Intel Graphics 24EU Mobile vs NVIDIA Rubin GPU Comparison

Head-to-Head Benchmarks

The database contains no direct benchmark scores for either the Intel Graphics 24EU Mobile or the NVIDIA Rubin GPU. The head-to-head comparison fields are empty, and both parts carry an average benchmark score of 0. This means there is no measured performance data to walk through win by win, and no percentile separation between the two parts can be derived from recorded results.

What the data does show is that both products sit at the 50th percentile against all GPUs in the database, but that percentile is assigned without any benchmark input. It reflects an unranked position rather than a measured performance tier. The Intel part has no nearest rivals listed, and the NVIDIA part has no nearest rivals listed, so there is no rival delta to reference either.

The absence of benchmark data is itself informative. The Intel Graphics 24EU Mobile is an integrated graphics processor with 192 shading units, 12 texture mapping units, and 4 raster operation units. Its compute throughput is recorded at 384.0 GFLOPS for FP32 and 768.0 GFLOPS for FP16 with a 2:1 ratio. The NVIDIA Rubin GPU is a server-class accelerator with 28,672 shading units, 896 texture mapping units, 24 raster operation units, and 896 tensor cores. Its FP32 throughput is recorded at 130.0 TFLOPS, and FP16 at 260.0 TFLOPS with a 2:1 ratio.

Without benchmark scores, the only quantitative comparison available is the raw specification gap. The NVIDIA part delivers approximately 338 times the FP32 throughput of the Intel part, based on the recorded figures of 130.0 TFLOPS versus 384.0 GFLOPS. That arithmetic follows directly from the data, but it is not a measured benchmark result. The database does not confirm how either part behaves in real workloads, so any statement about relative performance beyond the specification sheet remains unsupported.

Architecture Differences

The two parts come from different manufacturers, different foundries, and different design philosophies. The Intel Graphics 24EU Mobile uses the Xe-LP architecture, built on Intel's 10 nm process at Intel's own foundry. Its chip is codenamed Twin Lake, and it belongs to the HD Graphics-T (Twin Lake) generation. The NVIDIA Rubin GPU uses the Rubin architecture, built on TSMC's 3 nm process, with a chip codenamed GR100. It belongs to the Server Rubin (Rxx) generation and is the successor to Server Blackwell.

The process node gap is substantial: 10 nm versus 3 nm. That difference in manufacturing technology is paired with a massive difference in die size. The NVIDIA part has a die size of 1456 mm² and a transistor count of 336,000 million, which the database records as 336 billion. The Intel part lists transistor count as unknown and die size as unknown, so a direct density comparison is not possible for the Intel side. The NVIDIA part has an explicit transistor density of 230.8M per mm².

Memory architecture is entirely different. The Intel part uses System Shared memory, with a System Shared bus width and System Dependent bandwidth. Its memory clock is also System Shared. The NVIDIA part uses 288 GB of HBM4 memory on a 16384 bit bus, with a recorded bandwidth of 22.1 TB/s and a memory clock of 2695 MHz, noted as 10.8 Gbps effective.

The feature sets also diverge. The Intel part supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4. The NVIDIA part lists DirectX as N/A, OpenGL as N/A, and Vulkan as N/A. The NVIDIA part includes tensor cores, 896 of them, while the Intel part lists tensor cores as null. Neither part lists RT cores. The NVIDIA part has no display outputs, while the Intel part lists its display outputs as Portable Device Dependent.

Power and form factor differ sharply. The Intel part has a TDP of 6 W and uses an IGP slot width, with a Ring Bus interface. The NVIDIA part has a TDP of 2300 W, uses an SXM Module slot width, and requires a suggested PSU of 2700 W. Its bus interface is PCIe 6.0 x16. The Intel part has no power connectors listed and no suggested PSU, which matches its integrated nature. The NVIDIA part also has no power connectors listed, but the suggested PSU figure indicates a discrete module design.

The Verdict

The data supports only a specification-based verdict, not a performance-based one. For workloads that require massive compute throughput, high-bandwidth memory, and tensor operations, the NVIDIA Rubin GPU is the only part in this comparison that offers those capabilities. Its recorded figures include 130.0 TFLOPS FP32, 260.0 TFLOPS FP16, 22.1 TB/s memory bandwidth, and 896 tensor cores. It is built on a 3 nm process with 336,000 million transistors and a 1456 mm² die.

For integrated graphics duties in a portable device, the Intel Graphics 24EU Mobile is the only part that fits that role. It has a 6 W TDP, an IGP slot width, System Shared memory, and display outputs that depend on the portable device. It supports standard graphics APIs: DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4. The NVIDIA part lists no graphics API support and no display outputs, so it cannot serve as a display adapter in any conventional sense.

The recorded data does not include benchmark scores, so neither part can be declared a performance winner in any tested workload. The database shows zero wins for each part in the head-to-head fields. The only definitive statements are architectural: one is an integrated, low-power graphics solution; the other is a high-power server accelerator with no display path.

Users who need a GPU for a portable device should look at the Intel part, because the data shows it is an IGP with display outputs and API support. Users who need a server accelerator for compute-heavy tasks should look at the NVIDIA part, because the data shows it is an SXM module with tensor cores and enormous memory bandwidth. Neither part can substitute for the other based on the recorded specifications.

Specification Differences

The two parts differ in every major specification field except for a few null entries. The Intel Graphics 24EU Mobile uses a Twin Lake chip with Xe-LP architecture on a 10 nm process at Intel. The NVIDIA Rubin GPU uses a GR100 chip with Rubin architecture on a 3 nm process at TSMC.

