AMD Ryzen AI Max+ 392 vs Intel Core Ultra 5 235T Comparison

FAQ

Q: Which processor has more cores and threads?

A: The AMD Ryzen AI Max+ 392 uses 12 cores and 24 threads, while the Intel Core Ultra 5 235T has 14 cores and 14 threads. The AMD part offers simultaneous multithreading, doubling its thread count, whereas the Intel part runs one thread per core.

Q: How do the average benchmark scores compare?

A: The AMD Ryzen AI Max+ 392 records an average benchmark score of 90541, placing it in the 96th percentile of all CPUs. The Intel Core Ultra 5 235T averages 38561, which puts it in the 86th percentile. The AMD processor delivers roughly 2.35 times the average score of the Intel part.

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

A: The Intel Core Ultra 5 235T leads in single-thread performance with a PassMark single-thread score of 4339, compared to the AMD Ryzen AI Max+ 392's 3927. That is a 9.5% advantage for Intel in this specific test.

Q: What memory types do the two processors support?

A: The AMD Ryzen AI Max+ 392 supports LPDDR5X memory over a quad-channel interface with 256.0 GB/s of bandwidth. The Intel Core Ultra 5 235T supports DDR5 memory over a dual-channel interface with 102.4 GB/s of bandwidth. The AMD part also supports ECC memory, while the Intel part does not.

Q: What are the manufacturing process nodes?

A: The AMD Ryzen AI Max+ 392 is built on TSMC's 4 nm process. The Intel Core Ultra 5 235T is built on TSMC's 3 nm process. The Intel die measures 243 mm² and contains 17,800 million transistors, while the AMD die is listed as 2x 70.6 mm².

Q: Which processor has a higher TDP rating?

A: The Intel Core Ultra 5 235T has a TDP of 65 watts, while the AMD Ryzen AI Max+ 392 has a TDP of 55 watts. Despite the lower thermal envelope, the AMD processor wins 8 of the 11 head-to-head benchmark comparisons.

Architecture Differences

The AMD Ryzen AI Max+ 392 uses the Zen 5 architecture under the Strix Halo codename, built on a 4 nm process from TSMC. It is a mobile-market processor with 12 cores and 24 threads, enabled by simultaneous multithreading. The core complex uses 80 KB of L1 cache per core, 1 MB of L2 cache per core, and 64 MB of shared L3 cache. The memory controller supports LPDDR5X across a quad-channel bus, providing 256.0 GB/s of memory bandwidth. ECC memory is supported. The processor includes a Radeon 8060S integrated GPU and connects through PCIe Gen 4 with 16 CPU lanes.

The Intel Core Ultra 5 235T belongs to the Core Ultra Series 2 family, using the Arrow Lake architecture under the Arrow Lake-S codename. It is a desktop-market processor built on a 3 nm TSMC process. It has 14 cores and 14 threads, meaning no hyper-threading capability. The cache hierarchy uses 192 KB of L1 per core, 3 MB of L2 per core, and 24 MB of shared L3. Memory support is DDR5 across a dual-channel bus with 102.4 GB/s of bandwidth, and ECC is not supported. The integrated graphics are Arc Xe-LPG with 24 execution units. PCIe connectivity is Gen 5 with 20 CPU lanes.

The transistor counts and die sizes differ substantially. The Intel part contains 17,800 million transistors on a 243 mm² die, while the AMD die is reported as two chiplets of 70.6 mm² each. The node difference matters for density and power behavior, though the AMD part has the lower TDP at 55 watts versus 65 watts for Intel. The Intel processor also carries a launch MSRP of $247.

The two chips target different platforms. AMD uses Socket FP11, while Intel uses Socket 1851. The AMD part is classified as mobile, the Intel part as desktop. Both have locked multipliers, so neither offers unlocked overclocking. The AMD part's part number is 100-000001979, and the Intel part's part number is SRQES.

The core topology also differs beyond raw counts. AMD's 12 cores all support two threads each, giving 24 threads for parallel workloads. Intel's 14 cores each support one thread, giving 14 threads. This structural difference explains much of the multithreaded performance gap in the benchmark data.

The Verdict

The benchmark data presents a clear split between the two processors. The AMD Ryzen AI Max+ 392 wins 8 of the 11 head-to-head comparisons, including all the multithreaded and throughput-oriented tests. The Intel Core Ultra 5 235T wins 3 comparisons: floating-point math, single-thread, and singlethread (the latter two are the same PassMark test with different names).

For workloads that depend on parallel execution, data compression, encryption, extended instruction throughput, integer math, physics simulation, and random string sorting, the AMD processor holds the advantage. The largest margin is in extended instructions, where AMD leads by 96.7%. Data compression shows an 88% lead. Integer math is 80.9% ahead. Random string sorting is 68.2% ahead. Multithread performance is 46.3% ahead. Physics is 33.8% ahead. Even the closest multithreaded win, prime number search, shows a 0.6% edge for AMD.

For workloads that favor high single-core frequency and lighter thread counts, the Intel part is the better choice. It leads floating-point math by 6.6% and single-thread performance by 9.5%. The Intel part also has a higher boost clock of 5.00 GHz, matching the AMD part's boost clock, but its base clock is 2.20 GHz versus 3.20 GHz for AMD.

The percentile ranking reinforces the overall picture. The AMD part sits at the 96th percentile of all CPUs, while the Intel part sits at the 86th percentile. The AMD part's average score of 90541 places it near the Intel Xeon 654 (90717, delta -0.2%) and the AMD Ryzen 9 9955HX (91199, delta -0.7%). The Intel part's average score of 38561 places it near the Intel Core i5-14500 (38531, delta 0.1%) and the Intel Xeon w3-2525 (38392, delta 0.4%).

