Intel Core Ultra 9 285HX vs Intel Xeon 6517P Comparison

Head-to-Head Benchmarks

The benchmark data presents a clear split between these two processors, with the Intel Core Ultra 9 285HX winning 10 of the 17 head-to-head tests, while the Intel Xeon 6517P takes 7. The margins, however, tell a more nuanced story than the raw win count.

The Core Ultra 9 285HX dominates in floating-point math, scoring 194,998 against the Xeon's 127,497, a 52.9% advantage. This is its largest win in the entire comparison. The mobile chip also excels in data encryption, posting 48,567 versus 32,385, a 50% lead. These two results alone suggest the Core Ultra 9 has a significantly stronger compute pipeline for certain workloads.

Single-thread performance also favors the Core Ultra 9 in PassMark's test, with a score of 4,618 compared to 3,311, a 39.5% margin. The same chip wins the Cinebench R20 single-core test by 13.7% (2,856 vs 2,511), though the Xeon 6517P fights back in Cinebench R15 single-core with 602 versus 323.5, a 46.3% lead for the server part. This reversal between Cinebench versions is striking and indicates that the two chips respond very differently to the same workload family.

In multi-threaded Cinebench R23, the Xeon 6517P takes the crown with 42,352 versus 36,429.5, a 14% advantage. The Core Ultra 9 285HX counters in Cinebench R15 multi-core with 5,656.5 versus 4,268, a 32.5% win, and in Cinebench R20 multi-core with 20,236 versus 17,787, a 13.8% lead. The PassMark multi-thread test goes to the Core Ultra 9 with 56,902 versus 49,786, a 14.3% margin.

The Xeon 6517P shows strength in integer math, scoring 162,671 versus 155,076, a 4.7% win. It also wins in extended instructions (51,891 vs 49,148, a 5.3% lead) and data compression (653,338 vs 631,885, a 3.3% margin). The physics test goes to the Xeon by a substantial 21.9% (4,452 vs 3,476). The Core Ultra 9 285HX wins prime number finding by 37.3% (460 vs 335) and random string sorting by 14.4% (77,196 vs 67,480).

The average benchmark scores place the Core Ultra 9 at 76,155, which sits just 0.1% below the AMD Ryzen 9 8945HX and 0.5% above the AMD Ryzen 9 9950X3D. The Xeon 6517P averages 72,350, which is 0.1% below the Intel Xeon 6724P and 1.4% above the Intel Xeon Platinum 8270. The Core Ultra 9 also holds a 95th percentile ranking among all CPUs, while the Xeon sits at the 94th percentile.

Architecture Differences

The two processors come from entirely different design lineages. The Intel Core Ultra 9 285HX uses the Arrow Lake architecture, specifically the Arrow Lake-HX variant, built on a 3 nm process at TSMC. It packs 24 cores and 24 threads, meaning no hyperthreading is present. The Xeon 6517P uses Granite Rapids, built on Intel's 5 nm process, with 16 cores and 32 threads, enabling two threads per core.

The Core Ultra 9 features 192 KB of L1 cache per core, 3 MB of L2 per core, and 36 MB of shared L3. The Xeon has smaller per-core caches at 112 KB L1 and 2 MB L2, but a much larger 72 MB shared L3. This gives the Xeon a total cache capacity advantage despite having fewer cores.

Memory architecture differs dramatically. Both support DDR5, but the Core Ultra 9 uses dual-channel memory with a bandwidth of 102.4 GB/s. The Xeon 6517P employs eight-channel memory with a bandwidth of 409.6 GB/s, four times the throughput. Both support ECC memory.

PCIe connectivity also diverges sharply. The Core Ultra 9 provides Gen 5 with 20 lanes from the CPU, while the Xeon offers Gen 5 with 88 lanes. The mobile chip includes integrated Arc Xe-LPG graphics with 64 execution units, whereas the Xeon has no integrated graphics at all.

The Core Ultra 9 has an unlocked multiplier, making overclocking possible, while the Xeon is locked. The mobile part runs at a base clock of 2.80 GHz and boosts to 5.50 GHz, while the Xeon starts at 3.20 GHz and boosts to 4.20 GHz. The Core Ultra 9 has a 55 W TDP, while the Xeon consumes 190 W. The Core Ultra 9 uses Intel BGA 2114 socket, and the Xeon uses Intel Socket 4710. The Core Ultra 9 launched on January 12, 2025, with the Xeon following on February 23, 2025.

FAQ

Q: Which processor has higher single-thread performance in PassMark?

A: The Intel Core Ultra 9 285HX scores 4,618 in PassMark single-thread, which is 39.5% higher than the Xeon 6517P's 3,311.

Q: Does the Xeon 6517P ever beat the Core Ultra 9 in Cinebench?

A: Yes, the Xeon wins Cinebench R23 multi-core with 42,352 versus 36,429.5 (14% ahead) and Cinebench R15 single-core with 602 versus 323.5 (46.3% ahead).

Q: Which CPU has more memory bandwidth?

A: The Xeon 6517P supports eight-channel DDR5 with 409.6 GB/s bandwidth, while the Core Ultra 9 285HX uses dual-channel with 102.4 GB/s.

Q: What are the core and thread counts for each processor?

