AMD EPYC 9275F vs Intel Xeon 658X Comparison

AMD
AMD

AMD EPYC 9275F

CORE STATE Turin
CORE SPECS 24 Cores / 48 Threads
CLOCK SPEED 4.1 Base / 4.8 GHz Turbo
CACHE 256 MB (shared)
MAX TDP 320W
ARCHITECTURE Zen 5
nm
PROCESS 4 nm
LAUNCH DATE 2024
VS
Intel
INTEL

Xeon 658X

CORE STATE Granite Rapids
CORE SPECS 24 Cores / 48 Threads
CLOCK SPEED 3 Base / 4.9 GHz Turbo
CACHE 144 MB (shared)
MAX TDP 250W
ARCHITECTURE Granite Rapids
nm
PROCESS 5 nm
LAUNCH DATE 2026

PERFORMANCE BENCHMARKS

cinebench_cinebench_r15_multicore
7,250
6,296
cinebench_cinebench_r15_singlecore
1,023
888
cinebench_cinebench_r20_multicore
30,209
26,235
cinebench_cinebench_r20_singlecore
4,264
3,703
cinebench_cinebench_r23_multicore
71,927
62,466
cinebench_cinebench_r23_singlecore
10,154
8,818
passmark_data_compression
1,212,560
1,062,062
passmark_data_encryption
62,664
52,357
passmark_extended_instructions
94,889
84,626
passmark_find_prime_numbers
991
649
passmark_floating_point_math
201,888
210,480
passmark_integer_math
317,777
263,995
passmark_multithread
84,620
73,490
passmark_physics
12,089
6,470
passmark_random_string_sorting
144,037
103,028
passmark_single_thread
3,810
3,728
passmark_singlethread
3,810
3,728

Analysis: AMD EPYC 9275F vs Intel Xeon 658X

Head-to-Head Benchmarks

The recorded data is decisively one-sided: the AMD EPYC 9275F wins 16 of the 17 head-to-head tests, with the Intel Xeon 658X taking a single victory. The overall average benchmark score tells the same story, the EPYC 9275F posts 133,174 against 116,060 for the Xeon 658X, a gap of roughly 14.7 percent.

The largest margin comes in the PassMark physics test. The EPYC 9275F scores 12,089 against 6,470, a delta of 86.8 percent. That is nearly double the throughput. The PassMark find prime numbers test also shows a wide divide: 991 versus 649, a 52.7 percent advantage for the AMD part. Random string sorting favors the EPYC 9275F by 39.8 percent, with scores of 144,037 and 103,028 respectively.

In the Cinebench suite, the AMD processor is consistently ahead by around 15 percent. In Cinebench R15 multi-core, the EPYC 9275F scores 7,250 versus 6,296, a 15.2 percent lead. The single-core result is identical in percentage terms: 1,023 versus 888. Cinebench R20 multi-core shows 30,209 against 26,235, a 15.1 percent edge, and the single-core test has 4,264 versus 3,703, also 15.1 percent. Cinebench R23 multi-core delivers 71,927 against 62,466, 15.1 percent, while single-core is 10,154 versus 8,818, a 15.2 percent margin.

PassMark integer math favors the EPYC 9275F by 20.4 percent: 317,777 versus 263,995. Data encryption is a 19.7 percent win for AMD, with 62,664 against 52,357. Data compression shows a 14.2 percent edge: 1,212,560 versus 1,062,062. Extended instructions land at 94,889 versus 84,626, a 12.1 percent lead. The multithread score is 84,620 versus 73,490, a 15.1 percent advantage.

The single Intel victory is in floating point math. The Xeon 658X scores 210,480 against 201,888, a 4.1 percent lead. That is the only test where the Intel part comes out ahead. Even in single-threaded PassMark, where the margins are tighter, the EPYC 9275F wins: 3,810 versus 3,728, a 2.2 percent edge. The same result appears in the PassMark single thread test, identical scores and delta.

The nearest rival data places the EPYC 9275F at the 97th percentile among all CPUs, with the closest competitor being the Intel Xeon 6710E at 129,930 average score, a 2.5 percent gap. The Xeon 658X also sits at the 97th percentile, but its nearest rival is the AMD EPYC 9255 at 116,388, a 0.3 percent difference in favor of the Intel part. The Xeon 658X trails the Intel Xeon w7-3565X by 1.9 percent and the AMD EPYC 9384X by 3.6 percent.

