AMD EPYC 4345P vs Intel Core Ultra 7 265T Comparison

AMD
AMD

AMD EPYC 4345P

CORE STATE Grado
CORE SPECS 8 Cores / 16 Threads
CLOCK SPEED 3.8 Base / 5.5 GHz Turbo
CACHE 32 MB (shared)
MAX TDP 65W
ARCHITECTURE Zen 5
nm
PROCESS 4 nm
LAUNCH DATE 2025
VS
Intel
INTEL

Core Ultra 7 265T

CORE STATE Arrow Lake-S
CORE SPECS 20 Cores / 20 Threads
CLOCK SPEED 1.5 Base / 5.3 GHz Turbo
CACHE 30 MB (shared)
MAX TDP 35W
ARCHITECTURE Arrow Lake
nm
PROCESS 3 nm
LAUNCH DATE 2025

PERFORMANCE BENCHMARKS

cinebench_cinebench_r15_multicore
3,227
3,180
cinebench_cinebench_r15_singlecore
455
449
cinebench_cinebench_r20_multicore
13,448
13,254
cinebench_cinebench_r20_singlecore
1,898
1,871
cinebench_cinebench_r23_multicore
32,020
31,558
cinebench_cinebench_r23_singlecore
4,520
4,455
passmark_data_compression
421,490
370,158
passmark_data_encryption
23,313
29,687
passmark_extended_instructions
31,663
28,609
passmark_find_prime_numbers
167
322
passmark_floating_point_math
74,255
129,817
passmark_integer_math
123,774
104,943
passmark_multithread
36,123
37,084
passmark_physics
2,583
2,391
passmark_random_string_sorting
46,238
44,400
passmark_single_thread
4,408
4,338
passmark_singlethread
4,408
4,338

Analysis: AMD EPYC 4345P vs Intel Core Ultra 7 265T

The AMD EPYC 4345P and Intel Core Ultra 7 265T represent two fundamentally different approaches to high-performance computing within a constrained power envelope. The EPYC 4345P is a server-oriented processor on the AMD Socket AM5 platform, built on the Zen 5 architecture with 8 cores and 16 threads. The Core Ultra 7 265T is a desktop-focused chip on Intel Socket 1851, using the Arrow Lake architecture with 20 cores and 20 threads. Both processors achieve the 90th percentile among all CPUs, yet their benchmark profiles diverge sharply across different workload types. The data reveals a clear split: the AMD part dominates in single-threaded and integer-heavy tasks, while the Intel part excels in floating-point and specific parallel workloads. This analysis examines the benchmark results, architectural differences, and practical implications for each processor.

Where Each One Wins

The benchmark suite shows a decisive split in workload preferences. The AMD EPYC 4345P secures 13 wins out of 17 head-to-head tests, while the Intel Core Ultra 7 265T takes 4 wins. The AMD processor’s victories are concentrated in Cinebench rendering tests across all three versions—R15, R20, and R23—both multi-core and single-core. In these tests, the EPYC 4345P consistently leads by 1.3% to 1.5% margins, which are narrow but uniform across every Cinebench variant. This consistency suggests a general efficiency advantage in rendering workloads.

The Intel part’s wins are more specialized. It dominates in passmark_data_encryption with a 21.5% lead, passmark_find_prime_numbers with a 48.1% lead, and passmark_floating_point_math with a 42.8% lead. It also edges out the AMD chip in passmark_multithread by 2.6%. These results indicate that the Core Ultra 7 265T is substantially stronger in floating-point arithmetic and encryption tasks, where its wider core configuration pays off.

For general-purpose productivity, the AMD EPYC 4345P wins in passmark_integer_math with a 17.9% advantage and passmark_data_compression with a 13.9% lead. It also takes passmark_extended_instructions by 10.7%, passmark_physics by 8%, and passmark_random_string_sorting by 4.1%. The pattern is clear: the AMD chip handles integer math, compression, and physics simulations better, while the Intel chip excels at floating-point math, prime number finding, and encryption.

Architecture Differences

The two processors are built on different nodes and architectures, which explains their divergent benchmark behavior. The AMD EPYC 4345P uses the Zen 5 architecture with the codename Grado, manufactured on a 4 nm process at TSMC. It contains 8,315 million transistors on a 70.6 mm² die. The Intel Core Ultra 7 265T uses the Arrow Lake architecture with the codename Arrow Lake-S, manufactured on a 3 nm process, also at TSMC. It packs 17,800 million transistors on a 243 mm² die.

Core counts differ substantially: the EPYC 4345P has 8 cores and 16 threads, while the Core Ultra 7 265T has 20 cores and 20 threads, meaning the Intel chip has no hyperthreading. Cache hierarchies also differ: the AMD chip has 80 KB of L1 cache per core, 1 MB of L2 per core, and 32 MB of shared L3. The Intel chip has 192 KB of L1 per core, 3 MB of L2 per core, and 30 MB of shared L3.

