Intel Xeon 6736P vs Intel Xeon w5-3535X Comparison

Intel
INTEL

Intel Xeon 6736P

CORE STATE Granite Rapids
CORE SPECS 36 Cores / 72 Threads
CLOCK SPEED 2 Base / 4.1 GHz Turbo
CACHE 144 MB (shared)
MAX TDP 205W
ARCHITECTURE Granite Rapids
nm
PROCESS 5 nm
LAUNCH DATE 2025
VS
Intel
INTEL

Xeon w5-3535X

CORE STATE Sapphire Rapids
CORE SPECS 20 Cores / 40 Threads
CLOCK SPEED 2.9 Base / 4.8 GHz Turbo
CACHE 52.5 MB
MAX TDP 300W
ARCHITECTURE Sapphire Rapids
nm
PROCESS 10 nm
LAUNCH DATE 2024

PERFORMANCE BENCHMARKS

cinebench_cinebench_r15_multicore
4,290
4,634
cinebench_cinebench_r15_singlecore
605
654
cinebench_cinebench_r20_multicore
17,875
19,309
cinebench_cinebench_r20_singlecore
2,523
2,725
cinebench_cinebench_r23_multicore
42,561
45,974
cinebench_cinebench_r23_singlecore
6,008
6,490
passmark_data_compression
796,658
731,388
passmark_data_encryption
46,236
36,784
passmark_extended_instructions
55,563
60,183
passmark_find_prime_numbers
392
269
passmark_floating_point_math
149,770
145,924
passmark_integer_math
206,833
186,158
passmark_multithread
50,072
54,088
passmark_physics
6,531
3,547
passmark_random_string_sorting
103,723
73,618
passmark_single_thread
2,024
3,602
passmark_singlethread
2,024
3,602

Analysis: Intel Xeon 6736P vs Intel Xeon w5-3535X

Where Each One Wins

The benchmark split between these two Xeon parts is not subtle. The Intel Xeon w5-3535X takes the multi-core rendering crown, winning every Cinebench iteration by a consistent 7.4% to 7.5% margin. It also dominates single-threaded workloads with a 43.8% lead in PassMark single-thread scoring, which is a massive gap for two server-class processors. The w5-3535X also wins PassMark multithread by 7.4% and extended instructions by 7.7%. If your work is built around Cinebench-style rendering, threaded productivity, or instruction-heavy code paths, the w5-3535X is the clear pick.

The Intel Xeon 6736P fights back in a different arena. It wins 7 of the 17 head-to-head tests, and the wins are not marginal. Data encryption shows a 25.7% advantage, random string sorting is 40.9% ahead, find prime numbers is 45.7% ahead, and physics simulation is a staggering 84.1% ahead. Data compression, integer math, and floating point math also go its way, with leads of 8.9%, 11.1%, and 2.6% respectively. This is a processor that excels at encryption, compression, sorting, and scientific simulation workloads. The data suggests a clear division: the w5-3535X for latency-sensitive and clock-driven tasks, the 6736P for throughput-heavy data processing.

Architecture Differences

The two chips come from different Intel generations and manufacturing nodes. The Xeon 6736P is built on Granite Rapids architecture at a 5 nm process node, while the w5-3535X uses Sapphire Rapids on a 10 nm node. That process advantage helps the 6736P pack 36 cores and 72 threads into a 598 mm² die, versus the w5-3535X's 20 cores and 40 threads spread across a 4x 477 mm² multi-die design. Core count is not the only difference. The 6736P has 144 MB of shared L3 cache and 112 KB of L1 cache per core, while the w5-3535X offers 52.5 MB of L3 and 80 KB of L1 per core. Both use 2 MB of L2 per core.

Clock speeds tell the opposite story. The w5-3535X runs a 2.90 GHz base clock and boosts to 4.80 GHz, while the 6736P starts at 2.00 GHz and boosts to 4.10 GHz. That 0.70 GHz boost advantage explains much of the w5-3535X's single-thread dominance. The w5-3535X also has an unlocked multiplier, whereas the 6736P is locked. Power envelopes differ sharply: the 6736P is rated at 205 W TDP, the w5-3535X at 300 W TDP. Memory bandwidth favors the 6736P at 409.6 GB/s versus 307.2 GB/s, though both support DDR5 across eight channels. PCIe lane counts also diverge, with the w5-3535X offering 112 Gen 5 lanes versus 88 on the 6736P. Both use different sockets: Socket 4710 for the 6736P, Socket 4677 for the w5-3535X.

Head-to-Head Benchmarks

The w5-3535X sweeps the Cinebench suite. In Cinebench R15 multi-core, it scores 4634 against 4290, a 7.4% win. Single-core R15 shows 654 versus 605, a 7.5% margin. The pattern holds in R20 multi-core (19309 vs 17875, 7.4%) and R20 single-core (2725 vs 2523, 7.4%). Cinebench R23 multi-core sees 45974 versus 42561, again 7.4%, and R23 single-core is 6490 versus 6008, also 7.4%. The consistency across every Cinebench variant points to a fundamental clock advantage rather than workload-specific behavior.

PassMark multithread goes to the w5-3535X at 54088 versus 50072, a 7.4% lead. Extended instructions follow at 60183 versus 55563, a 7.7% edge. The single-thread result is the most lopsided of the entire comparison: 3602 versus 2024, a 43.8% gap. That is not a small difference; it means the w5-3535X delivers well over a third more single-thread performance in this test.

The 6736P answers with its own set of decisive wins. Physics simulation shows 6531 versus 3547, an 84.1% advantage. Find prime numbers delivers 392 versus 269, a 45.7% lead. Random string sorting is 103723 versus 73618, a 40.9% edge. Data encryption comes in at 46236 versus 36784, a 25.7% win. Integer math is 206833 versus 186158, an 11.1% lead. Data compression is 796658 versus 731388, an 8.9% win. Floating point math is closer at 149770 versus 145924, a 2.6% edge.

