Intel Iris Xe MAX Graphics vs NVIDIA GeForce GTX 680 Comparison

Intel Iris Xe MAX Graphics and NVIDIA GeForce GTX 680 occupy opposite ends of the GPU spectrum in both era and design philosophy, yet their benchmark data places them in a surprisingly close overall contest. The GeForce GTX 680, a keystone of NVIDIA’s Kepler generation from early 2012, carries a significantly higher raw compute ceiling in OpenCL, while the Intel Iris Xe MAX, a 2020 integrated-class part built on the DG1 chip, counters with modern architecture traits and a drastically lower power footprint. The head-to-head data shows the GTX 680 winning the only directly compared benchmark, but the overall average scores and percentile rankings tell a nuanced story where the Intel part is far more competitive than its 25 W envelope suggests.

Head-to-Head Benchmarks

The sole directly comparable benchmark in the data is Geekbench OpenCL, where the NVIDIA GeForce GTX 680 posts a score of 16906 against the Intel Iris Xe MAX’s 14315. This yields a -15.3% delta for Intel, meaning the GTX 680 outperforms the Iris Xe MAX by roughly fifteen percentage points in this compute-oriented test. That is a meaningful gap in raw performance, but it is not a landslide. For context, the GTX 680’s OpenCL score sits 0.9% above the NVIDIA Tesla K10 (14029) and -0.9% below the NVIDIA GeForce GTX 1070 Ti (14277), placing it in a tight cluster of high-end cards from different generations. The Iris Xe MAX, meanwhile, is -0.3% off the AMD Radeon Vega 11 (14352) and 0.3% ahead of the GTX 1070 Ti in the same benchmark, which shows that Intel’s part is not merely scraping by—it is trading blows with dedicated discrete GPUs that consume far more power.

Looking beyond the single direct comparison, the overall average benchmark scores reinforce this closeness. The Iris Xe MAX has an average score of 14315 across all its benchmark runs, while the GTX 680 averages 14150 across its three recorded tests (OpenCL, Metal, and Vulkan). That places the Intel part 1.2% ahead of the GTX 680 in overall average, despite losing the OpenCL head-to-head. The GTX 680’s additional benchmarks—7897 in Geekbench Metal and 17646 in Geekbench Vulkan—indicate that its OpenCL result is not its peak; the Vulkan score is notably higher, suggesting architectural strengths in certain API paths. However, the Intel part has no recorded Metal or Vulkan scores in this data, so a full cross-API comparison is impossible. The percentile rankings are nearly identical: the Iris Xe MAX sits at the 56th percentile of all GPUs, while the GTX 680 sits at the 55th, meaning they are statistically adjacent in the broader performance distribution.

Where Each One Wins

The GTX 680 wins decisively in raw compute throughput where its larger silicon and higher shader count can flex. Its OpenCL score of 16906 is the highest single benchmark result in either GPU’s record, and its Vulkan score of 17646 suggests it scales well with modern low-level APIs when drivers cooperate. This makes it the stronger choice for compute-heavy workloads that leverage OpenCL or Vulkan, provided the application can tolerate its older feature set. The GTX 680 also holds clear advantages in memory bandwidth (192.3 GB/s vs 68.26 GB/s) and texture fill rate (135.4 GTexel/s vs 79.20 GTexel/s), which historically translates to better performance in texture-bound scenarios like high-resolution gaming or certain rendering tasks.

The Iris Xe MAX wins in efficiency and architectural modernity. Its 25 W TDP is a fraction of the GTX 680’s 195 W, and its 10 nm Intel process node versus the GTX 680’s 28 nm TSMC node means it achieves competitive compute with far less power draw and heat output. In the average benchmark score, it edges out the GTX 680 by 1.2%, and its OpenCL result is within 15.3% of a card that draws nearly eight times the power. The Intel part also supports DirectX 12 (12_1) and Vulkan 1.4, while the GTX 680 is limited to DirectX 12 (11_0) and Vulkan 1.2.175, giving the Iris Xe MAX a feature-level edge in newer API-based workloads. For scenarios constrained by power budgets, thermal limits, or physical space—where the GTX 680’s dual-slot, 256 mm length and 2x 6-pin connectors are impractical—the Iris Xe MAX is the clear winner.

Architecture Differences

The two GPUs are separated by nearly a decade of architectural evolution. The Intel Iris Xe MAX uses the DG1 chip built on Generation 12.1 architecture, manufactured on a 10 nm process by Intel. It has 768 shading units, 48 TMUs, and 24 ROPs, with a die size of 95 mm². Its memory subsystem consists of 4 GB of LPDDR4X on a 128-bit bus, delivering 68.26 GB/s of bandwidth. The GTX 680, in contrast, uses the GK104 chip on NVIDIA’s Kepler architecture, built on a 28 nm process by TSMC. It packs 1536 shading units, 128 TMUs, and 32 ROPs, on a much larger 294 mm² die with 3,540 million transistors and a transistor density of 12.0M / mm². Its memory is 2 GB of GDDR5 on a 256-bit bus, providing 192.3 GB/s of bandwidth.

