AMD Ryzen 5 4600H vs Intel Core Ultra 5 115U Comparison

Head-to-Head Benchmarks

The head-to-head data reveals a fascinating split, with the AMD Ryzen 5 4600H taking 8 wins and the Intel Core Ultra 5 115U taking 7. The most dramatic Intel victory comes in Cinebench R23 multi-core, where the Core Ultra 5 115U scores 10,855 against the Ryzen's 8,072, a 34.5% advantage. This is a substantial margin in a modern rendering workload, suggesting the Intel part's architecture handles sustained multi-threaded loads with unusual efficiency despite its lower power class. The single-core Cinebench R23 result reinforces this, with Intel ahead by 34.1% (1,532 versus 1,142.5), a gap that points to a significant IPC advantage in the Core Ultra design.

However, the older Cinebench R15 benchmark tells a different story. The Ryzen 5 4600H wins multi-core by 23.2% (1,425 versus 1,094) and single-core by 12.5% (176 versus 154). This reversal between R15 and R23 is one of the most striking findings in the dataset, as it implies the two processors respond very differently to workload characteristics such as instruction mix, memory latency sensitivity, or thermal scaling under shorter versus longer render durations.

In PassMark suites, the pattern shifts again. The Ryzen dominates in data compression (202,030 versus 141,945, a 29.7% gap), data encryption (12,217 versus 8,606, a 29.6% gap), extended instructions (12,942 versus 8,620, a 33.4% gap), and random string sorting (21,689 versus 15,124, a 30.3% gap). These are heavily memory and integer-logic oriented tasks, where the Ryzen's higher base clock of 3.00 GHz and 45 W TDP likely contribute. The Intel part wins PassMark integer math only narrowly loses, 41,043 versus 45,698 (10.2% behind), but takes floating point math decisively at 30,538 versus 26,909 (13.5% ahead).

The most extreme Intel victory is in PassMark find prime numbers, where the Core Ultra 5 115U scores 53 against the Ryzen's 28, an 89.3% advantage. This test is notoriously sensitive to single-thread efficiency and branch prediction, and the result strongly favors Intel's newer core design. The Intel part also wins PassMark physics by 42.8% (898 versus 629) and PassMark single-thread by 26.4% (3,057 versus 2,418). Meanwhile, the Ryzen takes PassMark multi-thread by 8.5% (14,228 versus 13,024), a modest win that shows the AMD part's 12 threads versus Intel's 10 threads can still matter in certain parallel workloads, but not enough to overcome Intel's efficiency in other areas.

Where Each One Wins

The benchmark data suggests a clear use-case split. The Intel Core Ultra 5 115U is the stronger choice for modern rendering workloads, as evidenced by its 34.5% lead in Cinebench R23 multi-core and 34.1% lead in single-core. It also excels in floating point math, physics simulation, and prime number computation, making it well suited for scientific computing, physics-based simulations, and any workload that relies on modern instruction-level parallelism. The 89.3% lead in find prime numbers is particularly telling for mathematical and cryptographic research tasks that depend on single-thread integer throughput.

The Ryzen 5 4600H, by contrast, wins in data compression, encryption, extended instructions, random string sorting, and integer math. These are tasks that often benefit from higher memory bandwidth utilization or more aggressive multi-threading, and the Ryzen's 45 W TDP allows it to sustain higher clocks under load. For users working with large datasets, compression algorithms, or encryption-heavy workloads, the Ryzen offers a 29.7% to 33.4% advantage in the PassMark suite. The Ryzen also wins PassMark multi-thread by 8.5%, indicating that its 12 threads can be leveraged in well-parallelized, integer-heavy scenarios.

A nuance emerges in the Cinebench R15 versus R23 discrepancy. The Ryzen wins R15 multi-core by 23.2% but loses R23 multi-core by 34.5%. This suggests that the Ryzen may throttle or lose efficiency over longer, more complex render tasks, while the Intel part scales better with the newer workload. Users running short, bursty rendering tasks might see the Ryzen perform better, while longer, more demanding renders favor Intel.

FAQ

Q: Which processor has a higher Cinebench R23 multi-core score?

A: The Intel Core Ultra 5 115U scores 10,855, which is 34.5% higher than the AMD Ryzen 5 4600H's 8,072.

Q: Does the AMD Ryzen 5 4600H win any rendering benchmarks?

A: Yes, in Cinebench R15 multi-core the Ryzen scores 1,425 versus 1,094 for Intel, a 23.2% advantage, and in single-core it leads 176 versus 154, a 12.5% edge.

Q: How do the two compare in PassMark single-thread performance?

