Intel Xeon D-1581 vs Intel Xeon E5-1660 v4 Comparison

The Intel Xeon E5-1660 v4 and Intel Xeon D-1581 are both 14 nm Broadwell-based server-class processors, but they target fundamentally different deployment scenarios. Benchmark data shows the E5-1660 v4 wins all six head-to-head Cinebench tests, yet the margins are consistently under 1% in every workload, making the D-1581’s higher core count nearly irrelevant in these specific tests. The E5-1660 v4 is the pick for raw single-core responsiveness and slightly higher multi-core throughput, while the D-1581 is the pick for dense, low-power embedded or server environments where its 65 W TDP and dual-channel memory flexibility matter more than marginal benchmark wins.

The Verdict

From the data, the Intel Xeon E5-1660 v4 is the outright winner in every measured benchmark, taking all six head-to-head Cinebench tests with wins ranging from 0.6% to 0.9%. The largest margin is in Cinebench R20 single-core, where the E5-1660 v4 scores 669 versus 663 for the D-1581, a 0.9% advantage. In multi-core tests, the E5-1660 v4 leads by 0.8% in Cinebench R15, R20, and R23, with scores of 1137 vs 1128, 4739 vs 4702, and 11285 vs 11197 respectively. Single-core R15 and R23 show the same pattern: 160 vs 159 (0.6%) and 1593 vs 1580 (0.8%).

However, the verdict is not simply “buy the E5-1660 v4.” The D-1581 matches the E5-1660 v4 in overall percentile ranking, with both sitting at the 53rd percentile among all CPUs. The average benchmark scores are nearly identical: 3264 for the E5-1660 v4 versus 3238 for the D-1581, a difference of just 0.8%. The D-1581’s nearest rivals include the AMD Ryzen 5 1600X at 3249 (0.3% behind) and the AMD Ryzen 7 4800U at 3218 (0.6% behind), placing it in the same performance band as the E5-1660 v4’s rivals, which include the Intel Xeon E5-1680 v3 at 3265 (0% delta) and the AMD Ryzen Embedded V2516 at 3277 (0.4% ahead). The E5-1660 v4 is for users who demand the absolute best single-threaded score and accept a 140 W TDP. The D-1581 is for users who need 16 cores and 32 threads in a 65 W envelope, even if those cores run slower.

Architecture Differences

Both processors share the same 14 nm Intel process node and a 246 mm² die size, but the transistor counts differ slightly: the E5-1660 v4 packs 3,400 million transistors, while the D-1581 has 3,200 million. The E5-1660 v4 uses the Broadwell-EP architecture (codename Broadwell-EP, generation “Xeon E5 (Broadwell-EP)”), while the D-1581 uses the Broadwell-DE architecture (codename Broadwell, generation “Xeon D (Broadwell-DE)”). This architectural split explains the core and cache differences.

The E5-1660 v4 has 8 cores and 16 threads, with a base clock of 3.20 GHz and a boost clock of 3.80 GHz. The D-1581 doubles the core count to 16 cores and 32 threads, but its clocks drop dramatically: 1800.00 MHz base and 2.40 GHz boost. This is a 44% reduction in base clock and a 37% reduction in boost clock when comparing the E5-1660 v4’s higher figures. Cache layouts also diverge: the E5-1660 v4 has 20 MB of shared L3 cache, while the D-1581 has 1.5 MB per core, which totals 24 MB across 16 cores but is not shared in the same way. Both have 64 KB L1 and 256 KB L2 per core.

Memory support differs significantly. The E5-1660 v4 supports DDR4 over a quad-channel bus with a memory bandwidth of 76.8 GB/s. The D-1581 supports both DDR3 and DDR4 over a dual-channel bus, with no bandwidth figure listed in the data. PCIe connectivity also differs: the E5-1660 v4 provides Gen 3 with 40 lanes (CPU only), while the D-1581 provides Gen 3 with 24 lanes (CPU only). The E5-1660 v4 uses Intel Socket 2011-3, whereas the D-1581 uses Intel BGA 1667, meaning the D-1581 is soldered rather than socketed, which affects upgradeability and system design. Both support ECC memory and lack integrated graphics, and both have locked multipliers.

Head-to-Head Benchmarks

The E5-1660 v4 wins every single head-to-head test, but the margins are razor-thin. The largest advantage appears in Cinebench R20 single-core, where the E5-1660 v4 scores 669 against the D-1581’s 663, a 0.9% delta. This is the only test where the gap exceeds 0.8%. In Cinebench R15 single-core, the E5-1660 v4 scores 160 versus 159, a 0.6% win. For multi-core workloads, the pattern is uniform: Cinebench R15 shows 1137 vs 1128 (0.8%), Cinebench R20 shows 4739 vs 4702 (0.8%), and Cinebench R23 shows 11285 vs 11197 (0.8%). The Cinebench R23 single-core test shows 1593 vs 1580, another 0.8% win.

