NVIDIA GeForce RTX 4070 vs NVIDIA GeForce RTX 4090 Mobile Comparison

Architecture Differences

The NVIDIA GeForce RTX 4090 Mobile and NVIDIA GeForce RTX 4070 are both built on the Ada Lovelace architecture, using TSMC's 5 nm process, but they are fundamentally different chips with distinct configurations. The RTX 4090 Mobile uses the AD103 die, which contains 45,900 million transistors across a 379 mm² die, while the RTX 4070 uses the AD104 die with 35,800 million transistors on a 294 mm² die. Transistor density is nearly identical at 121.1M per mm² for the mobile part and 121.8M per mm² for the desktop part, confirming that both chips are produced on the same process node.

The compute resources diverge sharply. The RTX 4090 Mobile carries 9,728 shading units, 304 texture mapping units, and 112 ROPs, while the RTX 4070 has 5,888 shading units, 184 TMUs, and 64 ROPs. Ray tracing hardware follows the same pattern: the mobile flagship has 76 RT cores and 304 tensor cores, whereas the desktop RTX 4070 has 46 RT cores and 184 tensor cores. These differences translate directly into raw throughput: the RTX 4090 Mobile reaches 32.98 TFLOPS for both FP32 and FP16, while the RTX 4070 delivers 29.15 TFLOPS in both precisions.

Memory subsystems are also configured differently. The RTX 4090 Mobile uses 16 GB of GDDR6 memory on a 256-bit bus, yielding 576.0 GB/s of bandwidth. The RTX 4070 features 12 GB of GDDR6X memory on a 192-bit bus, producing 504.2 GB/s. Clock speeds tell another part of the story: the RTX 4090 Mobile runs at a 1335 MHz base and 1695 MHz boost, while the RTX 4070 operates at 1920 MHz base and 2475 MHz boost. The desktop card clocks substantially higher, which helps close the gap in some workloads despite having fewer shaders. The memory clocks also differ, with the mobile part at 2250 MHz (18 Gbps effective) and the desktop part at 1313 MHz (21 Gbps effective).

The physical and power profiles are vastly different. The RTX 4090 Mobile is an integrated graphics processor (IGP) with no power connectors and a 120 W TDP, while the RTX 4070 is a dual-slot card measuring 240 mm by 110 mm by 40 mm, requiring a single 16-pin connector and a 550 W suggested PSU. The RTX 4070 also provides fixed display outputs (1x HDMI 2.1 and 3x DisplayPort 1.4a), whereas the mobile part's outputs are portable device dependent. Both support PCIe 4.0 x16, DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. The release dates are close, with the RTX 4090 Mobile arriving on 2023-01-02 and the RTX 4070 on 2023-04-11. The production status differs as well: the mobile part is active, while the desktop part is end-of-life.

The Verdict

The data points to a clear split: the RTX 4090 Mobile wins in most GPU-centric workloads, but the RTX 4070 counters in specific areas, especially compute and 2D performance. Across the nine head-to-head benchmarks, the RTX 4090 Mobile takes five wins, while the RTX 4070 takes four. The average benchmark score puts the RTX 4090 Mobile at 43,667, which places it in the 84th percentile among all GPUs, while the RTX 4070 averages 37,648 and sits in the 81st percentile.

The RTX 4090 Mobile's largest victories are decisive. It leads by 24.5% in Passmark DirectX 10, by 16.8% in Geekbench OpenCL, and by 7.4% in Passmark DirectX 11. It also wins, more narrowly, in Passmark DirectX 12 (3.9%) and Passmark G3D (1.1%). The RTX 4070 strikes back with a 16.1% lead in Passmark GPU Compute, a 15.5% advantage in Passmark G2D, and a 1.9% margin in Geekbench Vulkan, plus a 3.1% win in Passmark DirectX 9.

For users prioritizing raw graphics throughput, especially in DirectX 10 and OpenCL workloads, the RTX 4090 Mobile is the stronger choice. Its 32.98 TFLOPS of FP32 performance, combined with 16 GB of memory and 576.0 GB/s of bandwidth, gives it a clear edge in shader-heavy and memory-intensive scenarios. However, the RTX 4070's compute advantage (14,720 vs 12,347 in Passmark GPU Compute) and its higher boost clock (2475 MHz vs 1695 MHz) make it the better option for tasks that respond well to clock speed and compute-focused architectures.

