Intel Core 9 273PE vs Intel Core Ultra 9 386H Comparison

Where Each One Wins

The Intel Core 9 273PE and Intel Core Ultra 9 386H are aimed at fundamentally different segments, and the benchmark data reflects that split clearly. The Core 9 273PE, a desktop part on Intel Socket 1700 with a 65 W TDP, wins 10 of the 17 recorded head-to-head comparisons. Its victories are concentrated in Cinebench multi-threaded workloads, integer math, and memory-sensitive tasks like data compression. The Core Ultra 9 386H, a mobile part on Intel BGA 2540 with a 25 W TDP, wins 7 comparisons, taking the single-thread PassMark tests, encryption, extended instructions, and prime number finding.

The Core 9 273PE is the stronger choice for sustained multi-core rendering and general productivity. In Cinebench R23 multi-core, it scores 31288 against 20547 for the Ultra 9 386H, a 52.3% advantage. It also leads in PassMark integer math by 59.7% (139410 vs 87284) and data compression by 15.2% (405885 vs 352365). These results point to a processor built for desktop workloads where power draw is less constrained and thread count matters.

The Core Ultra 9 386H, despite having fewer threads (16 vs 24), shows a different strength profile. It wins PassMark single-thread with 4218 vs 3650, a 13.5% margin. It also takes data encryption (27150 vs 22719, a 16.3% edge), extended instructions (29138 vs 24630, a 15.5% edge), and find prime numbers (341 vs 203, a 40.5% edge). Floating-point math is nearly tied, with the Ultra 9 386H ahead by just 0.6% (108527 vs 107884). This pattern suggests the newer Panther Lake architecture, built on a 3 nm node, brings efficiency-oriented IPC gains that show up in specific instruction paths.

The Cinebench R15 multi-core result is the one anomaly: the Ultra 9 386H wins 3223 vs 3153, a 2.2% margin, despite losing R20 and R23 multi-core by 2.5% and 52.3% respectively. This could reflect thermal or power behavior during shorter bursts, but the overall multi-core trend favors the desktop part.

FAQ

Q: Which processor has more cores and threads?

A: The Intel Core Ultra 9 386H has 16 cores but only 16 threads, meaning no hyper-threading. The Intel Core 9 273PE has 12 cores and 24 threads, using hyper-threading to double the thread count.

Q: What is the biggest single benchmark margin between the two?

A: The largest margin is in Cinebench R23 single-core, where the Core 9 273PE scores 4417 vs 2071.5 for the Ultra 9 386H, a 113.2% advantage. This is the only recorded test where one processor more than doubles the other's score.

Q: Does the mobile processor ever beat the desktop processor in multi-threaded tests?

A: Yes, in Cinebench R15 multi-core the Ultra 9 386H scores 3223 vs 3153, a 2.2% win. However, in R20 multi-core the Core 9 273PE wins by 2.5% (13140 vs 12820), and in R23 multi-core it wins by 52.3% (31288 vs 20547).

Q: How do the processors compare in overall average benchmark score?

A: The Core 9 273PE has an average benchmark score of 49845, while the Ultra 9 386H sits at 43210. The desktop part sits at the 90th percentile of all CPUs, the mobile part at the 88th.

Q: Which processor has a higher boost clock?

A: The Core 9 273PE boosts to 5.70 GHz, while the Ultra 9 386H boosts to 4.90 GHz. The base clocks are 2.30 GHz and 2.10 GHz respectively.

Q: What memory types does each processor support?

A: The Core 9 273PE supports DDR4 and DDR5. The Ultra 9 386H supports DDR5 and LPDDR5X. Both use dual-channel memory buses.

Head-to-Head Benchmarks

The Cinebench suite provides the clearest separation between the two. In R23 multi-core, the Core 9 273PE delivers 31288 points, which is 52.3% higher than the Ultra 9 386H's 20547. That is the second-largest margin in the entire comparison. In R23 single-core, the Core 9 273PE scores 4417, a 113.2% advantage over 2071.5. This is the single biggest delta recorded, and it indicates that the desktop part's 5.70 GHz boost clock and 12-core/24-thread configuration translate into far stronger per-thread performance in this particular test. The R20 results show a narrower gap: the Core 9 273PE wins multi-core by 2.5% (13140 vs 12820) and single-core by 2.5% (1855 vs 1809).

