AMD Ryzen AI 7 450G vs AMD Ryzen AI Max 390 Comparison

AMD
AMD

AMD Ryzen AI 7 450G

CORE STATE Gorgon Point
CORE SPECS 8 Cores / 16 Threads
CLOCK SPEED 2 Base / 5.1 GHz Turbo
CACHE 8 MB
MAX TDP 65W
ARCHITECTURE Gorgon Point
nm
PROCESS 4 nm
LAUNCH DATE 2026
VS
AMD
AMD

Ryzen AI Max 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

PERFORMANCE BENCHMARKS

passmark_data_compression
402,831
487,145
passmark_data_encryption
18,937
25,097
passmark_extended_instructions
28,223
38,716
passmark_find_prime_numbers
87
316
passmark_floating_point_math
63,683
90,594
passmark_integer_math
99,732
146,519
passmark_multithread
30,422
41,737
passmark_physics
1,440
2,761
passmark_random_string_sorting
42,663
53,113
passmark_single_thread
4,309
4,028
passmark_singlethread
4,309
4,028
cinebench_cinebench_r15_multicore
N/A
3,635
cinebench_cinebench_r15_singlecore
N/A
513
cinebench_cinebench_r20_multicore
N/A
15,146
cinebench_cinebench_r20_singlecore
N/A
2,138
cinebench_cinebench_r23_multicore
N/A
36,064
cinebench_cinebench_r23_singlecore
N/A
5,091

Analysis: AMD Ryzen AI 7 450G vs AMD Ryzen AI Max 390

The AMD Ryzen AI 7 450G and the AMD Ryzen AI Max 390 represent two distinct approaches to high-performance computing within AMD’s current mobile and desktop lineups. The Ryzen AI 7 450G is a desktop-oriented processor built for the AM5 platform, while the Ryzen AI Max 390 is a mobile-focused chip designed for the FP11 socket. The recorded data shows a clear split in their performance profiles: the Max 390 dominates nearly every multi-threaded and compute-heavy workload, while the 450G holds a decisive advantage in single-threaded performance. This analysis breaks down where each processor wins, what architectural choices drive those results, and which user profile each chip best serves according to the benchmark database.

Where Each One Wins

The benchmark results are unambiguous in their distribution of victories. The AMD Ryzen AI Max 390 wins 9 of the 11 recorded head-to-head tests, while the AMD Ryzen AI 7 450G wins only 2. The 450G’s wins come exclusively in the single-threaded category, where it scores 4309 points in both the Passmark single-thread and single-thread duplicated tests. The Max 390 scores 4028 in those same tests, giving the 450G a 7% advantage in both cases. This is a meaningful gap for applications that rely heavily on single-core responsiveness, such as legacy software, certain simulation tasks, and lightly threaded workloads.

The Max 390, by contrast, claims every other benchmark in the comparison. Its largest margin comes in the prime number finding test, where it scores 316 against the 450G’s 87, a 72.5% advantage. The physics test shows a 47.8% lead (2761 versus 1440), and the integer math test shows a 31.9% lead (146519 versus 99732). The floating point math test shows a 29.7% margin (90594 versus 63683). The extended instructions test and the multithread test both show identical 27.1% leads for the Max 390 (38716 versus 28223, and 41737 versus 30422, respectively). The data encryption test shows a 24.5% lead (25097 versus 18937), and the random string sorting test shows a 19.7% lead (53113 versus 42663). The smallest multi-threaded margin is in data compression, where the Max 390 scores 487145 against the 450G’s 402831, a 17.3% advantage.

The pattern is consistent: the Max 390 is the compute powerhouse, while the 450G is the single-thread specialist. The average benchmark score for the 450G is 63331, placing it in the 93rd percentile of all CPUs in the database. The Max 390 has an average score of 56273, placing it in the 91st percentile. Despite the Max 390 winning the vast majority of individual tests, its average score is lower because the single-thread tests in its benchmark set carry different weighting in the aggregation. The data shows that the 450G’s nearest rival, the AMD Ryzen AI Max PRO 390, sits only 0.7% below it in average score, while the Max 390’s closest rival, the AMD Ryzen AI 9 HX PRO 470, is 0.1% below it.

Architecture Differences

The architectural split between the two chips explains their divergent performance. The Ryzen AI 7 450G uses the Gorgon Point codename and belongs to the Ryzen AI 400 generation, built on a hybrid Zen 5 and Zen 5c core configuration. It has 8 cores and 16 threads, with a base clock of 2.00 GHz and a boost clock of 5.10 GHz. The Max 390 uses the Strix Halo codename and belongs to the Ryzen AI Max generation, built entirely on Zen 5 architecture. It has 12 cores and 24 threads, with a base clock of 3.20 GHz and a boost clock of 5.00 GHz. The Max 390’s higher base clock and additional cores give it a substantial raw throughput advantage, while the 450G’s higher boost clock contributes to its single-thread lead.

