AMD Ryzen AI Max PRO 390 vs Intel Core Ultra X7 368H Comparison

AMD
AMD

AMD Ryzen AI Max PRO 390

CORE STATE Strix Halo
CORE SPECS 12 Cores / 24 Threads
CLOCK SPEED 3.2 Base / 5 GHz Turbo
CACHE 64 MB (shared)
MAX TDP 55W
ARCHITECTURE Zen 5
nm
PROCESS 4 nm
LAUNCH DATE 2025
VS
Intel
INTEL

Core Ultra X7 368H

CORE STATE Panther Lake
CORE SPECS 16 Cores / 16 Threads
CLOCK SPEED 2 Base / 5 GHz Turbo
CACHE 18 MB (shared)
MAX TDP 25W
ARCHITECTURE Panther Lake
nm
PROCESS 3 nm
LAUNCH DATE 2026

PERFORMANCE BENCHMARKS

cinebench_cinebench_r15_multicore
3,918
2,844
cinebench_cinebench_r15_singlecore
303
401
cinebench_cinebench_r23_multicore
24,828
28,221
cinebench_cinebench_r23_singlecore
1,976
3,984
passmark_data_compression
506,170
312,927
passmark_data_encryption
26,027
25,228
passmark_extended_instructions
40,888
25,669
passmark_find_prime_numbers
323
321
passmark_floating_point_math
94,666
105,681
passmark_integer_math
148,508
88,083
passmark_multithread
42,912
32,956
passmark_physics
2,773
2,857
passmark_random_string_sorting
55,248
38,093
passmark_single_thread
3,967
4,005
passmark_singlethread
3,967
4,005
cinebench_cinebench_r20_multicore
N/A
11,852
cinebench_cinebench_r20_singlecore
N/A
1,673

Analysis: AMD Ryzen AI Max PRO 390 vs Intel Core Ultra X7 368H

Head-to-Head Benchmarks

The recorded data splits this comparison into two distinct profiles. The AMD Ryzen AI Max PRO 390 takes 8 of the 15 head-to-head benchmarks, while the Intel Core Ultra X7 368H takes 7, but the margins tell a more important story than the win count.

The AMD part dominates in throughput-heavy workloads. Its largest margin comes in PassMark integer math, where it scores 148508 against Intel's 88083, a 68.6 percent advantage. Data compression shows a similar gap: AMD scores 506170 versus Intel's 312927, a 61.8 percent lead. Extended instructions also favor AMD heavily, 40888 to 25669, a 59.3 percent difference. These three results indicate that the Ryzen AI Max PRO 390 is substantially stronger in raw computational throughput and data manipulation tasks.

Multi-threaded performance is also an AMD win, but with a smaller margin. In PassMark multithread, AMD scores 42912 against Intel's 32956, a 30.2 percent advantage. Cinebench R15 multi-core shows AMD ahead at 3918 versus 2844, a 37.8 percent lead. However, Cinebench R23 multi-core flips the result: Intel scores 28221 while AMD scores 24828, a 12 percent deficit for AMD. This inconsistency between the two Cinebench versions suggests the AMD chip's advantage is workload-specific rather than universal.

The Intel part wins decisively in single-thread performance. Cinebench R23 single-core shows Intel at 3984 versus AMD's 1976, a 50.4 percent lead. Cinebench R15 single-core shows a similar pattern: Intel at 401 versus AMD's 303, a 24.4 percent advantage. PassMark single-thread is closer, with Intel at 4005 and AMD at 3967, a 0.9 percent edge. This indicates Intel's architecture delivers significantly higher per-core performance in certain rendering workloads, though the PassMark result shows the gap narrows in other tests.

Floating-point math favors Intel modestly: 105681 versus 94666, a 10.4 percent lead. Physics scores are nearly identical, Intel at 2857 and AMD at 2773, a 2.9 percent edge. The remaining benchmarks are AMD wins with smaller margins: data encryption at 26027 versus 25228 (3.2 percent), find prime numbers at 323 versus 321 (0.6 percent), and random string sorting at 55248 versus 38093 (45 percent).

Overall average benchmark scores place AMD far ahead: 63765 versus 40518. AMD's percentile ranking is 93 against all CPUs, while Intel sits at 87. The nearest rivals for AMD include the Intel Core i9-13900KS at 64051 (0.4 percent ahead) and the AMD Ryzen AI 7 450G at 63331 (0.7 percent behind). Intel's nearest rivals include the Intel Core 7 253PE at 40557 (0.1 percent behind) and the AMD Ryzen 9 7940H at 40431 (0.2 percent behind).

