AMD Ryzen AI Max+ 392 vs Intel Xeon 6736P Comparison

The AMD Ryzen AI Max+ 392 and Intel Xeon 6736P represent two fundamentally different approaches to high-performance computing, one a mobile-first, power-efficient design and the other a server-class behemoth. Benchmark data shows a clear split: the Xeon dominates in nearly every multi-threaded and throughput-oriented task, while the Ryzen achieves a massive single-thread advantage. The average benchmark scores place the AMD part at 90,541 against the Intel’s 87,864, a 3% gap in favor of the Ryzen, yet this aggregate figure masks the distinct workload profiles where each processor excels.

FAQ

Q: Which processor has the higher average benchmark score?

A: The AMD Ryzen AI Max+ 392 scores 90,541 on average, which is 3% higher than the Intel Xeon 6736P’s 87,864. Both processors sit in the 96th percentile of all CPUs.

Q: How do the core counts compare between the two?

A: The Intel Xeon 6736P has 36 cores and 72 threads, while the AMD Ryzen AI Max+ 392 has 12 cores and 24 threads. This gives the Intel part a 3x core advantage and a 3x thread advantage.

Q: Which chip wins in single-threaded performance?

A: The AMD Ryzen AI Max+ 392 wins decisively, scoring 3,927 in the Passmark single-thread test compared to the Intel Xeon 6736P’s 2,024. This represents a 94% advantage for the AMD processor.

Q: What is the difference in memory bandwidth?

A: The Intel Xeon 6736P offers 409.6 GB/s of memory bandwidth via an eight-channel DDR5 interface, while the AMD Ryzen AI Max+ 392 provides 256.0 GB/s through a quad-channel LPDDR5X setup. The Intel part has 60% more bandwidth.

Q: Does the Intel Xeon have integrated graphics?

A: No, the Intel Xeon 6736P has no integrated graphics (N/A). The AMD Ryzen AI Max+ 392 includes a Radeon 8060S integrated GPU.

Q: What are the physical specifications for the two sockets?

A: The AMD Ryzen AI Max+ 392 uses AMD Socket FP11, while the Intel Xeon 6736P uses Intel Socket 4710. The AMD chip has a die size of 2x 70.6 mm², whereas the Intel die is 598 mm².

Architecture Differences

The architectural chasm between these two processors is stark. The AMD Ryzen AI Max+ 392 is built on TSMC’s 4 nm process node and uses the Zen 5 architecture under the Strix Halo codename. It features 12 cores and 24 threads, with a base clock of 3.20 GHz and a boost clock of 5.00 GHz. The Intel Xeon 6736P, by contrast, is manufactured on Intel’s own 5 nm node and uses the Granite Rapids architecture (Granite Rapids-SP for the Xeon 6 generation). It packs 36 cores and 72 threads, clocked at 2.00 GHz base and 4.10 GHz boost.

Cache hierarchies reveal different design philosophies. The AMD part has 80 KB of L1 cache per core, 1 MB of L2 per core, and 64 MB of shared L3 cache. The Intel Xeon doubles the per-core L1 to 112 KB and L2 to 2 MB, while its shared L3 cache is a massive 144 MB, more than double the AMD’s. This larger cache pool aligns with the Xeon’s server role, where data locality and large working sets are common.

Memory support further distinguishes them. The Ryzen AI Max+ 392 uses quad-channel LPDDR5X memory with a 256.0 GB/s bandwidth, a configuration suited for mobile and integrated systems. The Xeon 6736P employs eight-channel DDR5 memory, delivering 409.6 GB/s, a 60% bandwidth advantage. Both support ECC memory. PCIe connectivity also differs significantly: the AMD offers Gen 4 with 16 lanes, while the Intel provides Gen 5 with 88 lanes, a critical factor for expansion in server environments.

The Intel part is a server/workstation segment processor with a 205 W TDP and launch MSRP of $3351, released in February 2025. The AMD chip targets the mobile segment with a 55 W TDP and was released in January 2026. The AMD’s integrated Radeon 8060S GPU contrasts with the Xeon’s lack of integrated graphics, underscoring the Ryzen’s all-in-one mobile ambition versus the Xeon’s pure compute focus.

The Verdict

The data paints a clear picture for different buyers. The Intel Xeon 6736P is the choice for throughput-heavy, multi-threaded server workloads. Its 36 cores and 144 MB of L3 cache, combined with 409.6 GB/s memory bandwidth and 88 PCIe Gen 5 lanes, make it a dominant force in data compression, encryption, and floating-point math. It wins 9 of the 11 head-to-head benchmarks, including a 55.8% advantage in physics and a 42.6% edge in random string sorting.

The AMD Ryzen AI Max+ 392 is the pick for single-thread responsiveness and integrated-system efficiency. Its 94% lead in single-thread performance is the largest margin in any test, and its 55 W TDP versus the Xeon’s 205 W suggests a power profile that suits mobile or compact designs. The inclusion of a Radeon 8060S GPU means the Ryzen can handle graphics tasks without a discrete card, which the Xeon cannot do.

