AMD Ryzen AI Max 390 vs Intel Core Ultra 3 205 Comparison

AMD
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
VS
Intel
INTEL

Core Ultra 3 205

CORE STATE Arrow Lake-S
CORE SPECS 8 Cores / 8 Threads
CLOCK SPEED 3.8 Base / 4.9 GHz Turbo
CACHE 15 MB (shared)
MAX TDP 57W
ARCHITECTURE Arrow Lake
nm
PROCESS 3 nm
LAUNCH DATE 2025

PERFORMANCE BENCHMARKS

cinebench_cinebench_r15_multicore
3,635
2,001
cinebench_cinebench_r15_singlecore
513
282
cinebench_cinebench_r20_multicore
15,146
8,339
cinebench_cinebench_r20_singlecore
2,138
1,177
cinebench_cinebench_r23_multicore
36,064
19,856
cinebench_cinebench_r23_singlecore
5,091
2,803
passmark_data_compression
487,145
260,651
passmark_data_encryption
25,097
18,772
passmark_extended_instructions
38,716
22,578
passmark_find_prime_numbers
316
268
passmark_floating_point_math
90,594
75,136
passmark_integer_math
146,519
54,755
passmark_multithread
41,737
26,167
passmark_physics
2,761
2,016
passmark_random_string_sorting
53,113
31,083
passmark_single_thread
4,028
4,575
passmark_singlethread
4,028
4,575

Analysis: AMD Ryzen AI Max 390 vs Intel Core Ultra 3 205

The AMD Ryzen AI Max 390 and the Intel Core Ultra 3 205 serve completely different segments of the processor market. The Ryzen AI Max 390 is a mobile part, while the Core Ultra 3 205 is a desktop chip. The benchmark data shows a decisive performance gap, with the AMD part winning 15 of 17 head-to-head tests. The Intel part secures only two wins, both in single-threaded PassMark tests. The average benchmark score for the AMD processor is 56273, placing it in the 91st percentile of all CPUs, while the Intel processor averages 31473, sitting in the 82nd percentile.

Where Each One Wins

The AMD Ryzen AI Max 390 wins across the entire Cinebench suite and nearly all PassMark workloads. Its largest victory comes in PassMark integer math, where it scores 146519 against 54755, a delta of 167.6%. The data compression test shows an 86.9% delta, with scores of 487145 versus 260651. Extended instructions also favor the AMD part heavily, 38716 to 22578, a 71.5% delta. The multithreaded PassMark score is 41737 for AMD versus 26167 for Intel, a 59.5% advantage. Physics, floating point math, random string sorting, data encryption, and prime number finding all go to the AMD chip, with deltas ranging from 17.9% to 37%.

The Intel Core Ultra 3 205 takes the PassMark single-thread test with a score of 4575 against 4028, a 12% lead. This is the only category where the Intel part shows a clear advantage. The other Intel "win" is the duplicate singlethread test with the same scores. This suggests that while the Intel chip has a higher peak single-thread throughput in this specific benchmark, the AMD processor dominates in every other measured workload. For multi-threaded rendering, data processing, and compute-heavy tasks, the AMD Ryzen AI Max 390 is the clear choice. For a single, narrow single-threaded PassMark metric, the Intel part leads.

Architecture Differences

The two processors are built on different architectures and process nodes. The AMD Ryzen AI Max 390 uses Zen 5 architecture on a 4 nm TSMC process, codenamed Strix Halo. The Intel Core Ultra 3 205 uses Arrow Lake architecture on a 3 nm TSMC process, codenamed Arrow Lake-S. The AMD part has 12 cores and 24 threads, while the Intel part has 8 cores and 8 threads. This core and thread disparity explains much of the multi-threaded performance gap.

Cache configurations differ substantially. The AMD chip has 80 KB of L1 cache per core, 1 MB of L2 per core, and 64 MB of shared L3 cache. The Intel chip has 192 KB of L1 per core, 3 MB of L2 per core, and only 15 MB of shared L3. The AMD processor's larger L3 pool, 64 MB versus 15 MB, provides a significant advantage in data-heavy workloads. The die size also differs, with AMD using a dual-die setup of 2x 70.6 mm² versus Intel's single 243 mm² die. Intel lists 17,800 million transistors, while AMD does not report a transistor count.

