Intel Core Ultra 7 265T vs Intel Core Ultra X7 358H Comparison

Where Each One Wins

The benchmark split is decisively lopsided: the Intel Core Ultra 7 265T wins 15 of the 17 head-to-head comparisons, while the Intel Core Ultra X7 358H takes only 2. This is not a close contest. The 265T dominates in nearly every multi-threaded and single-threaded workload category, while the 358H secures its only victories in prime number finding and physics simulation.

For users prioritizing raw compute throughput, the 265T is the clear choice. Its Cinebench R23 multi-core score of 31,558 versus the 358H's 18,747 represents a 68.3% advantage, the largest gap in any multi-threaded test. The 265T also leads in PassMark multi-thread (37,084 vs 33,802, a 9.7% edge), integer math (104,943 vs 83,147, up 26.2%), floating-point math (129,817 vs 103,842, up 25%), and data compression (370,158 vs 332,508, up 11.3%). The 265T is the workstation-class performer here.

The 358H counters in two specific areas. It wins PassMark find prime numbers (337 vs 322, a 4.5% margin) and PassMark physics (3,021 vs 2,391, a 20.9% lead). The physics result is notable: the 358H outperforms the 265T by more than a fifth in this simulation-style workload, despite losing most other tests. This suggests the 358H's architecture handles certain algorithmic patterns with higher efficiency, even if overall throughput is lower.

For single-threaded tasks, the 265T is also ahead. Its Cinebench R23 single-core score of 4,455 beats the 358H's 2,080 by 114.2%, the largest single-test gap in the entire comparison. PassMark single-thread shows a narrower 5.2% lead (4,338 vs 4,124), and Cinebench R15 single-core favors the 265T by 48.9% (449 vs 301.5). The 265T's higher boost clock of 5.30 GHz versus 4.80 GHz explains part of this, but the sheer magnitude of the R23 single-core delta indicates a substantial IPC difference as well.

Architecture Differences

The two processors come from different Intel product lines and manufacturing strategies. The 265T belongs to Core Ultra Series 2, uses the Arrow Lake-S architecture, and is built on a 3 nm process by TSMC. It is a desktop part on Intel Socket 1851. The 358H belongs to Core Ultra Series 3, uses the Panther Lake-H architecture with codename Panther Lake, and is also built on a 3 nm process but by Intel's own foundry. It is a mobile part on Intel BGA 2540.

Core counts differ significantly. The 265T has 20 cores and 20 threads, while the 358H has 16 cores and 16 threads. Neither supports simultaneous multithreading, so thread counts equal core counts. The 265T's extra 4 cores directly contribute to its multi-threaded wins, especially in Cinebench R23 multi-core where the gap is largest.

Cache hierarchies diverge in L3 capacity. Both use 192 KB L1 per core and 3 MB L2 per core, but the 265T carries 30 MB shared L3 versus the 358H's 18 MB. This 12 MB difference helps the 265T in data-heavy workloads like compression and encryption, where larger shared cache reduces memory traffic.

Memory support differs by platform. The 265T uses DDR5 with dual-channel access and 102.4 GB/s bandwidth. The 358H uses LPDDR5X, also dual-channel, but with 153.6 GB/s bandwidth. The mobile part has 50% higher memory bandwidth, which may explain its physics simulation win where memory access patterns are less cache-friendly.

PCIe connectivity is another split. The 265T provides Gen 5 with 20 lanes (CPU only), while the 358H offers Gen 5 with only 4 lanes (CPU only). The desktop chip has far more expansion headroom for GPUs and NVMe drives. Integrated graphics differ as well: the 265T uses Arc Xe-LPG Graphics with 64 execution units, while the 358H uses Arc B390.

Production status is active for both, but release timing differs: the 265T launched on January 6, 2025, while the 358H launched on January 4, 2026. The 358H is effectively a year newer, yet it trails in most benchmarks. Transistor count and die size are recorded for the 265T (17,800 million transistors, 243 mm²) but are not listed for the 358H. The 265T has a launch MSRP of $384; the 358H has no recorded launch MSRP.

Head-to-Head Benchmarks

The largest single victory belongs to the 265T in Cinebench R23 single-core, where it scores 4,455 against the 358H's 2,080, a 114.2% advantage. This is not a marginal edge; the 265T more than doubles the 358H's score. The same test in multi-core shows a 68.3% lead (31,558 vs 18,747), confirming that the 265T wins on both per-core efficiency and aggregate throughput.

Cinebench R20 shows a smaller but consistent gap. Multi-core favors the 265T by 10.3% (13,254 vs 12,011), and single-core by 10.4% (1,871 vs 1,695). Cinebench R15 narrows further: multi-core is 5.1% higher for the 265T (3,180 vs 3,027), while single-core is 48.9% higher (449 vs 301.5). The R15 single-core result is anomalous compared to R20 and R23, suggesting the 358H may have been running at a lower power state during that test.

PassMark tests paint a similar picture. Integer math shows a 26.2% lead for the 265T (104,943 vs 83,147), and floating-point math shows 25% (129,817 vs 103,842). Data encryption is 14% higher (29,687 vs 26,046), while data compression is 11.3% (370,158 vs 332,508). Random string sorting favors the 265T by 10% (44,400 vs 40,357). Extended instructions are close, with only a 4.9% edge (28,609 vs 27,274). Single-thread and multi-thread PassMark scores both favor the 265T by 5.2% (4,338 vs 4,124) and 9.7% (37,084 vs 33,802) respectively.