The EPYC 4345P runs at a base clock of 3.80 GHz with a boost clock of 5.50 GHz. The Core Ultra 7 265T has a much lower base clock of 1.50 GHz but a boost clock of 5.30 GHz. Both are locked processors with no unlocked multiplier. The AMD part supports ECC memory, while the Intel part does not. Both support DDR5 memory with dual-channel buses, but the Intel chip has higher memory bandwidth at 102.4 GB/s compared to the AMD chip’s 89.6 GB/s.

PCIe support differs: the EPYC 4345P offers 24 lanes of Gen 5 from the CPU, while the Core Ultra 7 265T offers 20 lanes. Both have integrated graphics—the AMD chip has Radeon Graphics, and the Intel chip has Arc Xe-LPG Graphics 64EU. The AMD chip targets the Server/Workstation market segment, while the Intel chip targets Desktop. Their release dates differ by several months, with the Intel part launching earlier.

Head-to-Head Benchmarks

The Cinebench results show a remarkably consistent pattern. In Cinebench R15 multi-core, the EPYC 4345P scores 3227 against the Core Ultra 7 265T’s 3180, a 1.5% lead. Single-core R15 shows 455 versus 449, a 1.3% edge. In R20 multi-core, the scores are 13448 versus 13254, again a 1.5% lead. R20 single-core shows 1898 versus 1871, a 1.4% margin. R23 multi-core delivers 32020 versus 31558, and R23 single-core delivers 4520 versus 4455, both with 1.5% leads for the AMD chip.

The Passmark suite reveals larger divergences. The biggest AMD win comes in passmark_data_compression, where it scores 421490 against 370158, a 13.9% advantage. Passmark_integer_math shows 123774 versus 104943, a 17.9% lead. Passmark_extended_instructions gives 31663 versus 28609, a 10.7% edge. Passmark_physics shows 2583 versus 2391, an 8% lead. Passmark_random_string_sorting gives 46238 versus 44400, a 4.1% margin. Passmark_single_thread shows 4408 versus 4338, a 1.6% lead.

The Intel wins are equally pronounced. Passmark_floating_point_math is the largest gap: the Core Ultra 7 265T scores 129817 against the EPYC 4345P’s 74255, a 42.8% lead. Passmark_find_prime_numbers shows 322 versus 167, a massive 48.1% advantage. Passmark_data_encryption gives 29687 versus 23313, a 21.5% lead. Passmark_multithread shows 37084 versus 36123, a 2.6% edge.

Overall average benchmark scores put the EPYC 4345P at 48470, slightly ahead of the Core Ultra 7 265T’s 47697. The AMD part’s nearest rivals include the Intel Core i5-14600K with an average score of 48618 and a delta of -0.3%, and the Intel Core Ultra 5 245HX at 48287 with a delta of 0.4%. The Intel part’s nearest rivals include the AMD Ryzen 9 PRO 5945 at 47527 with a delta of 0.4% and the AMD Ryzen 9 7900X3D at 47908 with a delta of -0.4%.

FAQ

Q: Which processor has a higher single-core performance?

A: The AMD EPYC 4345P leads in every single-core benchmark. In Cinebench R23 single-core, it scores 4520 versus the Intel Core Ultra 7 265T’s 4455, a 1.5% advantage. Passmark single-thread shows 4408 versus 4338, a 1.6% lead.

Q: Why does the Intel Core Ultra 7 265T win in floating-point math?

A: The Intel chip scores 129817 in passmark_floating_point_math, which is 42.8% higher than the AMD chip’s 74255. This large margin likely stems from the Intel part’s 20-core configuration and larger L1 and L2 cache pools per core.

Q: Does the AMD EPYC 4345P support ECC memory?

A: Yes, the AMD EPYC 4345P supports ECC memory. The Intel Core Ultra 7 265T does not support ECC memory, which makes the AMD part more suitable for server and workstation environments where data integrity is critical.

Q: How do the core counts compare between the two processors?

A: The AMD EPYC 4345P has 8 cores and 16 threads, while the Intel Core Ultra 7 265T has 20 cores and 20 threads. The Intel chip has more physical cores but no hyperthreading, resulting in equal core and thread counts.

Q: Which processor has higher memory bandwidth?

A: The Intel Core Ultra 7 265T has higher memory bandwidth at 102.4 GB/s, compared to the AMD EPYC 4345P’s 89.6 GB/s. Both support DDR5 memory with dual-channel buses.

Q: What are the average benchmark scores for each processor?

A: The AMD EPYC 4345P has an average benchmark score of 48470, while the Intel Core Ultra 7 265T has an average score of 47697. Both place in the 90th percentile among all CPUs.

Specification Differences

The two processors differ in several key specifications. The AMD EPYC 4345P has 8 cores and 16 threads, while the Intel Core Ultra 7 265T has 20 cores and 20 threads. Base clocks differ significantly: the AMD chip runs at 3.80 GHz, while the Intel chip runs at 1.50 GHz. Boost clocks are closer, with the AMD chip at 5.50 GHz and the Intel chip at 5.30 GHz. The TDP ratings differ substantially: the AMD chip has a 65 TDP, while the Intel chip has a 35 TDP.

