Intel Core Ultra 5 238V vs Intel Xeon D-1746TER Comparison

The Intel Core Ultra 5 238V is the clear performance leader in this matchup, winning 14 of the 17 head-to-head benchmarks and posting a higher average benchmark score. However, the Intel Xeon D-1746TER is not without its strengths, claiming decisive victories in integer math, data compression, and random string sorting, which points to a distinct workload split despite the overall score advantage for the mobile chip.

Head-to-Head Benchmarks

The Core Ultra 5 238V dominates the Cinebench suite with remarkable consistency. Across all six Cinebench tests—R15, R20, and R23, both single-core and multi-core—the Ultra 5 wins by the same margin of 14.9%. This uniform delta is striking; for instance, the Ultra 5 scores 2208 in Cinebench R23 single-core versus the Xeon’s 1879, and 15645 versus 13311 in R23 multi-core. The consistency suggests a fundamental architectural efficiency advantage rather than a workload-specific quirk.

The gap widens dramatically in PassMark’s single-threaded tests. The Ultra 5 posts a PassMark single-thread score of 3890, which is 54.1% higher than the Xeon’s 1785. This is the largest percentage delta in either direction across the entire benchmark set. The Ultra 5 also shows overwhelming superiority in floating-point math, scoring 53160 versus 32772, a 38.4% advantage, and in the find prime numbers test, where it scores 174 versus 68, a massive 60.9% lead. Physics simulation follows suit, with the Ultra 5 at 1546 versus 859, a 44.4% win.

The Xeon D-1746TER fights back in three specific PassMark workloads. Its most significant victory is in integer math, where it scores 54482 against the Ultra 5’s 38889, a 40.1% advantage. It also wins data compression with 191594 versus 176532, an 8.5% lead, and random string sorting with 23732 versus 21585, a 9.9% edge. These are not marginal wins; the integer math result in particular shows that the Xeon’s multi-threaded, server-oriented design handles certain calculation types far better than the newer mobile processor.

The overall average benchmark scores reflect a near-tie, with the Ultra 5 at 21981 and the Xeon at 21635. The Ultra 5’s percentile rank matches the Xeon at 75%, placing both in the same performance tier despite their different designs. The data encryption test shows the Ultra 5 ahead by 28.5% (13072 versus 9343), and extended instructions also favor the Ultra 5 by 17.9% (15377 versus 12620). The Xeon’s wins are concentrated and specific, while the Ultra 5’s wins are broad and often overwhelming.

FAQ

Q: Which processor has the higher single-core performance?

A: The Intel Core Ultra 5 238V wins every single-core benchmark, including a 54.1% lead in PassMark single-thread (3890 versus 1785) and a 14.9% lead in Cinebench R23 single-core (2208 versus 1879).

Q: Is the Xeon D-1746TER better at any compute task?

A: Yes, the data shows it wins PassMark integer math by 40.1% (54482 versus 38889), data compression by 8.5% (191594 versus 176532), and random string sorting by 9.9% (23732 versus 21585).

Q: How do the multi-core scores compare?

A: The Core Ultra 5 238V is ahead by 14.9% in Cinebench R23 multi-core (15645 versus 13311) and by the same 14.9% margin in PassMark multithread (18407 versus 15660).

Q: What is the performance difference in floating-point operations?

A: The Ultra 5 238V leads significantly, scoring 53160 versus 32772 in PassMark floating-point math, a 38.4% advantage.

Q: Do both processors occupy the same performance percentile?

A: Yes, both the Intel Xeon D-1746TER and the Intel Core Ultra 5 238V are in the 75th percentile when compared to all CPUs, despite very different benchmark profiles.

Q: Which chip has the higher average benchmark score?

A: The Core Ultra 5 238V has a slightly higher average benchmark score of 21981, compared to the Xeon D-1746TER’s 21635.

Architecture Differences

The two processors represent fundamentally different design philosophies from Intel. The Xeon D-1746TER is built on the Ice Lake architecture, specifically the Ice Lake-D variant, using a 10 nm process node manufactured by Intel. Its socket is the Intel BGA 2227, and it is designed for the Server/Workstation market segment. The Core Ultra 5 238V, in contrast, uses the Lunar Lake architecture on a 3 nm process node fabricated by TSMC, fitting into the Intel BGA 2833 socket and targeting the Mobile segment.