The 358H's two wins are worth examining closely. PassMark physics gives the 358H a 20.9% lead (3,021 vs 2,391), which is its best relative performance. PassMark find prime numbers shows a narrower 4.5% edge (337 vs 322). These results indicate that the 358H's newer Panther Lake architecture has specific strengths in physics simulation and prime number generation, likely due to its higher base clock of 1.90 GHz versus the 265T's 1.50 GHz, combined with the LPDDR5X memory bandwidth advantage.

FAQ

Q: Which processor has more cores and threads?

A: The Intel Core Ultra 7 265T has 20 cores and 20 threads, while the Intel Core Ultra X7 358H has 16 cores and 16 threads. Neither supports multithreading, so thread counts match core counts.

Q: What is the largest benchmark gap between the two?

A: The largest gap is in Cinebench R23 single-core, where the 265T scores 4,455 versus the 358H's 2,080, a 114.2% difference. The second largest is Cinebench R23 multi-core at 68.3%.

Q: Does the 358H win any benchmarks?

A: Yes, the 358H wins PassMark physics (3,021 vs 2,391, a 20.9% lead) and PassMark find prime numbers (337 vs 322, a 4.5% lead). These are the only two tests where it outperforms the 265T.

Q: How do memory bandwidth and cache compare?

A: The 265T uses DDR5 with 102.4 GB/s bandwidth and 30 MB shared L3 cache. The 358H uses LPDDR5X with 153.6 GB/s bandwidth and 18 MB shared L3 cache. The 358H has higher memory bandwidth but less L3 cache.

Q: What are the platform differences?

A: The 265T is a desktop processor on Intel Socket 1851 with 20 PCIe Gen 5 lanes. The 358H is a mobile processor on Intel BGA 2540 with 4 PCIe Gen 5 lanes. The 265T has a 35 W TDP, while the 358H has a 25 W TDP.

Q: When did each processor launch?

A: The 265T launched on January 6, 2025, and the 358H launched on January 4, 2026. Both are currently active in production.

Specification Differences

The following fields differ between the two processors:

• Series: Core Ultra Series 2 (265T) vs Core Ultra Series 3 (358H)
• Cores: 20 (265T) vs 16 (358H)
• Threads: 20 (265T) vs 16 (358H)
• Base clock: 1.50 GHz (265T) vs 1.90 GHz (358H)
• Boost clock: 5.30 GHz (265T) vs 4.80 GHz (358H)
• TDP: 35 W (265T) vs 25 W (358H)
• Socket: Intel Socket 1851 (265T) vs Intel BGA 2540 (358H)
• Architecture: Arrow Lake (265T) vs not recorded for 358H
• Codename: Arrow Lake-S (265T) vs Panther Lake (358H)
• Foundry: TSMC (265T) vs Intel (358H)
• Transistors: 17,800 million (265T) vs not recorded for 358H
• Die size: 243 mm² (265T) vs not recorded for 358H
• L3 cache: 30 MB shared (265T) vs 18 MB shared (358H)
• Memory support: DDR5 (265T) vs LPDDR5X (358H)
• Memory bandwidth: 102.4 GB/s (265T) vs 153.6 GB/s (358H)
• PCIe lanes: 20 Gen 5 lanes (265T) vs 4 Gen 5 lanes (358H)
• Integrated graphics: Arc Xe-LPG 64EU (265T) vs Arc B390 (358H)
• Market segment: Desktop (265T) vs Mobile (358H)
• Release date: January 6, 2025 (265T) vs January 4, 2026 (358H)
• Launch MSRP: $384 (265T) vs not recorded for 358H
• Part number: SRQCS (265T) vs SA4RAQ9ET (358H)

The Verdict

The data is unambiguous: the Intel Core Ultra 7 265T is the superior processor for nearly all workloads. It wins 15 of 17 head-to-head tests, including every Cinebench version, all PassMark math tests, compression, encryption, and sorting. Its 20 cores and 30 MB L3 cache give it a decisive multi-threaded advantage, while its 5.30 GHz boost clock and 114.2% lead in Cinebench R23 single-core make it the clear choice for single-threaded performance as well.

The Intel Core Ultra X7 358H is not without merit. Its 20.9% win in PassMark physics and 4.5% win in prime number finding show that its newer Panther Lake architecture, combined with 153.6 GB/s LPDDR5X bandwidth, delivers specific benefits in simulation and number-theoretic workloads. Its lower 25 W TDP also makes it more suitable for battery-constrained mobile designs.

But these two wins do not offset the 265T's dominance elsewhere. The 358H trails by 68.3% in Cinebench R23 multi-core and by 25% or more in integer and floating-point math. For desktop users who need maximum compute throughput, the 265T is the only rational choice from this comparison. For mobile users who prioritize physics simulation or prime number workloads above all else, the 358H offers a targeted advantage, but it cannot compete on general-purpose performance. The database's average benchmark scores confirm this: the 265T averages 47,697 versus the 358H's 40,967, a 16.4% overall gap. The 265T also ranks in the 90th percentile of all CPUs, while the 358H sits in the 87th percentile. Both are capable processors, but the 265T is the higher-performing one by a wide margin.