Memory support and bandwidth are major differentiators. The AMD processor supports LPDDR5X memory on a quad-channel bus, delivering 256.0 GB/s of bandwidth. The Intel processor supports DDR5 on a dual-channel bus, with 102.4 GB/s of bandwidth. This 2.5x bandwidth advantage for AMD directly impacts memory-sensitive benchmarks. AMD also supports ECC memory, while Intel does not. PCIe capabilities favor Intel, which offers Gen 5 with 20 lanes, versus AMD's Gen 4 with 16 lanes. The integrated graphics differ as well: AMD uses a Radeon 8050S, while Intel uses Arc Xe-LPG Graphics 16EU.

The Verdict

The recorded data is unambiguous. The AMD Ryzen AI Max 390 outperforms the Intel Core Ultra 3 205 in nearly every benchmark category. The AMD part's average benchmark score of 56273 is 79% higher than Intel's 31473. The Intel chip's nearest rivals, such as the Intel Core 7 240H and AMD Ryzen 9 5980HX, have average scores around 31483 to 31495, placing the Core Ultra 3 205 in its correct performance tier. The AMD chip's rivals include the AMD Ryzen AI 9 HX PRO 470 and Intel Core i9-14900HX, with scores in the 56000 range.

The Intel Core Ultra 3 205 has a single meaningful advantage: the PassMark single-thread score of 4575, which is 12% ahead of AMD's 4028. This indicates that for a strictly single-threaded PassMark workload, the Intel part is faster. However, this does not translate into wins in Cinebench single-core tests, where AMD leads by 81.6% to 81.9%. The Intel part also has a higher base clock of 3.80 GHz versus 3.20 GHz, and a boost clock of 4.90 GHz versus 5.00 GHz. The Intel part is a desktop processor with a launch MSRP of $140, though pricing analysis is outside the scope of this data.

For any workload that uses multiple cores, threads, or substantial memory bandwidth, the AMD Ryzen AI Max 390 is the superior processor. The data shows a 81.7% lead in Cinebench R15 multicore, a 81.6% lead in R20 and R23 multicore, and a 167.6% lead in integer math. The only scenario where the Intel part wins is the narrow PassMark single-thread test. The AMD processor belongs to a higher performance class, as evidenced by its 91st percentile ranking versus Intel's 82nd percentile.

FAQ

Q: Which processor has more cores and threads?

A: The AMD Ryzen AI Max 390 has 12 cores and 24 threads, while the Intel Core Ultra 3 205 has 8 cores and 8 threads.

Q: What is the performance difference in Cinebench R23 multicore?

A: The AMD Ryzen AI Max 390 scores 36064, while the Intel Core Ultra 3 205 scores 19856, giving AMD an 81.6% lead.

Q: Does the Intel Core Ultra 3 205 win any benchmark?

A: Yes, the Intel part wins the PassMark single-thread test with a score of 4575 versus 4028 for AMD, a 12% advantage.

Q: What memory bandwidth does each processor support?

A: The AMD Ryzen AI Max 390 supports LPDDR5X with 256.0 GB/s bandwidth on a quad-channel bus. The Intel Core Ultra 3 205 supports DDR5 with 102.4 GB/s on a dual-channel bus.

Q: What is the process node for each chip?

A: The AMD Ryzen AI Max 390 uses a 4 nm TSMC process, while the Intel Core Ultra 3 205 uses a 3 nm TSMC process.

Q: How do the average benchmark scores compare?

A: The AMD Ryzen AI Max 390 has an average benchmark score of 56273, while the Intel Core Ultra 3 205 has an average of 31473.

Head-to-Head Benchmarks

The largest margin of victory for the AMD Ryzen AI Max 390 is in PassMark integer math, where it scores 146519 against 54755, a 167.6% delta. This indicates a massive advantage in arithmetic-heavy integer operations. The data compression test also shows a huge gap, with AMD at 487145 and Intel at 260651, an 86.9% delta. Extended instructions follow with a 71.5% delta, 38716 versus 22578. Random string sorting shows a 70.9% delta, 53113 versus 31083.

The Cinebench suite shows consistent 81.6% to 81.9% deltas in favor of AMD. In Cinebench R15 multicore, AMD scores 3635 versus 2001. In R15 singlecore, the scores are 513 versus 282. R20 multicore shows 15146 versus 8339, and R20 singlecore shows 2138 versus 1177. R23 multicore is 36064 versus 19856, and R23 singlecore is 5091 versus 2803. These results confirm that the AMD architecture delivers roughly 80% more performance across both single and multi-threaded rendering workloads.

