Intel Core Ultra 5 236V vs Intel Xeon D-1746TER Comparison

The Intel Core Ultra 5 236V and Intel Xeon D-1746TER are fundamentally different processors aimed at opposite ends of the computing spectrum. The benchmark data is unambiguous: the Core Ultra 5 236V wins 14 of 17 head-to-head tests, often by massive margins, while the Xeon D-1746TER claims only 3 wins, all in specific server-oriented workloads. The Core Ultra 5 236V is the clear performance champion for general and single-threaded tasks, while the Xeon D-1746TER holds a narrow, specialized edge in integer-heavy and data-compression scenarios. Choosing between them depends entirely on whether the workload prioritizes raw throughput in parallel server tasks or the swift execution of typical client-side applications.

Where Each One Wins

The Core Ultra 5 236V dominates every single-threaded and latency-sensitive benchmark. Its passmark single-thread score of 3893 is more than double the Xeon’s 1785, a 118.1% advantage that makes it the obvious choice for responsive, interactive workloads. This edge extends across all Cinebench tests, where the Ultra 5 leads by a consistent 17.4-17.5% in both single and multi-core runs. The Xeon D-1746TER cannot compete in any scenario where a single core’s speed dictates performance, such as typical desktop applications, light coding, or OS-level responsiveness.

Conversely, the Xeon D-1746TER wins exactly where its server pedigree shows. It leads in passmark integer math by 28.8% (54482 vs 38765) and in data compression by 7.8% (191594 vs 176554). It also edges out the Ultra 5 in random string sorting by 8.9% (23732 vs 21628). These are classic server tasks: database operations, file compression, and data processing that scale with many threads and benefit from large shared caches. For a dedicated server or workstation handling such workloads, the Xeon’s 10 cores and 20 threads provide a meaningful throughput advantage over the Ultra 5’s 8 cores and 8 threads.

The remaining 14 benchmark wins belong to the Ultra 5, including especially large margins in floating-point math (61% ahead), physics (75% ahead), and prime number finding (151.5% ahead). The data shows a clear split: the Ultra 5 is for compute-heavy, varied, and single-threaded work, while the Xeon is for narrow, highly parallel integer and compression tasks.

Architecture Differences

The two CPUs are built on entirely different foundations. The Core Ultra 5 236V uses the Lunar Lake architecture on a 3 nm process from TSMC, while the Xeon D-1746TER uses the Ice Lake-D architecture on Intel’s 10 nm process. This process advantage is a primary driver of the Ultra 5’s superior efficiency and clock speeds. The Ultra 5 has a base clock of 2.10 GHz and boosts to 4.70 GHz, compared to the Xeon’s 2.00 GHz base and 3.10 GHz boost. The higher boost clock, combined with the newer process, explains most of the single-thread performance gap.

Core counts and threading also differ sharply. The Ultra 5 has 8 cores and 8 threads, while the Xeon has 10 cores and 20 threads. The Xeon’s Hyper-Threading capability (implied by 20 threads from 10 cores) is a significant architectural feature for parallel workloads, yet the benchmark data shows the Ultra 5 still wins multi-threaded Cinebench tests by 17.4%. Cache hierarchies diverge as well: the Ultra 5 has 192 KB L1 and 2.5 MB L2 per core, with 8 MB shared L3, whereas the Xeon has 80 KB L1 and 1.25 MB L2 per core, but a larger 15 MB shared L3. The Xeon’s larger L3 cache likely contributes to its wins in data compression and integer math, where large working sets benefit from more shared cache.

Memory support is another major divergence. The Xeon D-1746TER supports DDR4 memory with a triple-channel bus and a rated bandwidth of 64.0 GB/s, while the Ultra 5’s memory support is listed as dependent on the motherboard. The Xeon also supports ECC memory, a critical feature for server reliability that the Ultra 5 lacks. PCIe capabilities differ too: the Xeon offers Gen 4 with 16 lanes, while the Ultra 5 offers Gen 5 with only 4 lanes. The Xeon’s additional lanes make it suitable for more expansion cards, storage controllers, and network adapters, while the Ultra 5’s Gen 5 bandwidth may benefit a single fast device.

Head-to-Head Benchmarks

The most decisive victory for the Core Ultra 5 236V is in passmark single-thread, where it scores 3893 against the Xeon’s 1785, a 118.1% lead. This is a catastrophic gap for the Xeon and underscores the Ultra 5’s superiority in any task that cannot be parallelized. The Ultra 5 also wins passmark physics by 75% (1503 vs 859) and floating-point math by 61% (52774 vs 32772), indicating a vastly stronger FPU and better handling of scientific or graphics-related calculations.

In Cinebench R23, the Ultra 5 scores 15628 multi-core and 2206 single-core, versus the Xeon’s 13311 and 1879, respectively. Both are 17.4% improvements for the Ultra 5. This is notable because the Xeon has 10 cores and 20 threads, yet still loses multi-core rendering to the Ultra 5’s 8 cores and 8 threads. The newer architecture and higher clocks of the Ultra 5 overcome the Xeon’s thread-count advantage. The pattern repeats in Cinebench R20 and R15, with the Ultra 5 leading by 17.4% in all multi-core and single-core variants.