A: The Core Ultra 9 285HX has 24 cores and 24 threads. The Xeon 6517P has 16 cores and 32 threads.

Q: Which chip includes integrated graphics?

A: Only the Core Ultra 9 285HX has integrated graphics, featuring Arc Xe-LPG Graphics with 64 execution units. The Xeon 6517P has no integrated graphics.

Q: How do the average benchmark scores compare?

A: The Core Ultra 9 285HX averages 76,155 across all benchmarks, while the Xeon 6517P averages 72,350. The Core Ultra 9 ranks at the 95th percentile of all CPUs, and the Xeon at the 94th.

Specification Differences

| Specification | Intel Core Ultra 9 285HX | Intel Xeon 6517P |

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

| Cores | 24 | 16 |

| Threads | 24 | 32 |

| Base Clock | 2.80 GHz | 3.20 GHz |

| Boost Clock | 5.50 GHz | 4.20 GHz |

| TDP | 55 W | 190 W |

| Socket | Intel BGA 2114 | Intel Socket 4710 |

| Architecture | Arrow Lake | Granite Rapids |

| Process Node | 3 nm | 5 nm |

| Foundry | TSMC | Intel |

| L1 Cache | 192 KB (per core) | 112 KB (per core) |

| L2 Cache | 3 MB (per core) | 2 MB (per core) |

| L3 Cache | 36 MB (shared) | 72 MB (shared) |

| Memory Bus | Dual-channel | Eight-channel |

| Memory Bandwidth | 102.4 GB/s | 409.6 GB/s |

| PCIe | Gen 5, 20 Lanes | Gen 5, 88 Lanes |

| Integrated Graphics | Arc Xe-LPG 64EU | N/A |

| Market Segment | Mobile | Server/Workstation |

| Release Date | 2025-01-12 | 2025-02-23 |

| Unlocked Multiplier | Yes | No |

| Part Number | SRVFJ | SRVU4 |

The Xeon 6517P has a launch MSRP of $1195. The Core Ultra 9 285HX has no recorded launch MSRP in the database.

The Verdict

The data paints two distinct portraits. The Intel Core Ultra 9 285HX is the stronger all-around performer for compute-heavy tasks that benefit from high clock speeds and modern architecture. Its wins in floating-point math, encryption, single-thread PassMark, and multi-thread PassMark make it the better choice for general-purpose high-performance computing in a mobile form factor.

The Intel Xeon 6517P excels in specific server-oriented workloads. Its 14% lead in Cinebench R23 multi-core and 21.9% lead in physics indicate that certain parallel workloads favor its 32 threads and larger L3 cache. The eight-channel memory system with 409.6 GB/s bandwidth is a significant advantage for memory-bound server applications.

For users who need a mobile processor with high single-thread capability and strong floating-point performance, the Core Ultra 9 285HX is the clear choice. It also offers integrated graphics, an unlocked multiplier, and a much lower TDP of 55 W versus 190 W. For server deployments where memory bandwidth and multi-threaded throughput in specific workloads matter more, the Xeon 6517P provides the necessary infrastructure with 88 PCIe lanes and eight-channel memory.

The average benchmark score favors the Core Ultra 9 (76,155 vs 72,350), and it holds a higher percentile ranking at 95 versus 94. The Xeon's nearest rivals include the Intel Xeon 6724P at nearly identical performance, while the Core Ultra 9 sits close to the AMD Ryzen 9 8945HX and Ryzen 9 9950X3D.

Where Each One Wins

The Intel Core Ultra 9 285HX wins in all of these categories based on the benchmark data:

• Floating-point math: 52.9% ahead of the Xeon. This is its largest margin and indicates a strong FPU pipeline.
• Data encryption: 50% ahead, suggesting superior cryptographic acceleration.
• Single-thread PassMark: 39.5% ahead, making it better for lightly threaded applications.
• Prime number finding: 37.3% ahead, which correlates with integer-heavy math loops.
• Cinebench R15 multi-core: 32.5% ahead, showing strength in older rendering workloads.
• Cinebench R20 multi-core: 13.8% ahead, with a similar 13.7% lead in single-core.
• PassMark multi-thread: 14.3% ahead, reinforcing its overall multi-tasking capability.
• Random string sorting: 14.4% ahead, indicating better memory access patterns for sorting algorithms.

The Intel Xeon 6517P wins in these areas:

• Cinebench R23 multi-core: 14% ahead, making it better for modern rendering and 3D workloads.
• Cinebench R15 single-core: 46.3% ahead, which is unusual given the Core Ultra 9 wins other single-thread tests.
• Cinebench R23 single-core: 63.4% ahead, a massive margin that suggests architectural advantages in this specific test.
• Physics: 21.9% ahead, indicating better performance in physics simulation workloads.
• Integer math: 4.7% ahead, a modest but consistent lead.
• Extended instructions: 5.3% ahead, showing better SIMD and instruction set utilization.
• Data compression: 3.3% ahead, useful for file compression and archival tasks.

The Xeon 6517P also provides eight-channel memory and 88 PCIe lanes, which are not benchmark categories but are critical for server workloads requiring high I/O throughput. The Core Ultra 9 counters with integrated graphics, which the Xeon lacks entirely.