FAQ

Q: Which processor has the higher boost clock?

A: The Intel Xeon 658X has a boost clock of 4.90 GHz, while the AMD EPYC 9275F boosts to 4.80 GHz. However, the AMD part has a much higher base clock of 4.10 GHz compared to 3.00 GHz for the Intel.

Q: How do the two compare in terms of memory bandwidth?

A: The AMD EPYC 9275F supports twelve-channel memory with a bandwidth of 576.0 GB/s. The Intel Xeon 658X uses eight-channel memory with a bandwidth of 409.6 GB/s. Both support DDR5 and ECC memory.

Q: Which one wins in the PassMark physics test?

A: The AMD EPYC 9275F wins decisively with a score of 12,089 against 6,470 for the Intel Xeon 658X, a delta of 86.8 percent. This is the largest margin in any test between the two.

Q: Is there any workload where the Intel Xeon 658X beats the AMD EPYC 9275F?

A: Yes, in floating point math. The Intel part scores 210,480 against 201,888, a 4.1 percent lead. That is the only recorded benchmark where the Intel processor outperforms the AMD one.

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

A: The AMD EPYC 9275F has an average benchmark score of 133,174. The Intel Xeon 658X averages 116,060. The AMD part is roughly 14.7 percent higher.

Q: Do both processors have the same number of cores and threads?

A: Yes, both have 24 cores and 48 threads. The difference lies in clock speeds, cache, memory channels, and architecture, not in core count.

Where Each One Wins

The AMD EPYC 9275F is the clear winner for compute-heavy, multi-threaded workloads. The PassMark multithread score of 84,620 versus 73,490 shows a 15.1 percent advantage. The Cinebench multi-core results, all around 15 percent higher, reinforce this. For tasks like data compression, encryption, and integer math, the AMD part leads by margins from 12.1 to 20.4 percent. The physics test, with an 86.8 percent delta, suggests a major advantage for simulation or physics-based workloads.

The Intel Xeon 658X has a narrow niche in floating point math, where it leads by 4.1 percent. That could matter for scientific computing or financial modeling that relies heavily on FPU throughput. However, this is a single test, and the overall pattern does not favor Intel.

For single-threaded performance, the AMD EPYC 9275F wins by 2.2 percent in PassMark single thread, but the Cinebench single-core tests show a much larger 15.2 percent edge. The AMD part is also ahead in data compression, which is relevant for database and storage workloads, and in random string sorting, where it leads by 39.8 percent.

The Intel processor has a higher boost clock, 4.90 GHz versus 4.80 GHz, but this does not translate into wins outside of floating point. The base clock difference, 4.10 GHz versus 3.00 GHz, likely explains the AMD advantage in sustained workloads.

Specification Differences

The two processors differ in several key specifications. The AMD EPYC 9275F has a base clock of 4.10 GHz and a boost clock of 4.80 GHz. The Intel Xeon 658X has a base clock of 3.00 GHz and a boost clock of 4.90 GHz. The AMD part has a TDP of 320 watts, while the Intel part is rated at 250 watts.

The AMD EPYC 9275F uses the AMD Socket SP5, while the Intel Xeon 658X uses the Intel Socket 4710. The AMD processor has an unlocked multiplier set to false, meaning it is locked, while the Intel part has an unlocked multiplier set to true.

Memory support differs: the AMD part uses twelve-channel DDR5 with 576.0 GB/s bandwidth, the Intel part uses eight-channel DDR5 with 409.6 GB/s. Both support ECC memory and PCIe Gen 5 with 128 lanes (CPU only).

The process node differs: the AMD EPYC 9275F is built on a 4 nm process by TSMC, while the Intel Xeon 658X uses a 5 nm process by Intel. The die size is also different: the AMD part has 8x 70.6 mm² dies, the Intel part has 2x 598 mm² dies.

The release dates are different: the AMD EPYC 9275F was released on 2024-10-09, while the Intel Xeon 658X has a release date of 2026-02-01. The launch MSRP for the AMD part is $3439, and for the Intel part it is $1699.

Architecture Differences

The AMD EPYC 9275F is based on the Zen 5 architecture, codenamed Turin, and is part of the EPYC 9005 series. It uses a 4 nm process from TSMC and has 66,520 million transistors. The cache hierarchy is: 80 KB of L1 per core, 1 MB of L2 per core, and 256 MB of shared L3 cache.

