AMD Ryzen AI Embedded P164 vs Intel Core Ultra 5 245T Comparison

Head-to-Head Benchmarks

The head-to-head data shows a clear split between the two processors, with the Intel Core Ultra 5 245T dominating most measured workloads. The Intel part wins 9 of the 11 recorded comparisons, while the AMD Ryzen AI Embedded P164 takes 2 wins. The margins, however, vary widely depending on the workload type.

The AMD processor's largest victory comes in PassMark data compression, where it scores 327891 against Intel's 283812, a 15.5% advantage. This is a substantial lead in a workload that typically benefits from high memory bandwidth and efficient core scheduling. The second AMD win is in extended instructions, with a score of 24193 versus 22071, a 9.6% margin. These two wins demonstrate that the Zen 5 architecture in the P164 has specific strengths in particular instruction patterns.

The Intel Core Ultra 5 245T, however, posts decisive wins in several categories. The most dramatic is in PassMark find prime numbers, where Intel scores 324 against AMD's 71, a 78.1% advantage. This is the largest delta in the entire comparison. Floating point math shows Intel at 108499 versus 55799, a 48.6% lead. Physics tests follow a similar pattern, with Intel at 2278 and AMD at 1210, a 46.9% gap. These three workloads indicate that the Intel part has substantially higher raw compute throughput in mathematically intensive operations.

Data encryption also favors Intel heavily, with a score of 23656 versus 16055, a 32.1% margin. This is notable because encryption workloads often depend on both instruction set extensions and core count. The Intel part's 14 cores appear to provide an advantage here.

Multithread performance shows Intel ahead by 16%, scoring 30833 against 25889. Single-thread performance is closer, with Intel at 4367 and AMD at 4029, a 7.7% gap. Integer math is nearly identical: Intel scores 88676, AMD scores 87940, a mere 0.8% difference. Random string sorting is similarly close, with Intel at 34931 and AMD at 34801, a 0.4% gap.

The average benchmark scores reflect these results differently. The AMD P164 has an average benchmark score of 52901, while the Intel 245T has an average of 38194. This discrepancy exists because the two processors were tested against different benchmark suites. The AMD part sits at the 91st percentile among all CPUs, while the Intel part sits at the 86th percentile. The AMD processor's nearest rivals include the AMD Ryzen 5 9500F with an average score of 52873 and a 0.1% delta, the Intel Xeon 634 at 52974 with a -0.1% delta, the AMD EPYC 7313P at 53206 with a -0.6% delta, and the AMD Ryzen 9 7900X at 53288 with a -0.7% delta. The Intel part's nearest rivals are the AMD Ryzen 7 250 at 38221 with a -0.1% delta, the Intel Core i5-14490F at 38149 with a 0.1% delta, the Intel Core i5-13600HX at 38261 with a -0.2% delta, and the Intel Core i5-13600KF at 38103 with a 0.2% delta.

The data shows that the Intel processor is the stronger performer in the majority of the head-to-head tests, particularly in floating point, physics, prime number finding, and encryption. The AMD processor counters with data compression and extended instruction wins, but those are narrower in relative scope. The single-thread and integer math results are close enough to be considered competitive, while the multithread gap favors Intel by a meaningful margin.

The Verdict

The recorded data supports a clear split in use cases. The Intel Core Ultra 5 245T is the better choice for workloads that depend on floating point math, physics simulation, prime number calculations, and data encryption. Its 78.1% lead in prime numbers and 48.6% lead in floating point math are decisive. The 16% multithread advantage and 32.1% encryption advantage reinforce this position for compute-heavy tasks.

The AMD Ryzen AI Embedded P164 is the better choice for data compression workloads, where its 15.5% lead is substantial. It also holds a 9.6% edge in extended instructions. For systems where compression and specialized instruction handling matter more than raw math throughput, the AMD part has a defensible position.

The Intel part also holds the single-thread advantage at 7.7%, which matters for lightly threaded applications. The AMD part trails in physics by 46.9%, which is a significant gap for any workload involving rigid body or particle simulation. The integer math and random string sorting results are statistically close, with deltas under 1%, so those workloads should not be a deciding factor.

The Intel part operates with a 65 TDP and is a desktop part, while the AMD part uses a 28 TDP and is a mobile part. That TDP difference is consistent with the performance gap in most tests. The Intel processor consumes more power budget and delivers higher absolute performance in most measured categories. The AMD processor delivers competitive results in specific workloads while operating at a much lower power envelope.

Architecture Differences

The two processors use fundamentally different architectures. The AMD Ryzen AI Embedded P164 is built on the Gorgon Point codename with a generation designation of Ryzen AI Embedded, using Zen 5 and Zen 5c core types. The Intel Core Ultra 5 245T uses the Arrow Lake architecture, specifically Arrow Lake-S, and belongs to the Core Ultra Series 2 generation.

The manufacturing processes differ. AMD uses a 4 nm process from TSMC. Intel also uses TSMC as the foundry but on a 3 nm node. The die sizes are close, with AMD at 233 mm² and Intel at 243 mm². Intel lists a transistor count of 17,800 million, while the AMD entry does not record a transistor count.