The PassMark suite tells a more mixed story. The Core 9 273PE dominates integer math with 139410 vs 87284, a 59.7% margin. It also wins multithread (36810 vs 35399, a 4% edge), physics (3120 vs 3028, a 3% edge), random string sorting (45098 vs 42135, a 7% edge), data compression (405885 vs 352365, a 15.2% edge), and the single-thread test appears on both sides: the Ultra 9 386H wins PassMark single-thread with 4218 vs 3650, a 13.5% margin, but the Core 9 273PE wins Cinebench R15 single-core by 46.6% (445 vs 303.5).

The Ultra 9 386H takes its wins in encryption, extended instructions, prime numbers, floating-point math, and the PassMark single-thread tests. The encryption margin is 16.3% (27150 vs 22719), extended instructions 15.5% (29138 vs 24630), and find prime numbers 40.5% (341 vs 203). Floating-point math is nearly even at 0.6%. The Cinebench R15 multi-core win for the Ultra 9 386H (3223 vs 3153, 2.2%) is its only multi-threaded victory in the Cinebench family. Across all 17 tests, the Core 9 273PE wins 10 and the Ultra 9 386H wins 7, but the magnitude of the Core 9 273PE's wins in R23 and integer math far outweighs the Ultra 9 386H's narrower edges.

Specification Differences

The two processors diverge on nearly every major specification except manufacturer and foundry. The Core 9 273PE has 12 cores and 24 threads; the Ultra 9 386H has 16 cores and 16 threads. Base clock is 2.30 GHz for the desktop part and 2.10 GHz for the mobile part. Boost clock is 5.70 GHz vs 4.90 GHz. TDP differs substantially: 65 W for the Core 9 273PE, 25 W for the Ultra 9 386H. Socket types are incompatible: Intel Socket 1700 for the desktop part, Intel BGA 2540 for the mobile part.

Cache layouts are different. The Core 9 273PE uses 80 KB of L1 per core, 2 MB of L2 per core, and 36 MB of shared L3. The Ultra 9 386H uses 192 KB of L1 per core, 2.5 MB of L2 per core, and 18 MB of shared L3. Memory support differs: the Core 9 273PE accepts DDR4 and DDR5, while the Ultra 9 386H accepts DDR5 and LPDDR5X. Memory bandwidth is higher on the mobile part at 115.2 GB/s vs 89.6 GB/s for the desktop part. ECC memory is supported on the Core 9 273PE but not on the Ultra 9 386H. PCIe lanes differ: 16 lanes on the desktop part vs 12 lanes on the mobile part, both Gen 5.

Integrated graphics are different: UHD Graphics 730 on the Core 9 273PE vs Intel Xe3 Graphics on the Ultra 9 386H. The market segments are Desktop and Mobile respectively. Release dates differ: the Core 9 273PE launched on 2026-03-08, the Ultra 9 386H on 2026-01-04. The desktop part has a launch MSRP of $549; no MSRP is recorded for the mobile part. Both have locked multipliers. The part numbers are SA4QD for the desktop and SA4R5Q9EH for the mobile.

Architecture Differences

The Core 9 273PE is built on Bartlett Lake, a 10 nm process from Intel. It belongs to the Core 9 generation. The Ultra 9 386H uses Panther Lake architecture, specifically Panther Lake-H, on a 3 nm process. It belongs to the Core Ultra Series 3 generation. Both are manufactured by Intel, but the process node difference is significant: 10 nm vs 3 nm. The smaller node on the Ultra 9 386H aligns with its lower 25 W TDP and mobile positioning.

The core counts and threading models reflect different design philosophies. The Core 9 273PE uses 12 cores with hyper-threading to reach 24 threads, maximizing throughput per core. The Ultra 9 386H uses 16 physical cores without hyper-threading, relying on raw core count and a newer architecture for efficiency. The L1 cache per core is much larger on the Ultra 9 386H (192 KB vs 80 KB), and L2 is also larger (2.5 MB vs 2 MB per core). However, the shared L3 cache is smaller on the Ultra 9 386H (18 MB vs 36 MB), which may explain why multi-threaded workloads that benefit from large shared caches, such as Cinebench R23, favor the desktop part.

The mobile part's higher memory bandwidth (115.2 GB/s vs 89.6 GB/s) and support for LPDDR5X indicate a design tuned for integrated graphics and power-constrained environments. The desktop part's support for DDR4 and DDR5, plus ECC memory, points to workstation or server-adjacent use cases. The integrated GPU differs as well: UHD Graphics 730 on the desktop part vs Intel Xe3 Graphics on the mobile part, with the latter likely benefiting from the newer architecture and higher memory bandwidth. The PCIe lane count (16 vs 12) reflects the desktop part's ability to drive more expansion cards, while the mobile part's lower count suits a compact BGA platform.