The Verdict

The data shows two processors built for different priorities. The AMD Ryzen AI Max PRO 390 is the stronger multi-threaded workhorse, with decisive leads in integer math (68.6 percent), data compression (61.8 percent), extended instructions (59.3 percent), random string sorting (45 percent), and PassMark multithread (30.2 percent). Its average benchmark score of 63765 places it in the 93rd percentile, within 0.7 percent of the AMD Ryzen AI 7 450G and 0.9 percent ahead of the Intel Core Ultra 7 265HX.

The Intel Core Ultra X7 368H is the single-thread specialist. Its Cinebench R23 single-core score of 3984 is more than double AMD's 1976, a 50.4 percent advantage. It also wins Cinebench R23 multi-core (28221 versus 24828), floating-point math (105681 versus 94666), and PassMark single-thread (4005 versus 3967). However, its average score of 40518 places it in the 87th percentile, and its nearest rivals are all within 0.2 percent, indicating a much tighter competitive field.

Users who run heavily threaded productivity workloads, data compression, encryption, or integer-heavy calculations should favor the AMD part based on these measurements. Users who prioritize single-thread rendering, floating-point math, or workloads that respond to the higher per-core performance should favor the Intel part. The Cinebench R23 multi-core result is the notable exception, where Intel wins despite losing in other multi-threaded tests, suggesting that specific rendering pipelines may behave differently than general-purpose multithreaded tasks.

Architecture Differences

The AMD Ryzen AI Max PRO 390 uses the Zen 5 architecture on the Strix Halo codename, built on a 4 nm process by TSMC. It has 12 cores and 24 threads. The Intel Core Ultra X7 368H uses the Panther Lake architecture (also its codename), built on a 3 nm process by Intel. It has 16 cores and 16 threads, meaning no simultaneous multithreading.

Cache configurations differ substantially. AMD provides 80 KB of L1 per core, 1 MB of L2 per core, and 64 MB of shared L3. Intel provides 192 KB of L1 per core, 2.5 MB of L2 per core, and 18 MB of shared L3. The larger shared L3 on AMD likely contributes to its data-heavy workload advantages, while Intel's larger per-core L1 and L2 may support its single-thread performance.

Memory architecture also differs. AMD uses quad-channel LPDDR5X with 256.0 GB/s bandwidth, while Intel uses dual-channel LPDDR5X with 153.6 GB/s bandwidth. AMD supports ECC memory; Intel does not. The memory bandwidth difference is substantial and likely explains part of AMD's lead in compression and sorting tasks.

PCIe support differs: AMD provides Gen 4 with 16 lanes (CPU only), while Intel provides Gen 5 with 4 lanes (CPU only). The integrated graphics differ as well: AMD uses the Radeon 8050S, while Intel uses Arc B390. Both are mobile parts with active production status.

Specification Differences

The core and thread counts differ: AMD has 12 cores and 24 threads, Intel has 16 cores and 16 threads. Base clocks differ: AMD runs at 3.20 GHz, Intel at 2.00 GHz. Boost clocks are identical at 5.00 GHz. TDP differs significantly: AMD is rated at 55 watts, Intel at 25 watts.

Sockets differ: AMD uses AMD Socket FP11, Intel uses Intel BGA 2540. Process nodes differ: AMD uses 4 nm from TSMC, Intel uses 3 nm from Intel's own foundry. Cache totals differ: AMD's L3 is 64 MB shared, Intel's is 18 MB shared. L1 per core differs: 80 KB versus 192 KB. L2 per core differs: 1 MB versus 2.5 MB.

Memory bandwidth differs: 256.0 GB/s versus 153.6 GB/s. Memory bus width differs: quad-channel versus dual-channel. ECC support differs: AMD supports it, Intel does not. PCIe generation and lane count differ: Gen 4 with 16 lanes versus Gen 5 with 4 lanes. Release dates differ: AMD launched January 5, 2025, Intel launched January 4, 2026. Part numbers differ: AMD is 100-000001421, Intel is SA4R7Q9EJ. Neither has a launch MSRP recorded, and neither has an unlocked multiplier.

FAQ

Q: Which processor has the higher average benchmark score?

A: The AMD Ryzen AI Max PRO 390 has an average benchmark score of 63765, while the Intel Core Ultra X7 368H has an average of 40518.

Q: Which processor wins in Cinebench R23 multi-core?

A: The Intel Core Ultra X7 368H wins with a score of 28221, compared to AMD's 24828, a 12 percent advantage.