The RTX 4070 also offers the practical benefit of being a desktop card with fixed display outputs and a dual-slot form factor, while the RTX 4090 Mobile is an IGP whose outputs depend on the portable device. The RTX 4070 carries a launch MSRP of 599 USD, noted once here for reference. The RTX 4090 Mobile has no listed launch MSRP.

FAQ

Q: Which GPU has more shading units?

A: The NVIDIA GeForce RTX 4090 Mobile has 9,728 shading units, compared to 5,888 for the RTX 4070, a difference of 3,840 units.

Q: How much memory bandwidth does each card provide?

A: The RTX 4090 Mobile offers 576.0 GB/s over a 256-bit bus with 16 GB GDDR6, while the RTX 4070 provides 504.2 GB/s over a 192-bit bus with 12 GB GDDR6X.

Q: Which GPU wins in Geekbench OpenCL?

A: The RTX 4090 Mobile scores 180,831 versus 154,858 for the RTX 4070, a 16.8% advantage for the mobile part.

Q: What about Geekbench Vulkan performance?

A: The RTX 4070 wins that test, scoring 174,152 against 170,774, a 1.9% margin for the desktop card.

Q: Which GPU has the higher boost clock?

A: The RTX 4070 boosts to 2475 MHz, well above the RTX 4090 Mobile's 1695 MHz boost clock. The desktop card also has a higher base clock at 1920 MHz versus 1335 MHz.

Q: What is the power draw difference?

A: The RTX 4090 Mobile has a 120 W TDP and requires no power connectors, while the RTX 4070 has a 200 W TDP and needs a single 16-pin connector with a suggested 550 W PSU.

Q: In which benchmark does the RTX 4070 have its largest win?

A: The RTX 4070 leads by 16.1% in Passmark GPU Compute, scoring 14,720 versus 12,347 for the RTX 4090 Mobile.

Specification Differences

| Specification | NVIDIA GeForce RTX 4090 Mobile | NVIDIA GeForce RTX 4070 |

|----------------|-------------------------------|------------------------|

| Chip | AD103 | AD104 |

| Transistors | 45,900 million | 35,800 million |

| Die Size | 379 mm² | 294 mm² |

| Transistor Density | 121.1M / mm² | 121.8M / mm² |

| Base Clock | 1335 MHz | 1920 MHz |

| Boost Clock | 1695 MHz | 2475 MHz |

| Memory Clock | 2250 MHz (18 Gbps effective) | 1313 MHz (21 Gbps effective) |

| Memory Size | 16 GB GDDR6 | 12 GB GDDR6X |

| Memory Bus Width | 256 bit | 192 bit |

| Memory Bandwidth | 576.0 GB/s | 504.2 GB/s |

| Shading Units | 9728 | 5888 |

| TMUs | 304 | 184 |

| ROPs | 112 | 64 |

| RT Cores | 76 | 46 |

| Tensor Cores | 304 | 184 |

| Pixel Rate | 189.8 GPixel/s | 158.4 GPixel/s |

| Texture Rate | 515.3 GTexel/s | 455.4 GTexel/s |

| FP32 / FP16 | 32.98 TFLOPS | 29.15 TFLOPS |

| TDP | 120 W | 200 W |

| Slot Width | IGP | Dual-slot |

| Power Connectors | None | 1x 16-pin |

| Suggested PSU | None | 550 W |

| Display Outputs | Portable Device Dependent | 1x HDMI 2.1, 3x DisplayPort 1.4a |

| Dimensions | Not specified | 240 mm x 110 mm x 40 mm |

| Production Status | Active | End-of-life |

| Release Date | 2023-01-02 | 2023-04-11 |

| Predecessor | GeForce 30 Mobile | GeForce 30 |

| Successor | GeForce 50 Mobile | GeForce 50 |

Head-to-Head Benchmarks

The RTX 4090 Mobile dominates in DirectX 10 rendering, where it scores 173 against the RTX 4070's 139, a 24.5% advantage. This is the single largest margin in the entire comparison. OpenCL performance also favors the mobile part substantially: 180,831 versus 154,858, a 16.8% gap. In DirectX 11, the RTX 4090 Mobile leads by 7.4% (262 vs 244), and in DirectX 12 it edges ahead by 3.9% (107 vs 103). The closest win for the RTX 4090 Mobile comes in Passmark G3D, where it scores 27,212 against 26,927, just 1.1% higher.