A: The Intel Core Ultra 5 115U leads with a score of 3,057 versus 2,418 for the Ryzen, a 26.4% difference.

Q: Which processor is better for data compression workloads?

A: The AMD Ryzen 5 4600H scores 202,030 in PassMark data compression, which is 29.7% higher than Intel's 141,945.

Q: What is the biggest percentage difference in any benchmark between the two?

A: The largest gap is in PassMark find prime numbers, where Intel scores 53 versus AMD's 28, an 89.3% difference.

Q: Do both processors have the same overall benchmark percentile?

A: Yes, both sit at the 70th percentile compared to all CPUs, and their average benchmark scores are close: 16,753 for Intel and 16,341 for AMD.

Specification Differences

The two processors diverge significantly in core and thread counts. The Intel Core Ultra 5 115U has 8 cores and 10 threads, while the AMD Ryzen 5 4600H has 6 cores and 12 threads. Clock speeds also differ: Intel's base clock is 1,500 MHz with a boost of 4.20 GHz, whereas AMD's base is 3.00 GHz with a boost of 4.00 GHz. The TDP gap is considerable, with Intel rated at 15 W and AMD at 45 W.

The memory subsystems differ as well. Intel supports DDR5 with a memory bandwidth of 89.6 GB/s, while AMD supports DDR4 with 51.2 GB/s. Both use dual-channel memory buses. PCIe support also varies: Intel offers Gen 4 with 12 lanes (CPU only), while AMD provides Gen 3. The integrated graphics differ, with Intel using Arc Xe-LPG 48EU and AMD using Radeon RX Vega 6.

The sockets are incompatible: Intel uses BGA 2049, and AMD uses Socket FP6. The release dates show Intel launched on 2023-12-13, while AMD launched on 2020-01-05. Neither processor has an unlocked multiplier, and both target the mobile market segment. The transistor count and die size are only listed for AMD: 9,800 million transistors and a die size of 156 mm². Intel's process node is 7 nm from Intel's foundry, while AMD's is also 7 nm but from TSMC.

Architecture Differences

The architectural divide is substantial. The Intel Core Ultra 5 115U is built on the Meteor Lake architecture, part of the Core Ultra Series 1, while the AMD Ryzen 5 4600H uses the Zen 2 architecture, codenamed Renoir, from the 4000 series. Intel's process node is 7 nm fabricated by Intel, whereas AMD's 7 nm node is fabricated by TSMC.

Cache hierarchies differ notably. Intel allocates 112 KB of L1 per core and 2 MB of L2 per core, with 10 MB of shared L3 cache. AMD provides 64 KB of L1 per core, 512 KB of L2 per core, and 8 MB of shared L3. This means Intel has a larger L3 cache overall and a much larger per-core L2, which may contribute to its single-thread advantages in certain tests. AMD's smaller cache per core is offset by its higher base clock and higher TDP.

The transistor count for AMD is listed at 9,800 million, with a die size of 156 mm². Intel's transistor count and die size are not recorded in the database. Memory support also differs at the architectural level: Intel pairs with DDR5 while AMD uses DDR4, which explains the memory bandwidth disparity of 89.6 GB/s versus 51.2 GB/s. Both processors lack ECC memory support and have locked multipliers.

The Verdict

The data paints a picture of two processors optimized for different priorities. The Intel Core Ultra 5 115U is the clear winner in modern rendering and single-thread intensive tasks, with a 34.5% lead in Cinebench R23 multi-core and a 34.1% lead in R23 single-core. Its 89.3% advantage in prime number finding and 42.8% lead in physics further cement its position for scientific and simulation workloads. The Intel part also has a higher memory bandwidth of 89.6 GB/s, which may explain its floating point math win at 13.5% over AMD.

The AMD Ryzen 5 4600H, despite being older, holds its own in integer-heavy and memory-latency-sensitive tasks. Its wins in data compression (29.7%), encryption (29.6%), extended instructions (33.4%), and random string sorting (30.3%) indicate that it remains a strong choice for data processing and security-related workloads. The Ryzen also wins PassMark multi-thread by 8.5%, showing that its 12 threads can outpace Intel's 10 threads in certain parallel tasks.

For a user deciding between these two, the choice hinges on workload type. Those running modern rendering engines, scientific simulations, floating point math, or single-thread dependent applications should choose the Intel Core Ultra 5 115U. Those dealing with compression, encryption, string processing, or integer-heavy parallel tasks may prefer the AMD Ryzen 5 4600H. The 70th percentile ranking for both processors confirms they are evenly matched overall, but the direction of the margin depends entirely on the task at hand. The data does not indicate a universal winner, only a context-dependent one.