Q: Which processor has the larger L3 cache?

A: The AMD Ryzen AI Max PRO 390 has 64 MB of shared L3, while the Intel Core Ultra X7 368H has 18 MB of shared L3.

Q: What are the thread counts for each processor?

A: The AMD Ryzen AI Max PRO 390 has 12 cores and 24 threads. The Intel Core Ultra X7 368H has 16 cores and 16 threads.

Q: Which processor supports ECC memory?

A: The AMD Ryzen AI Max PRO 390 supports ECC memory. The Intel Core Ultra X7 368H does not.

Q: How do the single-thread scores compare?

A: The Intel Core Ultra X7 368H leads in Cinebench R23 single-core with 3984 versus 1976, a 50.4 percent advantage. In PassMark single-thread, Intel scores 4005 versus AMD's 3967, a 0.9 percent edge.

Where Each One Wins

The AMD Ryzen AI Max PRO 390 wins in integer math, data compression, extended instructions, random string sorting, data encryption, find prime numbers, PassMark multithread, and Cinebench R15 multi-core. These are workloads that benefit from high memory bandwidth (256.0 GB/s), a large 64 MB L3 cache, and 24 threads. The 68.6 percent lead in integer math and the 61.8 percent lead in data compression are the strongest arguments for this processor in database, analytical, and compression-heavy applications.

The Intel Core Ultra X7 368H wins in Cinebench R23 single-core, Cinebench R15 single-core, Cinebench R23 multi-core, floating-point math, physics, and PassMark single-thread. The 50.4 percent lead in Cinebench R23 single-core is the standout result, indicating superior per-core execution in rendering pipelines. The floating-point lead of 10.4 percent suggests Intel's architecture handles scientific and simulation workloads more efficiently. The Cinebench R23 multi-core win, despite losing in other multi-threaded tests, indicates that certain rendering workloads respond to Intel's 16 physical cores and higher per-core clock behavior.

The TDP difference (55 watts for AMD, 25 watts for Intel) suggests the AMD part consumes more power to deliver its multi-threaded advantages, while Intel achieves its results within a lower power envelope. The memory bandwidth difference (256.0 GB/s versus 153.6 GB/s) directly correlates with AMD's wins in data-heavy tasks. The PCIe difference (Gen 4 with 16 lanes versus Gen 5 with 4 lanes) may matter for expansion options, though neither processor's integrated graphics performance is quantified in this data.

DETAILED SPECIFICATIONS

SPECIFICATION
AI Max PRO 390
Ultra X7 368H
Core Specs
Cores
12
16 +33.3%
Threads
24
16 -33.3%
Base Clock (GHz)
3.2
2 -37.5%
Boost Clock (GHz)
5
5 0.0%
Frequency (GHz)
3.2
2 -37.5%
Turbo Clock (GHz)
5
5 0.0%
Multiplier
32
20 -37.5%
SMP CPUs
1
1 0.0%
Cache
L1 Cache
80 KB (per core)
192 KB (per core)
L2 Cache
1 MB (per core)
2.5 MB (per core)
L3 Cache
64 MB (shared)
18 MB (shared)
Power
TDP (W)
55
25 -54.5%
Configurable TDP
45-120 W
45 W
Architecture
Architecture
Zen 5
Panther Lake
Codename
Strix Halo
Panther Lake
Generation
Ryzen AI Max PRO (Zen 5)
Ultra X7 (Panther Lake-H)
Process Size
4 nm
3 nm
Foundry
TSMC
Intel
Memory
Memory Support
LPDDR5X
LPDDR5X
Memory Bus
Quad-channel
Dual-channel
Memory Bandwidth
256.0 GB/s
153.6 GB/s
ECC Memory
Yes
No
Platform
Socket
AMD Socket FP11
Intel BGA 2540
PCIe
Gen 4, 16 Lanes(CPU only)
Gen 5, 4 Lanes(CPU only)
Intel Hybrid
Hybrid Cores
—
P-Cores: 4 E-Cores: 12
E-Core Frequency
—
1600 MHz up to 3.8 GHz
LP E-Cores
—
4
AI/NPU
NPU
Yes / 50 TOPS
Yes / 50 TOPS
Graphics
Integrated Graphics
Radeon 8050S
Arc B390
Other
Market
Mobile
Mobile
Production Status
Active
Active
Part Number
100-000001421
SA4R7Q9EJ
Package
FC-BGA
FC-BGA
Tj Max
100°C
100°C
View Ryzen AI Max PRO 390 Details View Core Ultra X7 368H Details