PassMark multithread shows AMD at 41737 versus 26167, a 59.5% delta. Physics tests show a 37% delta, with scores of 2761 and 2016. Data encryption shows a 33.7% delta, 25097 versus 18772. Floating point math has a 20.6% delta, 90594 versus 75136. The smallest AMD margin is in find prime numbers, where it scores 316 versus 268, a 17.9% delta. The only Intel victory is PassMark single-thread, where it scores 4575 against 4028, a 12% delta.

Specification Differences

The core count differs significantly: AMD has 12 cores and 24 threads, Intel has 8 cores and 8 threads. The base clock is higher on the Intel part at 3.80 GHz versus 3.20 GHz, but the boost clock is higher on AMD at 5.00 GHz versus 4.90 GHz. The TDP is 55 watts for AMD and 57 watts for Intel. The AMD processor uses AMD Socket FP11, while the Intel uses Intel Socket 1851. The market segment differs, with AMD as Mobile and Intel as Desktop.

The process node is 4 nm for AMD and 3 nm for Intel, both fabricated by TSMC. The die size is 2x 70.6 mm² for AMD and 243 mm² for Intel. Intel reports 17,800 million transistors, AMD does not. Cache allocation differs per core: AMD has 80 KB L1 and 1 MB L2 per core, Intel has 192 KB L1 and 3 MB L2 per core. The shared L3 is 64 MB for AMD and 15 MB for Intel. Memory support is LPDDR5X for AMD and DDR5 for Intel. The memory bus is quad-channel for AMD and dual-channel for Intel. Memory bandwidth is 256.0 GB/s for AMD and 102.4 GB/s for Intel. ECC memory is supported by AMD only. PCIe is Gen 4 with 16 lanes for AMD, Gen 5 with 20 lanes for Intel. Integrated graphics are Radeon 8050S for AMD and Arc Xe-LPG Graphics 16EU for Intel. The release dates are 2025-01-05 for AMD and 2025-07-31 for Intel.

DETAILED SPECIFICATIONS

SPECIFICATION
AI Max 390
Ultra 3 205
Core Specs
Cores
12
8 -33.3%
Threads
24
8 -66.7%
Base Clock (GHz)
3.2
3.8 +18.7%
Boost Clock (GHz)
5
4.9 -2.0%
Frequency (GHz)
3.2
3.8 +18.7%
Turbo Clock (GHz)
5
4.9 -2.0%
Multiplier
32
38 +18.8%
SMP CPUs
1
1 0.0%
Cache
L1 Cache
80 KB (per core)
192 KB (per core)
L2 Cache
1 MB (per core)
3 MB (per core)
L3 Cache
64 MB (shared)
15 MB (shared)
Power
TDP (W)
55
57 +3.6%
PL1
—
57 W
PL2
—
76 W
Configurable TDP
45-120 W
—
Architecture
Architecture
Zen 5
Arrow Lake
Codename
Strix Halo
Arrow Lake-S
Generation
Ryzen AI Max (Zen 5 (Strix Halo))
Ultra 3 (Arrow Lake)
Process Size
4 nm
3 nm
Transistors
—
17,800 million
Die Size
2x 70.6 mm²
243 mm²
Foundry
TSMC
TSMC
Memory
Memory Support
LPDDR5X
DDR5
Memory Bus
Quad-channel
Dual-channel
Memory Bandwidth
256.0 GB/s
102.4 GB/s
ECC Memory
Yes
No
Platform
Socket
AMD Socket FP11
Intel Socket 1851
Chipsets
—
Z890, B860, W880, Q870, H810
PCIe
Gen 4, 16 Lanes(CPU only)
Gen 5, 20 Lanes(CPU only)
Intel Hybrid
Hybrid Cores
—
P-Cores: 4 E-Cores: 4
E-Core Frequency
—
3.2 GHz up to 4.4 GHz
P-Core Turbo
—
4.7 GHz
AI/NPU
NPU
Yes / 50 TOPS
—
Graphics
Integrated Graphics
Radeon 8050S
Arc Xe-LPG Graphics 16EU
Other
Market
Mobile
Desktop
Production Status
Active
Active
Launch Price
—
$140
Part Number
100-000001423
SRVFC
Package
FC-BGA
FC-LGA18W
Tj Max
100°C
105°C
View Ryzen AI Max 390 Details View Core Ultra 3 205 Details