The Xeon’s wins are narrower but telling. In passmark integer math, the Xeon scores 54482 versus the Ultra 5’s 38765, a 28.8% advantage. This is the Xeon’s largest win and suggests its architecture is optimized for integer-heavy server code. In data compression, the Xeon scores 191594 versus 176554, a 7.8% lead, and in random string sorting it leads 23732 versus 21628, an 8.9% margin. These are consistent with a CPU designed for data processing and database workloads. However, the Ultra 5 counters with a 39.7% win in data encryption (13049 vs 9343) and a 22.4% win in extended instructions (15451 vs 12620), showing it is not weak in all server-like tasks.

FAQ

Q: Which CPU has better single-core performance?

A: The Intel Core Ultra 5 236V is dramatically ahead. Its passmark single-thread score is 3893, compared to the Xeon D-1746TER’s 1785, a 118.1% advantage. It also wins all Cinebench single-core tests by 17.4-17.5%.

Q: Which CPU is better for multi-threaded rendering?

A: The Core Ultra 5 236V wins every Cinebench multi-core test. In Cinebench R23, it scores 15628 versus the Xeon’s 13311, a 17.4% lead, despite having only 8 cores and 8 threads versus the Xeon’s 10 cores and 20 threads.

Q: Does the Xeon D-1746TER have any advantages?

A: Yes, it wins three benchmarks. It leads in passmark integer math by 28.8% (54482 vs 38765), data compression by 7.8% (191594 vs 176554), and random string sorting by 8.9% (23732 vs 21628). These point to server-oriented data tasks.

Q: Which CPU supports ECC memory?

A: Only the Intel Xeon D-1746TER supports ECC memory. The Core Ultra 5 236V does not, making the Xeon the appropriate choice for systems requiring error-correcting memory for data integrity.

Q: How do their memory channels compare?

A: The Xeon D-1746TER uses a triple-channel memory bus with DDR4 support and a rated bandwidth of 64.0 GB/s. The Core Ultra 5 236V uses a dual-channel bus, but its memory support is otherwise listed as dependent on the motherboard.

Q: Which CPU has more PCIe lanes?

A: The Xeon D-1746TER has 16 PCIe Gen 4 lanes, while the Core Ultra 5 236V has only 4 PCIe Gen 5 lanes. The Xeon offers more expansion capability, while the Ultra 5 offers faster Gen 5 bandwidth per lane.

The Verdict

The data points to a clear verdict: the Intel Core Ultra 5 236V is the superior processor for the vast majority of workloads. It wins 14 of 17 head-to-head benchmarks, including all Cinebench tests, all single-threaded tests, and heavy math workloads. Its 118.1% lead in single-thread performance and 61% lead in floating-point math make it the obvious choice for any user running general-purpose applications, development tools, or client-side software. The Ultra 5 also achieves this with a 17 W TDP versus the Xeon’s 67 W, making it far more power-efficient for mobile or compact systems.

The Intel Xeon D-1746TER is not without merit, but its advantages are narrow. A 28.8% lead in integer math and a 7.8% lead in data compression are significant for specific server tasks, and its support for ECC memory and 16 PCIe Gen 4 lanes make it suitable for reliable, expandable server infrastructure. However, its lower clock speeds and older 10 nm process leave it far behind in any latency-sensitive or single-threaded work. For a dedicated data-processing server where integer math and compression dominate, the Xeon is the correct pick. For anything else, the Core Ultra 5 236V is the definitive winner.

Specification Differences

| Specification | Intel Core Ultra 5 236V | Intel Xeon D-1746TER |

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

| Cores | 8 | 10 |

| Threads | 8 | 20 |

| Base Clock | 2.10 GHz | 2.00 GHz |

| Boost Clock | 4.70 GHz | 3.10 GHz |

| TDP | 17 W | 67 W |

| Socket | Intel BGA 2833 | Intel BGA 2227 |

| Architecture | Lunar Lake | Ice Lake |

| Process Node | 3 nm | 10 nm |

| Foundry | TSMC | Intel |

| L1 Cache | 192 KB (per core) | 80 KB (per core) |

| L2 Cache | 2.5 MB (per core) | 1.25 MB (per core) |

| L3 Cache | 8 MB (shared) | 15 MB (shared) |

| Memory Support | Depends on motherboard | DDR4 |

| Memory Bus | Dual-channel | Triple-channel |

| Memory Bandwidth | N/A | 64.0 GB/s |

| ECC Memory | No | Yes |

| PCIe | Gen 5, 4 Lanes | Gen 4, 16 Lanes |

| Integrated Graphics | Arc 130V | None |

| Market Segment | Mobile | Server/Workstation |

| Release Date | 2024-09-23 | 2022-02-23 |