AMD Ryzen AI Embedded P164 vs Intel Core 3 304 Comparison

The Verdict

The AMD Ryzen AI Embedded P164 is the clear performance winner in this comparison. It wins all 11 head-to-head benchmark comparisons in the database, with victory margins ranging from 4.4% to 256.9%. The data shows that the P164 delivers dramatically higher throughput in every measured workload category. It sits in the 91st percentile of all CPUs tested, while the Intel Core 3 304 sits in the 68th percentile. The average benchmark score of the AMD part is 52901, versus 13745 for the Intel part.

The Intel Core 3 304 is a low-power chip with a 15 TDP, while the AMD part uses 28 TDP. The Core 3 304 has a launch MSRP of $309. The AMD part has no launch MSRP listed in the database. For workloads that demand maximum compute throughput, the AMD Ryzen AI Embedded P164 is the only rational choice. The Intel Core 3 304 is suitable for scenarios where power draw must be minimized and single-threaded performance is the primary concern, but the benchmark data shows it trails the AMD part even there, with the P164 holding an 11.5% lead in single-thread tests.

Architecture Differences

The AMD Ryzen AI Embedded P164 uses the Gorgon Point codename, built on the Ryzen AI Embedded generation, which incorporates Zen 5 and Zen 5c cores. It is fabricated on a 4 nm process at TSMC, with a die size of 233 mm². The Intel Core 3 304 uses the Wildcat Lake codename, from the Core 3 generation, on a 3 nm process at Intel, with no die size listed in the database.

The core configurations differ substantially. The AMD part has 8 cores and 16 threads. The Intel part has 5 cores and 5 threads, meaning no simultaneous multithreading. The AMD part has a base clock of 2.00 GHz and a boost clock of 5.00 GHz. The Intel part has a base clock of 1.50 GHz and a boost clock of 4.30 GHz.

Cache structures are organized differently. The AMD part uses 80 KB of L1 per core, 1 MB of L2 per core, and 8 MB of L3. The Intel part uses 192 KB of L1 total, 2.5 MB of L2 total, and 6 MB of shared L3. The AMD part supports ECC memory; the Intel part does not. The AMD part has a dual-channel memory bus with 89.6 GB/s bandwidth, while the Intel part has a single-channel bus with 59.7 GB/s bandwidth. Both support DDR5 and LPDDR5X memory.

The AMD part has 16 PCIe Gen 4 lanes (CPU only), whereas the Intel part has 6 PCIe Gen 4 lanes (CPU only). Integrated graphics differ as well: the AMD part uses Radeon 880M, and the Intel part uses Intel Xe3 Graphics with 1 Xe core. The AMD part uses AMD Socket FP8, while the Intel part uses Intel BGA 1516. Both are mobile market segment parts with active production status. The AMD part was released in March 2026, and the Intel part in April 2026. Neither has an unlocked multiplier.

Head-to-Head Benchmarks

The PassMark suite in the database shows a consistent pattern of AMD dominance across all 11 tests. The largest margin appears in integer math, where the AMD part scores 87940 against 24640 for Intel, a 256.9% lead. This indicates a massive advantage in general arithmetic workloads. Data compression shows a 185.7% lead, with scores of 327891 versus 114775.

Random string sorting favors the AMD part by 154.8%, scoring 34801 against 13659. Extended instructions show a 149.8% lead, at 24193 versus 9686. Multithread performance is 122.7% higher for AMD, with scores of 25889 versus 11625. Data encryption shows an 88.9% advantage, at 16055 versus 8501. Floating point math is 87.7% higher, at 55799 versus 29722.

Physics calculations show a 39.4% lead for AMD, at 1210 versus 868. Even in single-thread tests, where the Intel part's lower core count might be expected to matter less, the AMD part leads by 11.5%, scoring 4029 versus 3614. The closest margin is in prime number finding, where the AMD part scores 71 against 68, a 4.4% lead.

The database shows the AMD part's nearest rivals include the AMD Ryzen 5 9500F, Intel Xeon 634, AMD EPYC 7313P, and AMD Ryzen 9 7900X. The P164 scores within 0.7% of all these parts. The Intel Core 3 304's nearest rivals include the AMD Ryzen Threadripper PRO 3975WX, Intel Core i7-8750H, Intel Core 5 120UL, and AMD EPYC 7443. The Core 3 304 sits within 1.4% of those parts. This means the AMD part competes with desktop and server class CPUs, while the Intel part competes with older mobile and lower-end server parts.

FAQ

Q: Which CPU has more cores and threads?

