Intel Xeon D-1518 vs Intel Xeon E5640 Comparison

FAQ

Q: Which processor is faster in multi-core rendering workloads?

A: The Intel Xeon D-1518 wins every recorded multi-core Cinebench test. In Cinebench R23 multi-core, the D-1518 scores 3877 versus 3303 for the E5640, a 14.8% lead. The margin is similar in Cinebench R20 multi-core (1628 versus 1387, also 14.8%).

Q: How do the two compare in single-core performance?

A: The D-1518 is again ahead. In Cinebench R23 single-core, the D-1518 scores 547 versus 466 for the E5640, a 14.8% difference. The R20 single-core test shows 229 versus 195, another 14.8% gap. The E5640 has no recorded single-core result for Cinebench R15, while the D-1518 scores 54 in that test.

Q: What are their average benchmark scores and percentiles?

A: The E5640 has an average benchmark score of 1137 and sits at the 32nd percentile of all CPUs. The D-1518 has an average score of 1121, also at the 32nd percentile. Despite the D-1518 winning every head-to-head test, the two land in the same percentile class and have nearly identical average scores.

Q: Which processor has a higher base clock?

A: The E5640 has a base clock of 2.67 GHz. The D-1518 has a base clock of 2.20 GHz. However, the E5640 also has a boost clock of 2.93 GHz, while the D-1518 has no recorded boost clock in the database.

Q: Are both processors unlocked for overclocking?

A: No. Neither processor has an unlocked multiplier. Both are locked parts, so the E5640 and D-1518 run at their factory-defined clock settings.

Q: What memory types do they support?

A: The E5640 supports DDR3 memory, and the memory bus is triple-channel. The D-1518 supports both DDR3 and DDR4, with a dual-channel memory bus. Both support ECC memory.

Architecture Differences

The two Xeons come from different eras of Intel's server roadmap. The E5640 is built on Westmere-EP, a 32 nm process with 1,170 million transistors on a 239 mm² die. The D-1518 is a Broadwell-DE part fabricated on 14 nm process with 3,200 million transistors on a 246 mm² die. The D-1518 transistors count is about 2.7 times higher in transistor count, packed into a slightly larger die area, 246 mm². Both processors share the same per-core L1 and L2 cache sizes, 64 KB per core for L1 and 256 KB per core for L2. The L3 cache structure differs in allocation. The E5640 has 12 MB of shared L3 cache, while the D-1518 uses a per-core L3 allocation of 1.5 MB per core. With 4 cores, that is 6 MB total L3 for the D-1518, so the E5640 holds twice the aggregate L3 capacity.

The socket and platform also differ substantially. The E5640 uses Intel Socket 1366, while the D-1518 uses Intel BGA 1667. The E5640 supports PCIe Gen 2, while the D-1518 supports PCIe Gen 3 with 24 lanes (CPU only). Memory support diverges: the E5640 is limited to DDR3 with triple-channel access, whereas the D-1518 supports both DDR3 and DDR4 with dual-channel access. The D-1518 has a much lower TDP, 35 watts versus 80 watts for the E5640.

The D-1518's process advantage, 14 nm versus 32 nm, is the clearest architectural differentiator. It allows a substantially higher transistor count, 3,200 million versus 1,170 million, in a similar die area, 246 mm² versus 239 mm², while consuming less than half the power, 35 watts versus 80 watts. The D-1518 also has a more modern PCIe generation, Gen 3 versus Gen 2. The E5640 was released in March 2010 and is end-of-life, while the D-1518 was released in March 2015 and is still listed as active production.

Head-to-Head Benchmarks

The recorded head-to-head results all point in the same direction: the D-1518 wins every test by a nearly uniform margin. In Cinebench R15 multi-core, the D-1518 scores 390 against 332 for the E5640, a 14.9% advantage. The R20 multi-core test shows 1628 versus 1387, a 14.8% lead. The R23 multi-core test shows 3877 versus 3303, also 14.8% ahead. These consistent margins across three generations of Cinebench suggest the D-1518's architectural advantages translate into a steady multi-threaded performance uplift rather than a workload-specific spike.

Single-core results mirror the multi-core pattern. In Cinebench R20 single-core, the D-1518 posts 229 versus 195, a 14.8% edge. In R23 single-core, the D-1518 reaches 547 versus 466, again 14.8% ahead. The D-1518 also has a recorded R15 single-core score of 54, a test the E5640 has no recorded result for. The consistency of the delta, always within a tenth of a point of 14.8%, indicates that the D-1518's per-core efficiency advantage is uniform across the benchmark suite.