For a server rack processing massive parallel workloads, the Xeon is the obvious choice. For a workstation or high-end laptop where per-core speed and energy efficiency matter more than raw core counts, the Ryzen AI Max+ 392 holds a clear edge. The average score gap of 3% in favor of the AMD part is notable, but it is driven almost entirely by the single-thread result; the Xeon dominates the multi-thread metrics that define server performance.

Specification Differences

The specifications diverge on nearly every measurable axis. The AMD Ryzen AI Max+ 392 has 12 cores and 24 threads, while the Intel Xeon 6736P has 36 cores and 72 threads. Clock speeds differ: the AMD runs at 3.20 GHz base and 5.00 GHz boost, versus the Intel’s 2.00 GHz base and 4.10 GHz boost. The TDP is a major differentiator — 55 W for AMD and 205 W for Intel.

The process nodes are close but not identical: AMD uses TSMC’s 4 nm, while Intel uses its own 5 nm. Die size varies dramatically, with the AMD at 2x 70.6 mm² and the Intel at 598 mm². Cache configurations differ in every level: L1 is 80 KB per core (AMD) versus 112 KB per core (Intel), L2 is 1 MB per core versus 2 MB per core, and L3 is 64 MB shared versus 144 MB shared.

Memory support separates them into different classes. The AMD uses LPDDR5X with a quad-channel bus and 256.0 GB/s bandwidth; the Intel uses DDR5 with an eight-channel bus and 409.6 GB/s. The AMD has PCIe Gen 4 with 16 lanes, while the Intel has PCIe Gen 5 with 88 lanes. Integrated graphics are present on the AMD (Radeon 8060S) and absent on the Intel. The sockets differ (FP11 vs. 4710), as do the release dates (January 2026 vs. February 2025). Both are active production parts with locked multipliers.

Head-to-Head Benchmarks

The Intel Xeon 6736P wins the majority of the head-to-head comparisons, but the margins vary widely. The largest Intel victory comes in the Passmark physics test, where it scores 6,531 against the AMD’s 2,887, a 55.8% advantage. This suggests a massive difference in simulation or physics-heavy workloads. Data compression shows the Xeon at 796,658 versus 554,760 for the AMD, a 30.4% lead. Encryption follows a similar pattern, with the Xeon scoring 46,236 against 27,784, a 39.9% gap.

Floating-point math is another decisive Intel win: 149,770 versus 99,548, a 33.5% edge. Integer math also favors the Xeon at 206,833 versus 152,414 (26.3% ahead). Random string sorting sees the Xeon at 103,723 versus 59,487, a 42.6% margin. Extended instructions show a smaller but still significant 17.8% lead for Intel (55,563 vs. 45,666). Find prime numbers yields a 392 to 320 result, an 18.4% Intel advantage. Even the multithread score, which might be expected to be closer given the core count difference, goes to the Xeon at 50,072 versus 45,231, a 9.7% win.

The AMD Ryzen AI Max+ 392 wins only two tests, but it wins them decisively. The Passmark single-thread test shows a 3,927 score for AMD versus 2,024 for Intel, a 94% advantage. This is the single largest delta in the entire benchmark set. The single-thread result is duplicated in the passmark_singlethread entry, confirming the consistency of this metric. The AMD’s win in single-thread performance is so large that it pulls the average benchmark score above the Xeon’s despite losing the other nine tests.

Where Each One Wins

The Intel Xeon 6736P wins in every multi-threaded and throughput-oriented category. Its 36 cores and 144 MB L3 cache drive victories in data compression (30.4% ahead), encryption (39.9% ahead), and floating-point math (33.5% ahead). Physics simulation shows the biggest gap at 55.8%, indicating a strong advantage for scientific computing or engineering workloads. Random string sorting, a test of memory and cache efficiency, goes to the Xeon by 42.6%, likely benefiting from the higher 409.6 GB/s memory bandwidth and larger cache. Extended instruction workloads and integer math also favor the Xeon, with leads of 17.8% and 26.3% respectively. Even the multithread score, where the AMD’s higher clocks could theoretically compensate for fewer cores, goes to the Xeon by 9.7%.

The AMD Ryzen AI Max+ 392 wins in single-thread performance by a 94% margin, a result that has no close parallel in the Intel’s victories. This suggests the AMD chip is better suited for lightly-threaded applications where per-core speed is paramount, such as legacy software, certain database queries, or interactive workloads that cannot fully utilize 72 threads. The AMD’s lower TDP of 55 W versus 205 W also implies a significant efficiency advantage, making it the logical choice for environments where power draw and thermal output are constrained. The integrated Radeon 8060S GPU gives it a capability the Xeon lacks entirely, enabling systems without discrete graphics. For mobile workstations or compact desktops where single-thread responsiveness and integrated graphics are priorities, the Ryzen AI Max+ 392 is the winner; for server racks and heavy parallel compute, the Xeon 6736P is unmatched.