The core configurations diverge sharply. The Xeon offers 10 cores and 20 threads, while the Ultra 5 provides 8 cores and 8 threads, meaning the Xeon supports simultaneous multithreading and the Ultra 5 does not. The cache hierarchy also differs: the Xeon allocates 80 KB of L1 and 1.25 MB of L2 per core, with a 15 MB shared L3 cache. The Ultra 5 uses larger per-core caches at 192 KB L1 and 2.5 MB L2, but has a smaller 8 MB shared L3. This suggests the Xeon is designed for heavily threaded, cache-sensitive workloads, while the Ultra 5 relies on faster single-core execution and newer process technology.

Memory support marks another major architectural split. The Xeon uses DDR4 memory with a triple-channel memory bus and an ECC memory capability, which is a hallmark of server platforms. The Ultra 5’s memory support is listed as dependent on the motherboard, with a dual-channel bus and no ECC support. The Xeon also features a Gen 4 PCIe interface with 16 CPU lanes, whereas the Ultra 5 offers Gen 5 PCIe with only 4 CPU lanes, but includes integrated Arc 130V graphics, which the Xeon lacks entirely.

The release dates and production statuses are notable: the Xeon launched on 2022-02-23, while the Ultra 5 arrived later on 2024-09-23. Both are listed as Active in production. The Xeon has a launch MSRP of $1069, a figure that is not available for the Ultra 5. Neither processor has an unlocked multiplier, and their part numbers are SRM1B for the Xeon and SRPN5SRPN4 for the Ultra 5.

Specification Differences

The Core Ultra 5 238V has a higher base clock at 2.10 GHz versus the Xeon’s 2.00 GHz, and a much higher boost clock at 4.70 GHz versus 3.10 GHz. The Ultra 5’s TDP is drastically lower at 17 watts compared to the Xeon’s 67 watts, underscoring their different thermal envelopes. The Xeon has more cores (10 versus 8) and more threads (20 versus 8). The Xeon’s L3 cache is larger at 15 MB shared, while the Ultra 5’s is 8 MB shared.

The Xeon supports DDR4 memory with a triple-channel bus and a memory bandwidth of 64.0 GB/s, while the Ultra 5’s memory support is motherboard-dependent with a dual-channel bus and no listed bandwidth. ECC memory is supported only on the Xeon. The PCIe configurations differ: the Xeon has Gen 4 with 16 lanes, and the Ultra 5 has Gen 5 with 4 lanes. The Ultra 5 includes integrated Arc 130V graphics; the Xeon has no integrated graphics. The sockets are different (BGA 2227 versus BGA 2833), and the process nodes differ (10 nm Intel versus 3 nm TSMC). The Xeon carries a launch MSRP of $1069, which is not specified for the Ultra 5.

Where Each One Wins

The Intel Core Ultra 5 238V is the winner in nearly every general-purpose and single-threaded workload. Its 54.1% lead in PassMark single-thread makes it the choice for responsiveness and lightly threaded applications. The 38.4% advantage in floating-point math and 44.4% lead in physics simulation point to strong performance in scientific computing and simulation tasks. Its consistent 14.9% wins across all Cinebench versions indicate superior rendering and media-encoding capability. The 28.5% lead in data encryption also makes it suitable for security-focused tasks that rely on modern instruction sets.

The Intel Xeon D-1746TER wins in three specific, compute-heavy areas. Its 40.1% lead in integer math is its most significant victory, suggesting a strong fit for database operations, financial calculations, and other integer-heavy server tasks. The 8.5% win in data compression and 9.9% win in random string sorting further indicate that the Xeon excels at data manipulation and sorting workloads, which are common in server and data-center environments. The Xeon’s ECC memory support and triple-channel DDR4 also give it a reliability and memory-bandwidth edge for sustained server operations.

The Verdict

The data clearly favors the Intel Core Ultra 5 238V for most users. It wins 14 of 17 benchmarks, has a higher average score (21981 versus 21635), and delivers a massive single-thread performance advantage of 54.1%. Its lower TDP of 17 watts, higher boost clock of 4.70 GHz, and integrated graphics make it the superior choice for mobile devices and any workload that benefits from fast single-core execution, floating-point math, or modern encryption instructions.

The Intel Xeon D-1746TER is the pick only for a narrow set of server-specific workloads where integer math, data compression, and string sorting are paramount. Its 10 cores and 20 threads, combined with 15 MB of L3 cache, ECC memory support, and triple-channel DDR4, make it a capable server processor despite its older 10 nm node. However, its 54.1% single-thread deficit and 38.4% floating-point gap mean it cannot match the Ultra 5 in general computing. The Xeon wins exactly three benchmarks, all in data-centric tasks, and its average score is lower. For any user not running those specific integer-heavy server applications, the Core Ultra 5 238V is the unequivocal choice based on the benchmark evidence.