Clock speeds tell a story of different design priorities. The Intel part has a base clock of 300 MHz and a boost of 1650 MHz, while the GTX 680 runs at a base of 1006 MHz and boosts to 1058 MHz. The GTX 680’s higher sustained clocks help its raw throughput, but the Intel part’s boost clock is nearly 60% higher, indicating a design tuned for short bursts of performance within a strict power envelope. The FP32 compute figures reflect this: the GTX 680 delivers 3.250 TFLOPS against the Iris Xe MAX’s 2.534 TFLOPS, a 28% advantage for NVIDIA. However, the Intel part has a dedicated FP16 path at 5.069 TFLOPS (2:1), which the GTX 680 lacks entirely (no FP16 data is recorded), making the Iris Xe MAX more suitable for mixed-precision workloads. The GTX 680 also has a higher pixel rate (33.86 GPixel/s vs 39.60 GPixel/s for Intel), but the Intel part wins texture rate due to its higher clock. The bus interface differs as well: the Iris Xe MAX uses PCIe 4.0 x8, while the GTX 680 uses PCIe 3.0 x16, though the older card’s wider interface does not compensate for its older protocol.

The Verdict

From the data, the choice is clear-cut based on your priorities. If you need maximum raw compute performance in OpenCL or Vulkan and have the power budget and physical space, the NVIDIA GeForce GTX 680 is the better pick. Its OpenCL score is 15.3% higher than the Iris Xe MAX, its Vulkan result of 17646 is the best single benchmark in this comparison, and its memory bandwidth and texture rate are multiples of the Intel part. It is an end-of-life product with a launch MSRP of 499 USD, but its performance class in the data justifies consideration for legacy compute tasks or older game titles that favor its architecture.

If your priority is efficiency, modern API support, or fitting a GPU into a power-constrained system, the Intel Iris Xe MAX is the data-backed choice. It matches the GTX 680’s overall average score within 1.2%, supports DirectX 12 (12_1) and Vulkan 1.4, and consumes 25 W versus 195 W. Its 4 GB of LPDDR4X memory doubles the GTX 680’s capacity, which helps in memory-hungry workloads despite lower bandwidth. The Intel part’s 56th percentile ranking versus the GTX 680’s 55th confirms that, on average, they are peers in performance—but the Intel part achieves that with a fraction of the power and no external power connectors. For a builder assembling a compact, low-power system, the Iris Xe MAX is the rational choice. For a desktop with a robust PSU and room for a dual-slot card, the GTX 680 offers higher peak performance in specific APIs.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The Intel Iris Xe MAX has an average benchmark score of 14315, which is 1.2% higher than the NVIDIA GeForce GTX 680’s average of 14150.

Q: How does the GTX 680 perform in the head-to-head OpenCL test?

A: The GTX 680 scores 16906 in Geekbench OpenCL, beating the Iris Xe MAX’s 14315 by 15.3%.

Q: What is the power consumption difference between the two cards?

A: The Intel Iris Xe MAX has a TDP of 25 W, while the NVIDIA GeForce GTX 680 has a TDP of 195 W. The Intel part also has a suggested PSU of 200 W, versus 450 W for the GTX 680.

Q: Does the Iris Xe MAX support newer graphics APIs than the GTX 680?

A: Yes. The Iris Xe MAX supports DirectX 12 (12_1) and Vulkan 1.4, while the GTX 680 supports DirectX 12 (11_0) and Vulkan 1.2.175.

Q: What are the memory sizes and types for each GPU?

A: The Iris Xe MAX has 4 GB of LPDDR4X on a 128-bit bus with 68.26 GB/s bandwidth. The GTX 680 has 2 GB of GDDR5 on a 256-bit bus with 192.3 GB/s bandwidth.

Q: Which GPU has a higher FP32 compute throughput?

A: The GTX 680 has 3.250 TFLOPS of FP32 compute, compared to the Iris Xe MAX’s 2.534 TFLOPS. However, the Iris Xe MAX has 5.069 TFLOPS of FP16 compute, which the GTX 680 does not have.

Specification Differences

| Specification | Intel Iris Xe MAX Graphics | NVIDIA GeForce GTX 680 |

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

| Architecture | Generation 12.1 | Kepler |

| Process Node | 10 nm | 28 nm |

| Foundry | Intel | TSMC |

| Die Size | 95 mm² | 294 mm² |

| Transistors | N/A | 3,540 million |

| Transistor Density | N/A | 12.0M / mm² |

| Base Clock | 300 MHz | 1006 MHz |

| Boost Clock | 1650 MHz | 1058 MHz |

| Memory Size | 4 GB | 2 GB |

| Memory Type | LPDDR4X | GDDR5 |

| Memory Bus Width | 128 bit | 256 bit |

| Memory Bandwidth | 68.26 GB/s | 192.3 GB/s |

| Shading Units | 768 | 1536 |

| TMUs | 48 | 128 |

| ROPs | 24 | 32 |

| Pixel Rate | 39.60 GPixel/s | 33.86 GPixel/s |

| Texture Rate | 79.20 GTexel/s | 135.4 GTexel/s |

| FP32 Performance | 2.534 TFLOPS | 3.250 TFLOPS |

| FP16 Performance | 5.069 TFLOPS (2:1) | N/A |

| TDP | 25 W | 195 W |

| Slot Width | IGP | Dual-slot |

| Power Connectors | N/A | 2x 6-pin |

| Suggested PSU | 200 W | 450 W |

| Bus Interface | PCIe 4.0 x8 | PCIe 3.0 x16 |

| Display Outputs | No outputs | 2x DVI, 1x HDMI 1.4a, 1x DisplayPort 1.2 |

| DirectX Support | 12 (12_1) | 12 (11_0) |

| Vulkan Support | 1.4 | 1.2.175 |

| Dimensions (LxHxW) | N/A | 256 mm x 111 mm x 38 mm (10.1 in x 4.4 in x 1.5 in) |

| Release Date | 2020-10-30 | 2012-03-21 |

| Launch MSRP | N/A | 499 USD |

| Predecessor | Graphics | GeForce 500 |

| Successor | Alchemist | GeForce 700 |