Intel Graphics 24EU Mobile vs NVIDIA L20 Comparison

Head-to-Head Benchmarks

The Intel Graphics 24EU Mobile and NVIDIA L20 occupy vastly different positions in the database. The Intel part is an integrated graphics solution with no recorded benchmark scores, while the NVIDIA L20 has two substantial entries. In Geekbench OpenCL, the L20 scores 274276 points. In Geekbench Vulkan, it records 228018 points. The average benchmark score for the L20 is 251147.

Comparing the L20 to its nearest rivals in the database, it sits 11.6% ahead of the NVIDIA PG506-232 in average score, which posts 225124 points. It also leads the AMD Radeon PRO W7900D by 14.2%, as that card averages 219827 points. Conversely, the L20 trails the NVIDIA L40 by 11.6%; the L40 averages 284111 points. The NVIDIA RTX 6000 Ada Generation is 12.6% ahead, with an average of 287237 points.

For the Intel side, there are no benchmark entries. Its percentile rank among all GPUs is 50, but with a zero average score, the data does not support any head-to-head numeric comparison. The wins tally in the database is 0 for Intel and 0 for NVIDIA, which reflects the absence of direct paired test results.

The L20's 99th percentile ranking places it in the top tier of all recorded GPUs. The Intel part, at the 50th percentile, sits at the median of the distribution, but this percentile is unverified by any actual benchmark run in the database.

Where Each One Wins

The NVIDIA L20 wins decisively in every measured category. Its compute throughput is recorded at 59.35 TFLOPS for FP32 and the same 59.35 TFLOPS for FP16 with a 1:1 ratio. The Intel part delivers 384.0 GFLOPS in FP32 and 768.0 GFLOPS in FP16 with a 2:1 ratio. This means the L20 offers over 150 times the FP32 throughput in raw terms, though the database does not provide a direct percentage delta between the two.

Texture and pixel processing follow the same pattern. The L20 reaches 927.4 GTexel/s and 322.6 GPixel/s. The Intel Graphics 24EU Mobile manages 12.00 GTexel/s and 4.000 GPixel/s. The L20 is roughly 77 times faster in texture rate and 80 times faster in pixel rate, calculated from the recorded figures.

Memory capacity is another clear differentiator. The L20 includes 48 GB of GDDR6 on a 384-bit bus with 864.0 GB/s bandwidth. The Intel part uses system shared memory with system dependent bandwidth and a system shared bus width. No fixed capacity or bandwidth figures exist for the Intel side.

Ray tracing and tensor hardware exist only on the L20. It has 92 RT cores and 368 tensor cores. The Intel part lists no RT or tensor core counts. For workloads using those features, the L20 is the only option between the two.

The L20 also supports DirectX 12 Ultimate (12_2), while the Intel part supports DirectX 12 (12_1). Both support OpenGL 4.6 and Vulkan 1.4, so those API levels are equivalent.

Architecture Differences

The Intel Graphics 24EU Mobile is built on Intel's 10 nm process and uses the Xe-LP architecture. Its chip is Twin Lake, part of the HD Graphics-T (Twin Lake) generation. The L20 uses the Ada Lovelace architecture on a 5 nm process from TSMC, with the AD102 chip in the Server Ada (Lxx) generation.

Foundry details differ. Intel fabricates its chip in-house. The L20 is built by TSMC. The L20's transistor count is recorded at 76,300 million, with a die size of 609 mm² and a transistor density of 125.3M per mm². The Intel part has unknown transistor and die size figures.

The L20 is a discrete card with a dual-slot design, 267 mm in length and 111 mm in height, using a PCIe 4.0 x16 bus interface and a single 16-pin power connector. The Intel part is an integrated graphics processor (IGP) with a ring bus interface and no power connectors listed. Its dimensions are not recorded.

The L20 draws 275 W TDP and has a suggested PSU of 600 W. The Intel part has a 6 W TDP. Clock behavior also diverges: the L20 has a 1440 MHz base and 2520 MHz boost, while the Intel part runs at 300 MHz base and 1000 MHz boost. The L20's memory clock is 2250 MHz with 18 Gbps effective, whereas the Intel part uses system shared memory.

Display outputs differ as well. The L20 provides 4x DisplayPort 1.4a. The Intel part's outputs are portable device dependent.