A: The AMD Ryzen AI Embedded P164 has 8 cores and 16 threads. The Intel Core 3 304 has 5 cores and 5 threads.

Q: What is the clock speed difference?

A: The AMD part has a 2.00 GHz base clock and a 5.00 GHz boost clock. The Intel part has a 1.50 GHz base clock and a 4.30 GHz boost clock.

Q: Which processor supports ECC memory?

A: The AMD Ryzen AI Embedded P164 supports ECC memory. The Intel Core 3 304 does not.

Q: How much memory bandwidth does each support?

A: The AMD part has a dual-channel memory bus with 89.6 GB/s bandwidth. The Intel part has a single-channel memory bus with 59.7 GB/s bandwidth.

Q: What is the process node for each?

A: The AMD part is fabricated on a 4 nm process at TSMC. The Intel part is fabricated on a 3 nm process at Intel.

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

A: The AMD Ryzen AI Embedded P164 scores 4029 in the PassMark single-thread test, which is 11.5% higher than the Intel Core 3 304's score of 3614.

Where Each One Wins

The AMD Ryzen AI Embedded P164 wins every benchmark category recorded in the database. The largest relative wins are in integer math, data compression, random string sorting, and extended instructions. These are compute-heavy workloads that benefit from the AMD part's 8 cores, 16 threads, and larger cache hierarchy. The multithread score of 25889 versus 11625 shows that heavily parallel workloads see more than double the throughput on the AMD part.

The AMD part also wins in memory-sensitive tasks. Its dual-channel memory bus with 89.6 GB/s bandwidth provides higher throughput than the Intel part's single-channel 59.7 GB/s. The data compression score of 327891 versus 114775 reflects this advantage. The AMD part's 8 MB of L3 cache, combined with 1 MB L2 per core, gives it more on-die storage for working sets.

The Intel Core 3 304 has no benchmark wins in the database. Its lower TDP of 15, versus 28 for the AMD part, suggests it is designed for power-constrained environments. The Intel part's single-channel memory bus and 6 MB of shared L3 cache are more modest. Its 5 cores and 5 threads limit multithreaded throughput. The Intel part does have a smaller process node at 3 nm versus 4 nm, which may indicate better transistor density, but the database shows no performance benefit from this.

For workloads that are purely single-threaded, the AMD part still leads by 11.5% in the PassMark single-thread test. The closest margin is prime number finding at 4.4%, which may indicate that the Intel part's architecture handles certain integer operations more efficiently relative to its other weaknesses. However, the AMD part still wins that test.

The data indicates the AMD Ryzen AI Embedded P164 is positioned for embedded and mobile systems that require substantial compute capability. Its 91st percentile ranking places it among high-end desktop and server processors. The Intel Core 3 304, at the 68th percentile, is positioned for lighter workloads where power efficiency and lower cost are the primary constraints.

Specification Differences

The two processors differ across nearly every specification in the database.

Cores and threads: The AMD part has 8 cores and 16 threads. The Intel part has 5 cores and 5 threads.

Clocks: The AMD part runs at 2.00 GHz base and 5.00 GHz boost. The Intel part runs at 1.50 GHz base and 4.30 GHz boost.

TDP: The AMD part uses 28 TDP. The Intel part uses 15 TDP.

Process node: The AMD part uses 4 nm at TSMC. The Intel part uses 3 nm at Intel.

Die size: The AMD part has a die size of 233 mm². No die size is listed for the Intel part.

Cache: The AMD part has 80 KB L1 per core, 1 MB L2 per core, and 8 MB L3. The Intel part has 192 KB L1 total, 2.5 MB L2 total, and 6 MB shared L3.

Memory: The AMD part supports dual-channel memory with 89.6 GB/s bandwidth and ECC. The Intel part supports single-channel memory with 59.7 GB/s bandwidth and no ECC.

PCIe: The AMD part has 16 PCIe Gen 4 lanes (CPU only). The Intel part has 6 PCIe Gen 4 lanes (CPU only).

Integrated graphics: The AMD part uses Radeon 880M. The Intel part uses Intel Xe3 Graphics with 1 Xe core.

Socket: The AMD part uses AMD Socket FP8. The Intel part uses Intel BGA 1516.

Release date: The AMD part was released on March 8, 2026. The Intel part was released on April 15, 2026.

Launch MSRP: The Intel part has a launch MSRP of $309. No launch MSRP is listed for the AMD part.

Part number: The Intel part has part number SAE3K. The AMD part lists its part number as unknown.

Market segment: Both are mobile parts. Both have active production status. Neither has an unlocked multiplier.