Apple 40-Core M5 Max vs NVIDIA RTX PRO 2000 Blackwell Full Specs

Summary

5,120 Shaders 1.62GHz	4,352 Shaders 1.95GHz
Shared Memory614.4GB/sLPDDR5X	16GB GDDR7288GB/s placeholder
Peak·FP32·FP16 16.59 TFLOPS	Peak·FP32·FP16 16.97 TFLOPS
Form Factor iGPU	Form Factor Soldered
TDP Shared	TDP 70W

Highlights

Peak OPS

Peak OPS FP32 FP16 Tensor FP4 FP4 Tensor Sparse Tensor FP8-16 FP8-16 Tensor Sparse Tensor FP8-32 FP8-32 Tensor Sparse Tensor FP16-16 FP16-16 Tensor Sparse Tensor FP16-32 FP16-32 Tensor Sparse FP64 BF16 Tensor BF16 BF16 Tensor Sparse Tensor TF32 Tensor INT8 Memory Bandwidth TDP

40-Core M5 Max33.18 TFLOPSFP8-16 Tensor

x1

RTX PRO 2000 Blackwell543.1 TFLOPSFP4 Tensor Sparse

x16.37

Show Top Score

Theoretical Performance

Clock Speed1620MHz UC·Boost·OC	Clock Speed1950MHz UC·Base·Boost·OC
Peak OPS 33.18 TFLOPSFP8-16 Tensor	Peak OPS 543.1 TFLOPSFP4 Tensor Sparse
-	Tensor FP4 271.6 TFLOPS FP4 Tensor Sparse 543.1 TFLOPS
Tensor FP8-16 33.18 TFLOPS FP8-16 Tensor Sparse - Tensor FP8-32 33.18 TFLOPS FP8-32 Tensor Sparse -	Tensor FP8-16 135.8 TFLOPS FP8-16 Tensor Sparse 271.6 TFLOPS Tensor FP8-32 67.89 TFLOPS FP8-32 Tensor Sparse 135.8 TFLOPS
Tensor FP16-16 16.59 TFLOPS FP16-16 Tensor Sparse - Tensor FP16-32 16.59 TFLOPS FP16-32 Tensor Sparse -	Tensor FP16-16 67.89 TFLOPS FP16-16 Tensor Sparse 135.8 TFLOPS Tensor FP16-32 33.95 TFLOPS FP16-32 Tensor Sparse 67.89 TFLOPS
-	BF16 16.97 TFLOPS Tensor BF16 33.95 TFLOPS BF16 Tensor Sparse 67.89 TFLOPS
Tensor TF32 -	Tensor TF32 16.97 TFLOPS
FP32 16.59 TFLOPS	FP32 16.97 TFLOPS
-	FP64 265.2 GFLOPS
-	Tensor INT8 135.8 TOPS
Ray -	Ray 51.48 TOPS
Pixel Rate 259.2 GPixel/s	Pixel Rate 124.8 GPixel/s
Texture Rate 518.4 GTexel/s	Texture Rate 265.2 GTexel/s

Core Config

Shaders 5,120 Shaders	Shaders 4,352 Shaders
TMUs 320 TMUs	TMUs 136 TMUs
ROPs 160 ROPs	ROPs 64 ROPs
Tensor Cores 40 T-Cores	Tensor Cores 136 T-Cores
RT Cores 40 RT-Cores	RT Cores 34 RT-Cores
EUs 640 EUs	SMs 34 SMs

Clocks

Base Clock -	Base Clock 790MHz
Boost Clock 1.62GHz	Boost Clock 1.95GHz

Cache

L2 Cache 8.2MB shared

L2 Cache 32.8MB shared

Memory

Shared MemoryLPDDR5X	16GB GDDR7 placeholder
Memory Bus 512-bit	Memory Bus 128-bit
Memory Speed 9.6GT/s	Memory Speed 18GT/s
Memory Bandwidth 614.4GB/s	Memory Bandwidth 288GB/s
ECC No	ECC Yes

Power & Thermals

TDP Shared

TDP 70W

Video Output

Multi-Monitor 4	Multi-Monitor 4
HDCP -	HDCP HDCP 2.3

Ports

-	4x DisplayPort 2.1

Encoder

Encoder Model 2x Apple Media Engine 3

Encoder Model NVENC 9

Decoder

Decoder Model 2x Apple Media Engine 3

Decoder Model NVDEC 6

Card

Form Factor iGPU	Form Factor Soldered
PCIe -	PCIe 1-Slots
-	Height 69 mm (2.72") Width 167 mm (6.58") Depth 20 mm (0.79")
Cooling Open-Air -	Cooling Blower 1x Fan

Chip

Manufacturer Apple	Manufacturer NVIDIA
Chip Designer Apple	Chip Designer NVIDIA
Architecture Apple Gen 8	Architecture Blackwell
Family M Series	Family RTX PRO Blackwell
Branding	Branding
Codename G17M	Codename NV196
Market Segment Desktop	Market Segment Workstation
Release Date Mar 3, 2026	Release Date Aug 11, 2025

Fabrication

Foundry TSMC	Foundry TSMC
Fabrication Node N3P	Fabrication Node 4NP
Die Size -	Die Size 183mm²
Transistor Count -	Transistor Count 23 Billion
Transistor Density -	Transistor Density 125.7 MTr/mm²

Images

No images available

No images available

Used In

CPUs M5 Max (18C40G)	CPUs -
Laptops MacBook Pro 14(M5 Max, Early 2026)MacBook Pro 16(M5 Max, Early 2026)	Laptops -

Competitors

Apple40-Core M5 Max vs GPU Apple32-Core M5 Max Apple40-Core M5 Max vs GPU Apple60-Core M3 Ultra Apple40-Core M5 Max vs GPU Apple20-Core M5 Pro Show 12 more

NVIDIARTX PRO 2000 Blackwell vs GPU NVIDIARTX PRO 4000 Blackwell SFF NVIDIARTX PRO 2000 Blackwell vs GPU NVIDIAGeForce RTX 5060 Laptop 80W NVIDIARTX PRO 2000 Blackwell vs GPU NVIDIAGeForce RTX 5060 Laptop 100W Show 12 more