Clock speeds: the Intel part runs at a base clock of 300 MHz and a boost clock of 1000 MHz, with a game clock listed as null. The NVIDIA part runs at a base clock of 700 MHz and a boost clock of 2267 MHz, with a game clock listed as null. Memory clocks differ: the Intel part lists System Shared, while the NVIDIA part lists 2695 MHz with 10.8 Gbps effective.

Memory capacity: the Intel part uses System Shared memory, while the NVIDIA part has 288 GB of HBM4. Memory type: System Shared versus HBM4. Bus width: System Shared versus 16384 bit. Bandwidth: System Dependent versus 22.1 TB/s.

Compute units: the Intel part has 192 shading units, 12 TMUs, and 4 ROPs. The NVIDIA part has 28,672 shading units, 896 TMUs, and 24 ROPs. Tensor cores: null on the Intel part, 896 on the NVIDIA part. RT cores are null on both.

Pixel rate: 4.000 GPixel/s on the Intel part, 54.41 GPixel/s on the NVIDIA part. Texture rate: 12.00 GTexel/s versus 2,031.2 GTexel/s. FP32: 384.0 GFLOPS versus 130.0 TFLOPS. FP16: 768.0 GFLOPS versus 260.0 TFLOPS, both with a 2:1 ratio.

TDP: 6 W versus 2300 W. Slot width: IGP versus SXM Module. Suggested PSU: null versus 2700 W. Bus interface: Ring Bus versus PCIe 6.0 x16. Display outputs: Portable Device Dependent versus No outputs. DirectX: 12 (12_1) versus N/A. OpenGL: 4.6 versus N/A. Vulkan: 1.4 versus N/A.

Release dates: the Intel part is dated 2024-12-31, and the NVIDIA part is dated 2025-12-31. Production status is Active for both. The Intel part has no predecessor or successor listed. The NVIDIA part has Server Blackwell as a predecessor and no successor. Transistors, die size, and dimensions are unknown or null on the Intel part; the NVIDIA part lists 336,000 million transistors, a 1456 mm² die, and a transistor density of 230.8M per mm².

FAQ

Q: Which GPU has higher FP32 compute throughput?

A: The NVIDIA Rubin GPU has an FP32 throughput of 130.0 TFLOPS, compared to the Intel Graphics 24EU Mobile's 384.0 GFLOPS.

Q: Can the NVIDIA Rubin GPU be used in a portable device with a display?

A: The data shows the NVIDIA part has no display outputs and lists graphics APIs as N/A, while the Intel part has display outputs described as Portable Device Dependent and supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4.

Q: What memory configuration does each GPU use?

A: The Intel part uses System Shared memory with System Dependent bandwidth. The NVIDIA part uses 288 GB of HBM4 memory on a 16384 bit bus with 22.1 TB/s bandwidth.

Q: How do the power requirements compare?

A: The Intel part has a TDP of 6 W and an IGP slot width. The NVIDIA part has a TDP of 2300 W, an SXM Module slot width, and a suggested PSU of 2700 W.

Q: Does either GPU include tensor cores?

A: The NVIDIA Rubin GPU includes 896 tensor cores. The Intel Graphics 24EU Mobile lists tensor cores as null.

Q: Which GPU is built on a smaller manufacturing process?

A: The NVIDIA Rubin GPU is built on a 3 nm process at TSMC. The Intel Graphics 24EU Mobile is built on a 10 nm process at Intel.

Where Each One Wins

The Intel Graphics 24EU Mobile wins in categories related to integration and portability. It has a 6 W TDP, which is dramatically lower than the 2300 W TDP of the NVIDIA part. It uses an IGP slot width, meaning it is designed to be built into a processor rather than installed as a discrete module. It has display outputs described as Portable Device Dependent, so it can drive a screen in a mobile system. It supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.4, which makes it usable for standard graphics workloads and API-level compatibility.

The NVIDIA Rubin GPU wins in categories related to raw compute and memory capacity. It delivers 130.0 TFLOPS FP32 and 260.0 TFLOPS FP16, compared to 384.0 GFLOPS and 768.0 GFLOPS on the Intel part. It has 288 GB of HBM4 memory with 22.1 TB/s bandwidth, while the Intel part depends on System Shared memory with System Dependent bandwidth. It includes 896 tensor cores, which the Intel part does not offer. Its pixel rate of 54.41 GPixel/s and texture rate of 2,031.2 GTexel/s far exceed the Intel part's 4.000 GPixel/s and 12.00 GTexel/s.

The transistor and die data favor the NVIDIA part as well. It has 336,000 million transistors on a 1456 mm² die, built on a 3 nm process. The Intel part has unknown transistor and die figures on a 10 nm process. The NVIDIA part also has a higher boost clock at 2267 MHz versus 1000 MHz on the Intel part.

The release timeline differs, with the Intel part dated 2024-12-31 and the NVIDIA part dated 2025-12-31. Both are marked Active in production status. The NVIDIA part has a predecessor in Server Blackwell, while the Intel part has no predecessor or successor recorded.

The practical split is clear from the data. For a portable system that needs integrated graphics, display output, and low power consumption, the Intel Graphics 24EU Mobile is the only viable option in this comparison. For a server installation that needs maximum compute throughput, tensor core acceleration, and massive memory bandwidth, the NVIDIA Rubin GPU is the only viable option. The two parts do not overlap in their intended use cases, and the recorded specifications reflect that separation.