop powered by an RTX 1000 Ada as of late February, we have plenty of performance data for the RTX 4050 Laptop. Based on that, we expect a run-of-the-mill RTX 2000 Ada to deliver:

• a Blender 3.3 Classroom CUDA score of around 54 seconds
• a 3DMark 11 GPU score of around 27,000 points
• around 50 fps in GTA V (1440p - Highest settings possible, 16x AF, 4x MSAA, FXAA)
• upwards of 30 fps in Cyberpunk 2077 (1440p - High settings, Ultra RT, "Quality" DLSS)

Nvidia's marketing materials mention "up to 12.1 TFLOPS" of performance, a downgrade compared to 14.5 TFLOPS delivered by the RTX 2000 Ada.

Your mileage may vary depending on how competent the cooling solution of your laptop is and how high the TGP power target of the RTX 1000 Ada is.

Power consumption

Nvidia no longer divides its laptop graphics cards into Max-Q and non-max-Q models. Instead, laptop makers are free to set the TGP according to their needs, and the range can sometimes be shockingly wide. This is exactly the case with the RTX 1000, as the lowest value recommended for it sits at just 35 W while the highest is 300% higher at 140 W (this most likely includes Dynamic Boost). The slowest system built around an RTX 1000 Ada can easily be half as fast as the fastest one.

Last but not the least, the improved 5 nm process (TSMC 4N) the RTX 1000 is built with makes for decent energy efficiency, as of early 2024.

Apple M3 Pro 14-Core GPU

The Apple M3 Pro 14-Core GPU is a self-designed graphics card in the Apple M3 Pro (with 11 CPU cores) and offers fourteen of the eighteen cores available on the chip. The graphics card uses a new architecture and now offers dynamic caching, mesh shading and ray tracing acceleration.

The GPU can access up to 24 GB LPDDR5-6400 unified memory via a 192-bit memory bus (max 150 GBit/s). Thanks to dynamic caching, the GPU only reserves as much memory as is required.

The entire chip is manufactured in the current 3nm process (probably N3B) at TSMC.