Both processors are manufactured on TSMC’s 4 nm process node, but their die configurations differ significantly. The 450G has a die size of 195 mm², while the Max 390 uses a dual-die design with two 70.6 mm² dies, for a total of 141.2 mm². The cache hierarchy also diverges sharply. Both chips have 80 KB of L1 cache per core and 1 MB of L2 cache per core. The L3 cache, however, is a major differentiator: the 450G has 8 MB of L3 cache, while the Max 390 has 64 MB of shared L3 cache. That eightfold difference in L3 capacity is critical for workloads that repeatedly access large data sets, which helps explain the Max 390’s strong performance in data compression and random string sorting.

Memory support further separates the two. The 450G supports both DDR5 and LPDDR5X memory over a dual-channel bus, providing 89.6 GB/s of memory bandwidth. The Max 390 supports only LPDDR5X memory over a quad-channel bus, delivering 256.0 GB/s of memory bandwidth. That is roughly 2.9 times the bandwidth of the 450G, a factor that directly benefits memory-intensive compute tasks such as encryption and extended instruction workloads. Both chips support ECC memory. The 450G uses AMD Socket AM5, while the Max 390 uses AMD Socket FP11. The 450G has an unlocked multiplier, while the Max 390 does not, which indicates that the desktop part is intended for overclocking while the mobile part is fixed.

PCIe connectivity also differs. The 450G provides Gen 4 with 12 lanes from the CPU, while the Max 390 provides Gen 4 with 16 lanes from the CPU. The integrated graphics differ as well: the 450G uses the Radeon 860M, while the Max 390 uses the Radeon 8050S. The 450G is classified as a desktop part with a TDP of 65 watts, while the Max 390 is classified as a mobile part with a TDP of 55 watts. The release dates also differ, with the Max 390 launching earlier in 2025 and the 450G launching later in 2026.

The Verdict

The data points to a clear division of roles. The AMD Ryzen AI Max 390 is the superior choice for any workload that scales across cores and benefits from large cache capacity and high memory bandwidth. Its 12 cores and 24 threads, combined with 64 MB of L3 cache and 256.0 GB/s of memory bandwidth, deliver decisive wins in every multi-threaded benchmark in the comparison. The 27.1% lead in multithread performance, the 31.9% lead in integer math, and the 29.7% lead in floating point math make it the obvious pick for rendering, scientific computation, data processing, and any parallel compute task.

The AMD Ryzen AI 7 450G, however, holds the single-thread crown. Its 7% lead in single-thread performance, driven by a 5.10 GHz boost clock, makes it the better option for applications that are latency-sensitive or that cannot utilize more than a few cores. The 450G also carries an unlocked multiplier, which suggests headroom for user-controlled frequency adjustments, and its 65-watt TDP in a desktop socket context indicates a different thermal and power envelope than the mobile-oriented Max 390.

For a user building a desktop system on the AM5 platform, the 450G offers a strong blend of high single-thread performance and respectable multi-threaded capability, with an average benchmark score that places it in the 93rd percentile of all CPUs. For a user seeking maximum compute throughput in a mobile form factor, the Max 390 delivers that in nearly every measured category, despite its slightly lower overall percentile ranking. The database shows that the Max 390’s 91st percentile placement reflects its lower single-thread scores, but its dominant multi-threaded results make it the more capable all-around compute processor in this pair.

FAQ

Q: Which processor has the higher single-thread benchmark score?

A: The AMD Ryzen AI 7 450G scores 4309 in the Passmark single-thread test, which is 7% higher than the AMD Ryzen AI Max 390’s score of 4028 in the same test.

Q: How large is the multi-threaded performance gap between the two?

A: The AMD Ryzen AI Max 390 scores 41737 in the Passmark multithread test, which is 27.1% higher than the AMD Ryzen AI 7 450G’s score of 30422.

Q: What is the difference in L3 cache capacity?

A: The AMD Ryzen AI Max 390 has 64 MB of shared L3 cache, while the AMD Ryzen AI 7 450G has 8 MB of L3 cache.

Q: Which processor supports quad-channel memory?

A: The AMD Ryzen AI Max 390 supports quad-channel LPDDR5X memory with 256.0 GB/s of bandwidth. The AMD Ryzen AI 7 450G supports dual-channel DDR5 and LPDDR5X memory with 89.6 GB/s of bandwidth.

Q: Which processor has more cores and threads?

A: The AMD Ryzen AI Max 390 has 12 cores and 24 threads. The AMD Ryzen AI 7 450G has 8 cores and 16 threads.