Core organization differs substantially. The AMD part has 8 cores and 16 threads, indicating simultaneous multithreading. The Intel part has 14 cores and 14 threads, meaning it does not use hyper-threading. The core count difference explains some of the multithread performance gap, though the AMD part's thread count is higher than its core count while Intel's is equal.

Cache hierarchies are structured differently. AMD provides 80 KB of L1 per core, 1 MB of L2 per core, and 8 MB of L3. Intel provides 192 KB of L1 per core, 3 MB of L2 per core, and 24 MB of shared L3. The Intel part has substantially more L3 cache, which likely contributes to its advantages in floating point and physics workloads.

Memory support differs in type and bandwidth. AMD supports DDR5 and LPDDR5X with dual-channel memory and 89.6 GB/s bandwidth. Intel supports DDR5 with dual-channel memory and 102.4 GB/s bandwidth. Both support ECC memory. The Intel part has a 12.8 GB/s bandwidth advantage.

PCIe capabilities differ. AMD uses Gen 4 with 16 lanes from the CPU. Intel uses Gen 5 with 20 lanes from the CPU. This gives Intel both a newer PCIe generation and more available lanes.

Integrated graphics differ. AMD includes a Radeon 880M. Intel includes Arc Xe-LPG Graphics with 64 execution units. Neither part has a fully unlocked multiplier, as both list multiplierUnlocked as false.

Specification Differences

The socket types differ: AMD uses AMD Socket FP8, while Intel uses Intel Socket 1851. This means platform compatibility is entirely separate between the two.

Clock speeds are close but favor Intel slightly. AMD has a base clock of 2.00 GHz and a boost clock of 5.00 GHz. Intel has a base clock of 2.20 GHz and a boost clock of 5.10 GHz. The Intel part has a 0.20 GHz higher base clock and a 0.10 GHz higher boost clock.

TDP differs by more than double. The AMD part is rated at 28 W, while the Intel part is rated at 65 W.

Market segments differ. AMD is classified as a mobile part, while Intel is classified as a desktop part.

Release dates differ. The Intel part was released on 2025-01-06, while the AMD part has a release date of 2026-03-08.

The Intel part has a launch MSRP of $270. The AMD part has no recorded launch MSRP.

Part numbers differ, with Intel listing SRVFF and AMD listing unknown.

The Intel part has a production status of Active, as does the AMD part.

FAQ

Q: Which processor has more cores?

A: The Intel Core Ultra 5 245T has 14 cores, while the AMD Ryzen AI Embedded P164 has 8 cores. The AMD part has 16 threads while the Intel part has 14 threads.

Q: Which processor is faster in single-thread performance?

A: The Intel Core Ultra 5 245T scores 4367 in PassMark single-thread tests, while the AMD Ryzen AI Embedded P164 scores 4029. This gives Intel a 7.7% advantage.

Q: How do the two compare in data compression?

A: The AMD Ryzen AI Embedded P164 leads in data compression with a score of 327891 against Intel's 283812, a 15.5% advantage.

Q: Which processor has more L3 cache?

A: The Intel Core Ultra 5 245T has 24 MB of shared L3 cache, while the AMD Ryzen AI Embedded P164 has 8 MB of L3 cache.

Q: What is the TDP difference between the two?

A: The AMD Ryzen AI Embedded P164 has a TDP of 28 W, while the Intel Core Ultra 5 245T has a TDP of 65 W.

Q: Which processor wins in floating point math?

A: The Intel Core Ultra 5 245T scores 108499 in PassMark floating point math, against the AMD Ryzen AI Embedded P164's 55799, a 48.6% lead for Intel.

Where Each One Wins

The Intel Core Ultra 5 245T is the stronger choice for floating point math, physics simulation, prime number finding, and data encryption. Its 48.6% lead in floating point, 46.9% lead in physics, and 78.1% lead in prime numbers are the largest margins in the comparison. These results point to workloads such as scientific computing, financial modeling, and encryption-heavy applications.

The Intel part also wins in multithread performance by 16%, in single-thread by 7.7%, and in data encryption by 32.1%. Its memory bandwidth advantage of 102.4 GB/s over 89.6 GB/s supports these results. The 24 MB of shared L3 cache provides additional headroom for data-heavy workloads.

The AMD Ryzen AI Embedded P164 wins in data compression by 15.5% and in extended instructions by 9.6%. These are the only two head-to-head wins for AMD, but they are meaningful for workloads that rely heavily on compression routines or specialized instruction sets. The lower TDP of 28 W versus 65 W means the AMD part can deliver competitive results in these areas within a much smaller power budget, which matters for mobile or embedded deployments.

The integer math and random string sorting results are effectively tied, with deltas of 0.8% and 0.4% respectively. Neither processor has a meaningful advantage in these categories, so they should not drive a purchasing decision.

The overall percentile rankings place the AMD part at the 91st percentile versus Intel's 86th percentile, but this is based on different average benchmark scores across different test suites. The head-to-head results are the more direct comparison, and they favor Intel in most measured workloads. The AMD part's strengths are concentrated in compression and extended instructions, while the Intel part spreads its wins across math, physics, encryption, and multithread performance.