AMD Ryzen 9 9900X3D vs Intel Core Ultra 5 336H Comparison

FAQ

Q: Which processor has the higher average benchmark score?

A: The AMD Ryzen 9 9900X3D records an average benchmark score of 54762, while the Intel Core Ultra 5 336H records 34485. That puts the AMD part at the 91st percentile of all CPUs versus the 84th percentile for the Intel part.

Q: How do the two chips compare in multi-core rendering?

A: In Cinebench R23 multi-core, the AMD Ryzen 9 9900X3D scores 47737 versus 23991 for the Intel Core Ultra 5 336H, a 99% advantage. The Cinebench R20 multi-core result shows the same pattern: 20049 versus 10076, also a 99% gap.

Q: Which processor wins in single-threaded workloads?

A: The AMD Ryzen 9 9900X3D wins every single-thread test. In Cinebench R23 single-core it scores 6739 versus 3387, a 99% lead. PassMark single-thread shows 4646 versus 4013, a 15.8% lead.

Q: What is the biggest performance gap between the two?

A: The largest delta appears in PassMark integer math, where the AMD Ryzen 9 9900X3D scores 179727 versus 62070, a 189.6% advantage. PassMark data compression shows a 144.4% gap (685074 versus 280340).

Q: Are there any benchmarks where the Intel part wins?

A: No. Across all 17 head-to-head benchmark comparisons, the AMD Ryzen 9 9900X3D wins every test. The Intel Core Ultra 5 336H registers zero wins.

Q: What are the nearest rivals for each processor?

A: The AMD Ryzen 9 9900X3D sits closest to the Intel Core Ultra 5 245KF (average score 55093, a 0.6% difference). The Intel Core Ultra 5 336H sits closest to the Intel Core i7-13700HX (average score 34554, a 0.2% difference).

Architecture Differences

The AMD Ryzen 9 9900X3D belongs to the 9000 series and uses the Zen 5 architecture on the Granite Ridge codename. It is built on a 4 nm process at TSMC with 16,630 million transistors across a dual-chiplet design with each die measuring 70.6 mm². The processor uses AMD Socket AM5 and targets the desktop market segment. Its cache hierarchy consists of 80 KB of L1 per core, 1 MB of L2 per core, and a large 128 MB L3 cache. The chip supports DDR5 memory in a dual-channel configuration with 89.6 GB/s of bandwidth, includes ECC memory support, and offers PCIe Gen 5 with 24 CPU lanes. Integrated graphics come in the form of Radeon Graphics. The multiplier is unlocked, and the part number is 100-000001368.

The Intel Core Ultra 5 336H belongs to the Core Ultra Series 3 and uses the Panther Lake architecture on the Panther Lake-H codename. It is built on a 3 nm process at Intel. The processor uses Intel BGA 2540 and targets the mobile market segment. Its cache hierarchy consists of 192 KB of L1 per core, 2.5 MB of L2 per core, and 18 MB of shared L3 cache. The chip supports both DDR5 and LPDDR5X memory in a dual-channel configuration with 115.2 GB/s of bandwidth, does not support ECC memory, and offers PCIe Gen 5 with 12 CPU lanes. Integrated graphics come in the form of Intel Xe3 Graphics. The multiplier is locked, and the part number is SA4RD.

The node difference is notable: AMD uses a 4 nm TSMC process while Intel uses a 3 nm process. The AMD chip has 12 cores and 24 threads, while the Intel chip has 16 cores and 16 threads. The Intel part has no simultaneous multithreading, which explains the thread count matching the core count. The AMD part has a much larger L3 cache at 128 MB versus 18 MB, a difference driven by the 3D V-Cache design. The AMD chip has a base clock of 4.40 GHz and a boost clock of 5.50 GHz, while the Intel chip has a base clock of 1.90 GHz and a boost clock of 4.60 GHz. The TDP values differ sharply: 120 W for the AMD desktop part versus 25 W for the Intel mobile part.

Head-to-Head Benchmarks

The head-to-head data shows a clean sweep. The AMD Ryzen 9 9900X3D wins all 17 recorded comparisons, with zero wins for the Intel Core Ultra 5 336H. The margins vary considerably by workload type.

In Cinebench multi-core tests, the AMD part roughly doubles the Intel part. Cinebench R15 multi-core shows 4811 versus 2418, a 99% lead. Cinebench R20 multi-core shows 20049 versus 10076, also a 99% lead. Cinebench R23 multi-core shows 47737 versus 23991, again a 99% lead. The consistency of that doubling across three Cinebench versions indicates the performance ratio holds regardless of rendering workload intensity.

Single-core Cinebench tests follow the same pattern. Cinebench R15 single-core shows 679 versus 341, a 99.1% lead. Cinebench R20 single-core shows 2830 versus 1422, a 99% lead. Cinebench R23 single-core shows 6739 versus 3387, a 99% lead. The single-core margin is proportionally similar to the multi-core margin, which suggests the AMD core design holds a strong per-thread advantage over the Intel mobile core.