The process nodes differ: the AMD chip uses a 4 nm process, while the Intel chip uses a 3 nm process. Transistor counts differ: the AMD chip has 8,315 million transistors, while the Intel chip has 17,800 million. Die sizes differ: the AMD chip is 70.6 mm², while the Intel chip is 243 mm². Cache configurations differ: the AMD chip has 80 KB L1 per core, 1 MB L2 per core, and 32 MB shared L3. The Intel chip has 192 KB L1 per core, 3 MB L2 per core, and 30 MB shared L3.

Memory bandwidth differs: the AMD chip has 89.6 GB/s, while the Intel chip has 102.4 GB/s. ECC support differs: the AMD chip supports ECC, while the Intel chip does not. PCIe lanes differ: the AMD chip offers 24 Gen 5 lanes, while the Intel chip offers 20 Gen 5 lanes. The AMD chip targets the Server/Workstation market segment, while the Intel chip targets Desktop. The AMD chip launched on Socket AM5, while the Intel chip uses Socket 1851.

The Verdict

The data points to a clear split based on workload requirements. For server and workstation environments where ECC memory support is essential, the AMD EPYC 4345P is the only choice between these two, as the Intel Core Ultra 7 265T lacks ECC support. The AMD chip also wins the majority of benchmark tests, including all Cinebench rendering tests and integer-heavy Passmark tests. Its 17.9% lead in integer math and 13.9% lead in data compression make it well-suited for general productivity and data processing tasks.

The Intel Core Ultra 7 265T is the better option for floating-point-intensive workloads. Its 42.8% lead in floating-point math and 48.1% lead in prime number finding are substantial. The 21.5% lead in data encryption also makes it attractive for cryptographic workloads. Its lower TDP of 35 watts, compared to the AMD chip’s 65 watts, suggests better power efficiency for sustained floating-point operations.

For multi-threaded performance, the results are mixed. The Intel chip leads in passmark_multithread by 2.6%, but the AMD chip leads in all Cinebench multi-core tests by 1.5%. The AMD chip’s higher boost clock of 5.50 GHz likely contributes to its Cinebench advantage, while the Intel chip’s 20 physical cores provide an edge in passmark_multithread.

The average benchmark scores are close: the AMD chip averages 48470, while the Intel chip averages 47697, a difference of less than 2%. Both processors sit at the 90th percentile of all CPUs. The AMD EPYC 4345P wins 13 of 17 head-to-head tests, while the Intel Core Ultra 7 265T wins 4. However, the margins of the Intel wins are often larger, particularly in floating-point math and prime number finding.

The choice ultimately depends on the specific workload mix. For general server workloads, ECC support, and integer-heavy applications, the AMD EPYC 4345P is the stronger candidate. For floating-point-heavy computation, encryption, and lower power consumption, the Intel Core Ultra 7 265T is the better fit. The AMD chip’s higher base clock and boost clock make it more responsive in single-threaded tasks, while the Intel chip’s core count provides raw parallel throughput in specific scenarios.

DETAILED SPECIFICATIONS

SPECIFICATION
EPYC 4345P
Ultra 7 265T
Core Specs
Cores
8
20 +150.0%
Threads
16
20 +25.0%
Base Clock (GHz)
3.8
1.5 -60.5%
Boost Clock (GHz)
5.5
5.3 -3.6%
Frequency (GHz)
3.8
1.5 -60.5%
Turbo Clock (GHz)
5.5
5.3 -3.6%
Multiplier
38
15 -60.5%
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
32 MB (shared)
30 MB (shared)
Power
TDP (W)
65
35 -46.2%
PL1
—
35 W
PL2
—
112 W
PPT
88 W
—
Architecture
Architecture
Zen 5
Arrow Lake
Codename
Grado
Arrow Lake-S
Generation
EPYC (Zen 5 (Grado))
Ultra 7 (Arrow Lake)
Process Size
4 nm
3 nm
Transistors
8,315 million
17,800 million
Die Size
70.6 mm²
243 mm²
Foundry
TSMC
TSMC
Memory
Memory Support
DDR5
DDR5
Memory Bus
Dual-channel
Dual-channel
Memory Bandwidth
89.6 GB/s
102.4 GB/s
ECC Memory
Yes
No
Platform
Socket
AMD Socket AM5
Intel Socket 1851
Chipsets
—
Z890, B860, W880, Q870, H810
PCIe
Gen 5, 24 Lanes(CPU only)
Gen 5, 20 Lanes(CPU only)
Intel Hybrid
Hybrid Cores
—
P-Cores: 8 E-Cores: 12
E-Core Frequency
—
1200 MHz up to 4.6 GHz
P-Core Turbo
—
5.2 GHz
AMD Multi-Die
IO Process Size
6 nm
—
Graphics
Integrated Graphics
Radeon Graphics
Arc Xe-LPG Graphics 64EU
Other
Market
Server/Workstation
Desktop
Production Status
Active
Active
Launch Price
$329
$384
Part Number
100-000001556
SRQCS
Package
FC-LGA1718
FC-LGA18W
Tj Max
95°C
105°C
Bundled Cooler
None
—
View EPYC 4345P Details View Core Ultra 7 265T Details