The Intel Xeon 658X is based on the Granite Rapids architecture, specifically Granite Rapids-WS, and is part of the Xeon 600 series. It uses a 5 nm process from Intel. The die size is 2x 598 mm². The cache layout is: 112 KB of L1 per core, 2 MB of L2 per core, and 144 MB of shared L3 cache.

The AMD part has a larger L3 cache, 256 MB versus 144 MB, but the Intel part has larger L1 and L2 caches per core. The Intel part has a larger die area, suggesting more complex physical design, but the AMD part uses a smaller process node and has a transistor count listed, while the Intel part does not have a transistor count in the data.

Both support the same memory type, DDR5, and both have no integrated graphics. Both are listed as active production parts, targeting the server and workstation market segment.

The Verdict

The data points to the AMD EPYC 9275F as the stronger processor for almost all measured workloads. With 16 wins out of 17 tests, including all Cinebench multi-core and single-core benchmarks, the AMD part is consistently ahead. The largest margins, 86.8 percent in physics and 52.7 percent in find prime numbers, show a substantial performance advantage in specific compute tasks.

The Intel Xeon 658X has a single win in floating point math, but this does not offset the overall pattern. The AMD part is ahead by double digits in most tests, and even in single-threaded PassMark, where the margin is smallest, it still wins.

For users prioritizing raw compute performance, multi-threaded throughput, or memory bandwidth, the AMD EPYC 9275F is the clear choice. The twelve-channel memory support and larger L3 cache likely contribute to its wins in data-heavy tests like compression and encryption. The higher base clock also helps in sustained workloads.

The Intel Xeon 658X could be considered for workloads dominated by floating point math, where it leads by 4.1 percent. It also has a lower TDP, 250 watts versus 320 watts, which may matter for power-constrained environments. The lower launch MSRP, $1699 versus $3439, is a notable difference, though the performance data does not favor the Intel part.

In short, the AMD EPYC 9275F is the recommended processor for general server and workstation use, based on the benchmark results. The Intel Xeon 658X is a viable alternative only for specific floating-point-heavy tasks or where power consumption and cost are primary concerns.

DETAILED SPECIFICATIONS

SPECIFICATION
EPYC 9275F
658X
Core Specs
Cores
24
24 0.0%
Threads
48
48 0.0%
Base Clock (GHz)
4.1
3 -26.8%
Boost Clock (GHz)
4.8
4.9 +2.1%
Frequency (GHz)
4.1
3 -26.8%
Turbo Clock (GHz)
4.8
4.9 +2.1%
Multiplier
41
30 -26.8%
SMP CPUs
2
1 -50.0%
Cache
L1 Cache
80 KB (per core)
112 KB (per core)
L2 Cache
1 MB (per core)
2 MB (per core)
L3 Cache
256 MB (shared)
144 MB (shared)
Power
TDP (W)
320
250 -21.9%
Configurable TDP
320-400 W
—
Architecture
Architecture
Zen 5
Granite Rapids
Codename
Turin
Granite Rapids
Generation
EPYC (Zen 5 (Turin))
Xeon 600 (Granite Rapids-WS)
Process Size
4 nm
5 nm
Transistors
66,520 million
—
Die Size
8x 70.6 mm²
2x 598 mm²
Foundry
TSMC
Intel
Memory
Memory Support
DDR5
DDR5
Memory Bus
Twelve-channel
Eight-channel
Memory Bandwidth
576.0 GB/s
409.6 GB/s
ECC Memory
Yes
Yes
Platform
Socket
AMD Socket SP5
Intel Socket 4710
Chipsets
—
W890
PCIe
Gen 5, 128 Lanes(CPU only)
Gen 5, 128 Lanes(CPU only)
AMD Multi-Die
IO Process Size
6 nm
10 nm
Interconnect
CXL
Gen 2.0
Gen 2.0 (Shared with PCI-E)
Other
Market
Server/Workstation
Server/Workstation
Production Status
Active
Active
Launch Price
$3439
$1699
Part Number
100-000001144
SA2D2
Package
FC-LGA6096
FC-LGA18N
Tj Max
—
99°C
Bundled Cooler
—
None
View EPYC 9275F Details View Xeon 658X Details