Looking at the aggregate, the 6736P posts an average benchmark score of 87864 against 81115 for the w5-3535X, and its nearest rivals include the Xeon 6731P at 87756 (0.1% behind) and the Xeon w7-2575X at 88172 (0.3% ahead). The w5-3535X sits near the Core i9-14900KS at 81127 and the Ryzen 9 8940HX at 81103, both essentially tied. Both CPUs rank in the 96th percentile of all CPUs, so neither is a weak performer, but their strengths are clearly separated.

FAQ

Q: Which CPU wins in Cinebench multi-core rendering?

A: The Intel Xeon w5-3535X wins every Cinebench multi-core test by 7.4% across R15, R20, and R23.

Q: Which CPU is better for data encryption workloads?

A: The Intel Xeon 6736P is 25.7% ahead of the w5-3535X in PassMark data encryption.

Q: What is the single-thread performance gap between the two?

A: The w5-3535X leads by 43.5% in PassMark single-thread scoring, with a score of 3602 versus 2024.

Q: Which CPU has more cores and threads?

A: The Xeon 6736P has 36 cores and 72 threads, compared to 20 cores and 40 threads on the w5-3535X.

Q: How do their memory bandwidth figures compare?

A: The 6736P provides 409.6 GB/s of memory bandwidth, while the w5-3535X provides 307.2 GB/s.

Q: Which CPU is better for physics simulation?

A: The 6736P leads by 84.1% in PassMark physics, scoring 6531 versus 3547.

Specification Differences

The two processors differ across nearly every major specification. The Xeon 6736P uses 36 cores and 72 threads, while the w5-3535X uses 20 cores and 40 threads. Base clocks are 2.00 GHz versus 2.90 GHz, and boost clocks are 4.10 GHz versus 4.80 GHz. TDP is 205 W versus 300 W. The 6736P runs on Socket 4710, the w5-3535X on Socket 4677. Architecture is Granite Rapids versus Sapphire Rapids, with process nodes of 5 nm versus 10 nm. Die size is 598 mm² versus 4x 477 mm². L1 cache is 112 KB per core versus 80 KB per core, while L3 cache is 144 MB shared versus 52.5 MB. Memory bandwidth is 409.6 GB/s versus 307.2 GB/s. PCIe lanes are 88 lanes versus 112 lanes. The w5-3535X has an unlocked multiplier, the 6736P is locked. Launch dates differ too, with the 6736P launching on February 23, 2025 and the w5-3535X on August 23, 2024. The launch MSRP of the 6736P is $3351, while the w5-3535X is $1689.

The Verdict

Choose the Intel Xeon w5-3535X if your workload is dominated by single-threaded tasks, Cinebench rendering, or code that is sensitive to clock speed. The 43.8% single-thread lead and consistent 7.4% Cinebench advantage make it the better fit for interactive work or lightly threaded applications. The unlocked multiplier is a bonus for those willing to push clocks further, though the 300 W TDP should be accounted for in cooling and power planning.

Choose the Intel Xeon 6736P if your tasks involve encryption, compression, sorting, physics simulation, or integer-heavy math. The 84.1% physics lead, 45.7% prime-number advantage, and 25.7% encryption edge are decisive for scientific computing and data processing. The 36 cores, 144 MB of L3 cache, and 409.6 GB/s memory bandwidth provide the raw throughput that many server workloads require. The 205 W TDP also makes it the more power-efficient option for dense deployments. The two chips are not interchangeable; they are aimed at different corners of the workstation and server market, and the benchmark data makes that split clear.

DETAILED SPECIFICATIONS

SPECIFICATION
6736P
w5-3535X
Core Specs
Cores
36
20 -44.4%
Threads
72
40 -44.4%
Base Clock (GHz)
2
2.9 +45.0%
Boost Clock (GHz)
4.1
4.8 +17.1%
Frequency (GHz)
2
2.9 +45.0%
Turbo Clock (GHz)
4.1
4.8 +17.1%
Multiplier
20
29 +45.0%
SMP CPUs
2
1 -50.0%
Cache
L1 Cache
112 KB (per core)
80 KB (per core)
L2 Cache
2 MB (per core)
2 MB (per core)
L3 Cache
144 MB (shared)
52.5 MB
Power
TDP (W)
205
300 +46.3%
Architecture
Architecture
Granite Rapids
Codename
Granite Rapids
Sapphire Rapids
Generation
Xeon 6 (Granite Rapids-SP)
Xeon W (Sapphire Rapids)
Process Size
5 nm
10 nm
Die Size
598 mm²
4x 477 mm²
Foundry
Intel
Intel
Memory
Memory Support
DDR5
DDR5
Memory Bus
Eight-channel
Eight-channel
Memory Bandwidth
409.6 GB/s
307.2 GB/s
ECC Memory
Yes
Yes
Platform
Socket
Intel Socket 4710
Intel Socket 4677
PCIe
Gen 5, 88 Lanes(CPU only)
Gen 5, 112 Lanes(CPU only)
DMI
4.0 x8
AMD Multi-Die
IO Process Size
10 nm
Interconnect
UPI Links
4 x24 24 GT/s
CXL
Gen 2.0, 64 Lanes (Shared with PCI-E)
Other
Market
Server/Workstation
Server/Workstation
Production Status
Active
Active
Launch Price
$3351
$1689
Part Number
SRVNW
SRN76
Package
FC-LGA18N
FC-LGA16A
Tj Max
98°C
Bundled Cooler
None
View Xeon 6736P Details View Xeon w5-3535X Details