The RTX 4070's biggest victory is in Passmark GPU Compute, where it scores 14,720 versus 12,347, a 16.1% margin. This result is notable because it reverses the overall compute hierarchy suggested by FP32 TFLOPS: the RTX 4090 Mobile has higher theoretical FP32 throughput (32.98 TFLOPS) but falls behind in this specific compute test. The RTX 4070 also wins Passmark G2D by 15.5% (1,164 vs 984), indicating stronger 2D graphics performance. In Geekbench Vulkan, the desktop card leads 174,152 to 170,774, a 1.9% edge. The smallest win for the RTX 4070 is in Passmark DirectX 9: 320 versus 310, a 3.1% margin.

The overall average benchmark scores confirm the mobile part's higher standing: 43,667 for the RTX 4090 Mobile versus 37,648 for the RTX 4070. The nearest rivals in the database further contextualize these results. The RTX 4090 Mobile sits 0.8% above the NVIDIA Quadro M6000 (43,301), 0.9% above both the RTX 5050 Mobile (43,268) and the Quadro M6000 24 GB (43,262), and 0.9% below the RTX A6000 (44,075). The RTX 4070 is 0.1% above the Tesla P4 (37,628), 0.4% above the Radeon RX Vega 56 (37,507), 1.3% below the RTX 4080 Mobile (38,135), and 1.3% above the Radeon PRO W6400 (37,157).

Where Each One Wins

The RTX 4090 Mobile is the pick for users who need maximum graphics throughput in DirectX 10, DirectX 11, and OpenCL workloads. Its 24.5% DirectX 10 lead and 16.8% OpenCL advantage are the most pronounced differences in the dataset, and its higher pixel rate (189.8 GPixel/s) and texture rate (515.3 GTexel/s) support this profile. The 16 GB memory capacity and 576.0 GB/s bandwidth also make it the better choice for large textures and high-resolution rendering scenarios, where the extra 4 GB and 71.8 GB/s of bandwidth provide headroom.

The RTX 4070 wins in compute-heavy tasks, as evidenced by its 16.1% lead in Passmark GPU Compute. This result suggests that the desktop card's higher clocks (2475 MHz boost) and its specific compute scheduling give it an advantage in general-purpose GPU workloads, despite having fewer shaders and lower theoretical FP32. Its 15.5% win in Passmark G2D indicates superior 2D graphics performance, which matters for desktop productivity and non-3D rendering tasks. The narrow Vulkan win (1.9%) and DirectX 9 win (3.1%) round out a profile that favors lower-level and compute-oriented APIs.

For gaming and typical 3D workloads, the two cards are much closer than their specifications suggest. The RTX 4090 Mobile wins Passmark G3D by just 1.1%, and the DirectX 12 margin is only 3.9%. The RTX 4070's higher clock speeds compensate for its smaller chip, making it competitive in modern APIs. However, the RTX 4090 Mobile's superior memory bandwidth and larger frame buffer give it an edge in memory-bound scenarios, while the RTX 4070's compute advantage makes it the more versatile choice for mixed-use systems that handle both graphics and compute tasks.

In terms of physical integration, the RTX 4090 Mobile is designed for laptops and compact devices, with a 120 W TDP and no external power requirement. The RTX 4070, as a dual-slot desktop card with a 200 W TDP and a 550 W suggested PSU, requires a full desktop chassis. The RTX 4070 also provides standard display outputs, while the mobile part's outputs depend on the host device. The production status difference (active for the mobile part, end-of-life for the desktop part) suggests that the RTX 4090 Mobile remains in current production, while the RTX 4070 has been phased out.