PassMark integer math delivers the largest gap in the dataset. The AMD part scores 179727 versus 62070, a 189.6% advantage. That workload heavily favors the higher clock speeds and desktop power envelope of the AMD chip. PassMark data compression shows a 144.4% gap (685074 versus 280340), and PassMark extended instructions shows a 125.8% gap (55426 versus 24542). PassMark random string sorting shows a 108.9% gap (71864 versus 34402).

PassMark multi-thread shows 56154 versus 28545, a 96.7% lead. PassMark physics shows 5340 versus 2682, a 99.1% lead. PassMark find prime numbers shows 544 versus 299, an 81.9% lead. PassMark data encryption shows 33282 versus 21291, a 56.3% lead. PassMark floating point math shows 119622 versus 82415, a 45.1% lead.

The smallest margin appears in PassMark single-thread: 4646 versus 4013, a 15.8% lead. That result indicates the Intel Core Ultra 5 336H has a respectable single-thread performance level relative to the AMD part, even though it still loses. The gap narrows considerably in single-threaded integer work compared to the multi-threaded and vectorized workloads.

Specification Differences

| Specification | AMD Ryzen 9 9900X3D | Intel Core Ultra 5 336H |

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

| Series | 9000 series | Core Ultra Series 3 |

| Manufacturer | AMD | Intel |

| Cores | 12 | 16 |

| Threads | 24 | 16 |

| Base clock | 4.40 GHz | 1.90 GHz |

| Boost clock | 5.50 GHz | 4.60 GHz |

| TDP | 120 W | 25 W |

| Socket | AMD Socket AM5 | Intel BGA 2540 |

| Architecture | Zen 5 | Panther Lake |

| Codename | Granite Ridge | Panther Lake |

| Process node | 4 nm | 3 nm |

| Foundry | TSMC | Intel |

| Transistors | 16,630 million | Not listed |

| Die size | 2x 70.6 mm² | Not listed |

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

| L2 cache | 1 MB (per core) | 2.5 MB (per core) |

| L3 cache | 128 MB | 18 MB (shared) |

| Memory support | DDR5 | DDR5, LPDDR5X |

| Memory bus | Dual-channel | Dual-channel |

| Memory bandwidth | 89.6 GB/s | 115.2 GB/s |

| ECC memory | True | False |

| PCIe | Gen 5, 24 Lanes (CPU only) | Gen 5, 12 Lanes (CPU only) |

| Integrated graphics | Radeon Graphics | Intel Xe3 Graphics |

| Market segment | Desktop | Mobile |

| Release date | 2025-01-05 | 2026-01-04 |

| Multiplier unlocked | True | False |

| Part number | 100-000001368 | SA4RD |

The memory bandwidth figure favors the Intel part: 115.2 GB/s versus 89.6 GB/s. The L1 and L2 cache per core also favor Intel: 192 KB versus 80 KB for L1, and 2.5 MB versus 1 MB for L2. The L3 cache strongly favors AMD: 128 MB versus 18 MB. The Intel part supports LPDDR5X in addition to DDR5, while the AMD part supports only DDR5. The AMD part has ECC memory support; the Intel part does not. The PCIe lane count favors AMD at 24 lanes versus 12. The AMD part has an unlocked multiplier; the Intel part is locked. The release dates differ by approximately one year, with AMD launching on January 5, 2025, and Intel launching on January 4, 2026. The AMD part carries a launch MSRP of $599. No launch MSRP is recorded for the Intel part.

The Verdict

The benchmark data presents an unambiguous result: the AMD Ryzen 9 9900X3D wins every recorded comparison against the Intel Core Ultra 5 336H. The average benchmark score of 54762 versus 34485 places the AMD part at the 91st percentile of all CPUs and the Intel part at the 84th percentile. The AMD part sits nearest to the Intel Core Ultra 5 245KF with a 0.6% average score difference, while the Intel part sits nearest to the Intel Core i7-13700HX with a 0.2% difference.

The workload profile matters. The AMD Ryzen 9 9900X3D delivers its largest advantages in integer math, data compression, extended instructions, and random string sorting, all of which exceed 100% deltas. The smallest advantage appears in single-thread PassMark, where the Intel part trails by only 15.8%. That suggests the Intel Core Ultra 5 336H has competitive single-thread efficiency for its class, but the AMD part still holds the lead.

The specification data clarifies the context. The AMD part is a desktop processor with a 120 W TDP, an unlocked multiplier, 24 PCIe Gen 5 lanes, and 128 MB of L3 cache. The Intel part is a mobile processor with a 25 W TDP, a locked multiplier, 12 PCIe Gen 5 lanes, and 18 MB of L3 cache. The Intel part offers higher memory bandwidth at 115.2 GB/s and supports LPDDR5X, which suits mobile power-constrained designs. The AMD part offers ECC memory support and a much larger L3 cache.

The data indicates that users requiring maximum compute throughput should look to the AMD Ryzen 9 9900X3D. The 99% advantage across Cinebench multi-core tests and the 189.6% advantage in integer math leave no ambiguity for rendering or computation-heavy workloads. The Intel Core Ultra 5 336H, with its lower TDP and mobile socket, serves a different purpose: it delivers 84th percentile performance in a 25 W envelope with a 2026 release date. The selection between the two should follow the target platform: desktop with an AM5 socket and high power budget for the AMD part, mobile with a BGA 2540 socket and low power budget for the Intel part.