The E5640's closest rivals in the database are the Intel Core i5-2550K and Intel Core i5-3570T, both with an average score of 1136, just 0.1% behind the E5640's 1137. The AMD Athlon Silver PRO 3125GE scores 1138, 0.1% ahead. The Intel Core i3-N300 scores 1135, 0.2% behind. These near-ties place the E5640 in a dense cluster of mid-range desktop and server parts.

The D-1518's nearest rivals include the AMD Athlon 240GE with an identical average score of 1121, a 0% delta. The Intel Core i7-2860QM scores 1120, 0.1% behind, and the Intel Core i3-4150 scores 0.1% behind at 1119, while the Intel Core i5-3550S is 0.2% behind at 1118. The D-1518 sits just above the E5640's nearest-rival cluster in raw score difference, 1121 versus 1137, yet wins every head-to-head test. This is because the head-to-head tests are a direct comparison of the two parts themselves, not their positions against other CPUs.

The average benchmark scores, 1137 for the E5640 and 1121 for the D-1518, initially appear to contradict the head-to-head data, but they do not. The head-to-head tests are a direct comparison of the two parts, and the D-1518 wins all of them. The average score is an aggregate across all recorded tests for each processor, and the two parts land within 1.4% of each other in overall average, 1137 versus 1121. The D-1518's win rate in the direct comparisons is 100%, with five wins and zero losses.

Specification Differences

The table below lists only the fields where the two processors differ.

| Specification | Intel Xeon E5640 | Intel Xeon D-1518 |

|---|---|---|

| Base clock | 2.67 GHz | 2.20 GHz |

| Boost clock | 2.93 GHz | Not recorded |

| TDP | 80 W | 35 W |

| Socket | Intel Socket 1366 | Intel BGA 1667 |

| Architecture | Westmere | Broadwell |

| Codename | Westmere-EP | Broadwell |

| Process node | 32 nm | 14 nm |

| Transistors | 1,170 million | 3,200 million |

| Die size | 239 mm² | 246 mm² |

| L3 cache | 12 MB (shared) | 1.5 MB (per core) |

| Memory support | DDR3 | DDR3, DDR4 |

| Memory bus | Triple-channel | Dual-channel |

| PCIe | Gen 2 | Gen 3, 24 lanes (CPU only) |

| Production status | End-of-life | Active |

| Release date | 2010-03-15 | 2015-03-08 |

| Part number | SLBVC | SR2DN |

The E5640 has a higher base clock, 2.67 GHz versus 2.20 GHz, and a boost clock of 2.93 GHz, while the D-1518 has no recorded boost clock. The D-1518 counters with a much smaller process node, 14 nm versus 32 nm, and a TDP of 35 watts versus 80 watts. The E5640's L3 cache is larger in aggregate, 12 MB shared, but the D-1518 arranges its L3 as 1.5 MB per core. The memory systems differ in both type and channel count, with the E5640 using triple-channel DDR3 and the D-1518 using dual-channel DDR3 or DDR4. The PCIe generation also differs, Gen 2 for the E5640 and Gen 3 for the D-1518. The D-1518 has a launch MSRP of $193.

The Verdict

The data points to a clear winner for raw benchmark performance. The Intel Xeon D-1518 wins all five recorded head-to-head Cinebench tests, with a consistent 14.8% advantage in four of the five and a 14.9% edge in the fifth. Its single-core and multi-core results both improve by this same margin, indicating a broad architectural uplift rather than a niche optimization. The D-1518 achieves this while drawing 35 watts against the E5640's 80 watts, a substantial efficiency advantage. It also uses a much newer 14 nm process, supports DDR4 memory, and offers PCIe Gen 3.

The E5640 is not without its own advantages. It has a higher base clock, 2.67 GHz versus 2.20 GHz, and a boost clock of 2.93 GHz. Its L3 cache is larger in aggregate, 12 MB shared versus the D-1518's per-core arrangement. It also supports triple-channel memory, which can be relevant for memory-bandwidth-sensitive workloads. The E5640's average benchmark score, 1137, is actually slightly higher than the D-1518's 1121, despite losing every direct head-to-head comparison. This reflects the fact that the average score aggregates all recorded tests for each part, and the two parts land in the same 32nd percentile.

For a user choosing between these two server processors, the D-1518 is the better performer in the recorded Cinebench suite and offers a dramatically lower power envelope. It is the only one of the two still in active production, and it supports modern DDR4 memory with PCIe Gen 3. The E5640's triple-channel DDR3 and larger shared L3 cache may appeal in scenarios where those specific features matter, but the benchmark data does not show a single test where the E5640 comes out ahead. The D-1518 wins every recorded benchmark, and that is the decisive fact in this comparison.