Release timing is separated by roughly a year. The L20 was released on 2023-11-15. The Intel part followed on 2024-12-31. Both are currently marked as Active in production status.

FAQ

Q: What is the average benchmark score for each GPU?

A: The NVIDIA L20 has an average benchmark score of 251147. The Intel Graphics 24EU Mobile has an average benchmark score of 0, as no benchmark entries exist for it.

Q: How does the L20 compare to the NVIDIA PG506-232?

A: The L20's average score is 11.6% higher than the PG506-232's average of 225124 points.

Q: Which GPU has more shading units?

A: The NVIDIA L20 has 11776 shading units. The Intel Graphics 24EU Mobile has 192 shading units.

Q: What memory configurations are recorded?

A: The L20 has 48 GB of GDDR6 on a 384-bit bus with 864.0 GB/s bandwidth. The Intel part uses system shared memory with system dependent bandwidth.

Q: Are ray tracing cores present in both GPUs?

A: No. The NVIDIA L20 has 92 RT cores. The Intel Graphics 24EU Mobile lists no RT cores.

Q: What DirectX versions do the two GPUs support?

A: The L20 supports DirectX 12 Ultimate (12_2). The Intel part supports DirectX 12 (12_1).

The Verdict

The database shows a complete mismatch in capability. The NVIDIA L20 is a server-class accelerator with 48 GB of GDDR6, 59.35 TFLOPS of FP32 performance, and support for ray tracing and tensor operations. It ranks in the 99th percentile of all GPUs. The Intel Graphics 24EU Mobile is a 6 W integrated solution with no recorded benchmark scores and a 50th percentile rank that is not backed by any test entry.

For compute-heavy workloads, AI inference, or graphics rendering that uses ray tracing, the L20 is the only viable choice between these two. Its 11776 shading units, 368 TMUs, and 128 ROPs provide the raw resources, and its 864.0 GB/s bandwidth supports large data sets. The Intel part, with 192 shading units, 12 TMUs, and 4 ROPs, is suited only to basic display output or light workloads on a portable device.

The L20's nearest rival data confirms it is a strong mid-to-high tier server GPU, sitting ahead of the PG506-232 and Radeon PRO W7900D, and only behind the L40 and RTX 6000 Ada Generation. The Intel part has no rival data to interpret.

For anyone choosing based strictly on the recorded measurements, the NVIDIA L20 delivers in every measurable category. The Intel Graphics 24EU Mobile offers no benchmark data and no fixed memory bandwidth, so its performance cannot be quantified from the database. The choice depends entirely on whether the workload requires a discrete accelerator or an integrated part with minimal power draw.

Specification Differences

| Specification | Intel Graphics 24EU Mobile | NVIDIA L20 |

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

| Process node | 10 nm | 5 nm |

| Foundry | Intel | TSMC |

| Architecture | Xe-LP | Ada Lovelace |

| Base clock | 300 MHz | 1440 MHz |

| Boost clock | 1000 MHz | 2520 MHz |

| Memory size | System Shared | 48 GB |

| Memory type | System Shared | GDDR6 |

| Memory bus width | System Shared | 384 bit |

| Memory bandwidth | System Dependent | 864.0 GB/s |

| Shading units | 192 | 11776 |

| TMUs | 12 | 368 |

| ROPs | 4 | 128 |

| RT cores | None listed | 92 |

| Tensor cores | None listed | 368 |

| Pixel rate | 4.000 GPixel/s | 322.6 GPixel/s |

| Texture rate | 12.00 GTexel/s | 927.4 GTexel/s |

| FP32 performance | 384.0 GFLOPS | 59.35 TFLOPS |

| FP16 performance | 768.0 GFLOPS (2:1) | 59.35 TFLOPS (1:1) |

| TDP | 6 W | 275 W |

| Slot width | IGP | Dual-slot |

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

| Suggested PSU | None | 600 W |

| Bus interface | Ring Bus | PCIe 4.0 x16 |

| Display outputs | Portable Device Dependent | 4x DisplayPort 1.4a |

| DirectX version | 12 (12_1) | 12 Ultimate (12_2) |

| Release date | 2024-12-31 | 2023-11-15 |

| Production status | Active | Active |