Q: What is the largest performance margin in any single benchmark?

A: The largest margin is in the Passmark find prime numbers test, where the AMD Ryzen AI Max 390 scores 316 versus the AMD Ryzen AI 7 450G’s 87, a 72.5% advantage for the Max 390.

Head-to-Head Benchmarks

The head-to-head data reveals a lopsided contest. The AMD Ryzen AI Max 390 wins 9 of 11 tests, and its margins range from 17.3% to 72.5%. The AMD Ryzen AI 7 450G wins only the two single-thread tests, with a 7% margin in each.

Starting with the Max 390’s biggest win, the prime number finding test shows a score of 316 against the 450G’s 87, a 72.5% gap. This test is highly sensitive to core count and memory bandwidth, both of which favor the Max 390. The physics test shows a 47.8% lead (2761 versus 1440), which reflects the Max 390’s superior multi-core throughput. The integer math test shows a 31.9% lead (146519 versus 99732), and the floating point math test shows a 29.7% lead (90594 versus 63683). The extended instructions test shows a 27.1% lead (38716 versus 28223), matching the multithread test’s 27.1% margin (41737 versus 30422).

The encryption test shows a 24.5% lead (25097 versus 18937), a workload that benefits heavily from the Max 390’s 256.0 GB/s memory bandwidth. The random string sorting test shows a 19.7% lead (53113 versus 42663), which also depends on memory throughput and cache size. The data compression test shows the smallest Max 390 margin at 17.3% (487145 versus 402831), still a substantial advantage.

On the 450G side, the single-thread test shows a score of 4309 against the Max 390’s 4028, a 7% lead. The duplicate single-thread test shows the identical 4309 versus 4028 result. That 7% margin in single-thread performance is the only area where the 450G surpasses the Max 390, and it is consistent across both recorded entries.

The overall benchmark picture is that the Max 390 is the dominant compute processor in this comparison, with wins across all multi-threaded and memory-sensitive tests. The 450G’s single-thread advantage is real but narrow, and it does not offset the Max 390’s broad superiority in the other nine tests. The average benchmark scores reflect this tension: the 450G averages 63331, while the Max 390 averages 56273, but that gap is due to the inclusion of the single-thread tests in the aggregate, where the 450G’s higher score pulls its average up despite losing most individual tests. The data shows that for users prioritizing raw multi-threaded performance, the Max 390 is the clear winner, while the 450G appeals to those who need maximum single-thread responsiveness.

DETAILED SPECIFICATIONS

SPECIFICATION
AI 7 450G
AI Max 390
Core Specs
Cores
8
12 +50.0%
Threads
16
24 +50.0%
Base Clock (GHz)
2
3.2 +60.0%
Boost Clock (GHz)
5.1
5 -2.0%
Frequency (GHz)
2
3.2 +60.0%
Turbo Clock (GHz)
5.1
5 -2.0%
Multiplier
20
32 +60.0%
SMP CPUs
1
1 0.0%
Cache
L1 Cache
80 KB (per core)
80 KB (per core)
L2 Cache
1 MB (per core)
1 MB (per core)
L3 Cache
8 MB
64 MB (shared)
Power
TDP (W)
65
55 -15.4%
PPT
88 W
—
Configurable TDP
—
45-120 W
Architecture
Architecture
—
Zen 5
Codename
Gorgon Point
Strix Halo
Generation
Ryzen AI 400 (Zen 5 / Zen 5c)
Ryzen AI Max (Zen 5 (Strix Halo))
Process Size
4 nm
4 nm
Die Size
195 mm²
2x 70.6 mm²
Foundry
TSMC
TSMC
Memory
Memory Support
DDR5, LPDDR5X
LPDDR5X
Memory Bus
Dual-channel
Quad-channel
Memory Bandwidth
89.6 GB/s
256.0 GB/s
ECC Memory
Yes
Yes
Platform
Socket
AMD Socket AM5
AMD Socket FP11
Chipsets
X870E, X870, B850, B840, X670E, X670, B650E, B650, A620
—
PCIe
Gen 4, 12 Lanes(CPU only)
Gen 4, 16 Lanes(CPU only)
Intel Hybrid
Hybrid Cores
4 + 4
—
E-Core Frequency
2000 MHz up to 3.6 GHz
—
AI/NPU
NPU
Yes / 50 TOPS
Yes / 50 TOPS
Graphics
Integrated Graphics
Radeon 860M
Radeon 8050S
Other
Market
Desktop
Mobile
Production Status
Active
Active
Part Number
100-000001917
100-000001423
Package
FC-LGA1718
FC-BGA
Tj Max
95°C
100°C
View Ryzen AI 7 450G Details View Ryzen AI Max 390 Details