The AMD processor is the stronger overall compute part in this comparison, particularly for threaded and memory-bandwidth-intensive tasks. The Intel processor is competitive in single-threaded and floating-point scenarios, and it offers a desktop platform with PCIe Gen 5 connectivity. The choice depends entirely on whether the workload is multithreaded throughput (AMD) or single-thread latency and floating-point math (Intel).

Specification Differences

The AMD Ryzen AI Max+ 392 and Intel Core Ultra 5 235T differ across nearly every major specification category.

Cores and threads: AMD has 12 cores and 24 threads; Intel has 14 cores and 14 threads.

Clock speeds: AMD base clock is 3.20 GHz, boost clock is 5.00 GHz. Intel base clock is 2.20 GHz, boost clock is 5.00 GHz.

TDP: AMD is 55 watts; Intel is 65 watts.

Socket: AMD uses AMD Socket FP11; Intel uses Intel Socket 1851.

Architecture and codename: AMD uses Zen 5 with codename Strix Halo; Intel uses Arrow Lake with codename Arrow Lake-S.

Process node and foundry: AMD is 4 nm TSMC; Intel is 3 nm TSMC.

Transistors and die size: Intel lists 17,800 million transistors and 243 mm² die size; AMD lists no transistor count but a die size of 2x 70.6 mm².

Cache: AMD L1 is 80 KB per core, L2 is 1 MB per core, L3 is 64 MB shared. Intel L1 is 192 KB per core, L2 is 3 MB per core, L3 is 24 MB shared.

Memory: AMD supports LPDDR5X, quad-channel, 256.0 GB/s, with ECC. Intel supports DDR5, dual-channel, 102.4 GB/s, without ECC.

PCIe: AMD is Gen 4 with 16 CPU lanes; Intel is Gen 5 with 20 CPU lanes.

Integrated graphics: AMD uses Radeon 8060S; Intel uses Arc Xe-LPG Graphics 24EU.

Market segment: AMD is mobile; Intel is desktop.

Release date: AMD released on 2026-01-05; Intel released on 2025-01-06. The Intel part has a launch MSRP of $247; the AMD part has no listed launch MSRP.

Head-to-Head Benchmarks

The PassMark suite provides 11 comparable metrics. The AMD Ryzen AI Max+ 392 wins 8, and the Intel Core Ultra 5 235T wins 3.

The largest AMD win is in extended instructions, scoring 45666 versus 23212, a 96.7% advantage. This test measures the throughput of SIMD and other extended instruction sets, and the Zen 5 architecture with 24 threads produces nearly double the output of the Intel part.

Data compression shows a similar pattern. AMD scores 554760 against Intel's 295100, an 88% lead. The combination of 64 MB of shared L3 cache, quad-channel LPDDR5X memory at 256.0 GB/s, and 24 threads gives the AMD part a substantial edge in this memory-intensive workload.

Integer math is another decisive AMD win. The AMD score of 152414 compares to Intel's 84244, an 80.9% margin. Integer throughput benefits directly from the thread count advantage, and the data confirms that AMD's 24 threads outperform Intel's 14 threads in this test.

Random string sorting shows AMD at 59487 versus Intel's 35377, a 68.2% lead. This workload depends on memory latency and cache behavior. AMD's larger 64 MB L3 cache and higher memory bandwidth help explain the margin.

Multithread performance, the aggregate PassMark multithread score, shows AMD at 45231 versus Intel at 30918, a 46.3% advantage. This is the broadest measure of parallel capability and reflects the cumulative effect of more threads, more cache, and higher memory bandwidth.

Physics simulation favors AMD at 2887 versus 2157, a 33.8% lead. The physics test is typically sensitive to both core count and memory throughput, both of which favor the AMD part.

Data encryption shows AMD at 27784 versus Intel's 23457, an 18.4% advantage. The AMD part's AES and cryptographic instruction throughput outpaces the Intel part, though the gap is smaller than in other AMD wins.

Prime number search is the closest AMD win. AMD scores 320 versus Intel's 318, a 0.6% margin. This test is nearly tied, indicating that the two architectures handle this specific integer workload at similar efficiency per thread.

The Intel wins are concentrated in single-thread and floating-point performance. Floating-point math shows Intel at 106546 versus AMD's 99548, a 6.6% advantage. Despite having fewer threads, the Intel part executes floating-point operations at a higher rate, likely due to its 3 nm process and per-core resources.

Single-thread performance shows Intel at 4339 versus AMD's 3927, a 9.5% advantage. The same result appears under the passmark_singlethread test name, which is the identical metric. This is the largest Intel lead in the entire comparison.

The average benchmark scores place the two parts in different competitive tiers. AMD's 90541 average puts it in the same class as the Intel Xeon 654 (90717, delta -0.2%) and AMD Ryzen 9 9955HX (91199, delta -0.7%). Intel's 38561 average places it alongside the Intel Core i5-14500 (38531, delta 0.1%) and Intel Xeon w3-2525 (38392, delta 0.4%). The AMD part outperforms its closest rivals by up to 3%, while the Intel part trades within 0.7% of its nearest competitors.

The data confirms that the AMD Ryzen AI Max+ 392 is the stronger overall performer in the PassMark suite, winning by wide margins in threaded and memory-heavy workloads. The Intel Core Ultra 5 235T holds a clear single-thread and floating-point advantage, but it cannot overcome the AMD part's multithreaded dominance in the aggregate benchmarks.