The D-1581’s 16 cores and 32 threads do not translate into a multi-core advantage in these tests, which is surprising given the 2x core count difference. The D-1581’s much lower boost clock of 2.40 GHz versus the E5-1660 v4’s 3.80 GHz appears to offset its core advantage entirely. The data shows that in Cinebench R23 multi-core, the E5-1660 v4 with 8 cores beats the D-1581 with 16 cores by 88 points (11285 vs 11197). This suggests that the E5-1660 v4’s higher per-core performance, driven by its superior clock speeds, compensates for having half the cores. The average benchmark scores reinforce this: the E5-1660 v4 averages 3264, while the D-1581 averages 3238, a difference of 26 points or 0.8%.

Neither processor approaches the top of the benchmark charts. At the 53rd percentile, both sit in the middle of the CPU performance distribution. The E5-1660 v4’s nearest rival, the Intel Xeon E5-1680 v3, scores 3265, effectively identical. The D-1581’s nearest rival, the AMD Ryzen 5 1600X, scores 3249, also within 0.3%. These benchmark results indicate that neither chip is a performance outlier; they are comparable to mid-range desktop and embedded parts from the same era.

FAQ

Q: Which processor has the higher single-core performance in Cinebench R23?

A: The Intel Xeon E5-1660 v4 scores 1593 in Cinebench R23 single-core, while the Intel Xeon D-1581 scores 1580. This gives the E5-1660 v4 a 0.8% advantage.

Q: Does the D-1581’s 16-core count give it a multi-core advantage?

A: No. In every multi-core test (Cinebench R15, R20, R23), the E5-1660 v4 wins by 0.8% despite having only 8 cores. The D-1581’s lower boost clock of 2.40 GHz versus 3.80 GHz negates its core advantage.

Q: What is the TDP difference between these two CPUs?

A: The E5-1660 v4 has a TDP of 140 W, while the D-1581 has a TDP of 65 W. The D-1581 consumes less than half the power budget.

Q: Do both processors support ECC memory?

A: Yes, both support ECC memory. However, the E5-1660 v4 supports only DDR4 over a quad-channel bus, while the D-1581 supports both DDR3 and DDR4 over a dual-channel bus.

Q: What is the L3 cache configuration on each chip?

A: The E5-1660 v4 has 20 MB of shared L3 cache. The D-1581 has 1.5 MB per core, which totals 24 MB across its 16 cores, but this is not listed as shared.

Q: How do their overall benchmark averages compare?

A: The E5-1660 v4 has an average benchmark score of 3264, and the D-1581 has 3238. Both are at the 53rd percentile of all CPUs.

Where Each One Wins

The E5-1660 v4 wins in every benchmark category measured here, but its advantages are narrow. It leads in single-core performance across all three Cinebench generations: R15 (160 vs 159), R20 (669 vs 663), and R23 (1593 vs 1580). This makes it the better choice for lightly threaded applications where per-core clock speed is the bottleneck. Its quad-channel memory bus with 76.8 GB/s bandwidth also gives it a theoretical memory throughput advantage over the D-1581’s dual-channel setup, though no memory benchmark is provided in the data. The E5-1660 v4’s 40 PCIe Gen 3 lanes versus the D-1581’s 24 lanes make it more suitable for systems with multiple add-in cards or high-bandwidth peripherals.

The D-1581 wins in power efficiency and core density. At 65 W TDP, it uses less than half the power of the E5-1660 v4’s 140 W, while still delivering 16 cores and 32 threads. This makes it the logical choice for passive-cooled or low-power embedded systems, network appliances, or storage servers where thermal and power constraints are critical. Its support for both DDR3 and DDR4 memory provides flexibility in system design, and its BGA 1667 socket (soldered) is typical for compact, board-level deployments. The D-1581’s 24 PCIe Gen 3 lanes are fewer than the E5-1660 v4’s 40, but they are sufficient for many embedded workloads. The data shows the D-1581’s multi-core scores are within 0.8% of the E5-1660 v4, meaning it sacrifices almost nothing in throughput while offering a far lower power ceiling.

In practical terms, the E5-1660 v4 is the pick for a socketed workstation or server that prioritizes single-threaded performance and memory bandwidth. The D-1581 is the pick for a fixed, low-power appliance where the 16-core count supports concurrent workloads but the system cannot accommodate the E5-1660 v4’s power draw or socket type.

Specification Differences

| Specification | Intel Xeon E5-1660 v4 | Intel Xeon D-1581 |

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

| Cores | 8 | 16 |

| Threads | 16 | 32 |

| Base Clock | 3.20 GHz | 1800.00 MHz |

| Boost Clock | 3.80 GHz | 2.40 GHz |

| TDP | 140 W | 65 W |

| Socket | Intel Socket 2011-3 | Intel BGA 1667 |

| Codename | Broadwell-EP | Broadwell |

| Generation | Xeon E5 (Broadwell-EP) | Xeon D (Broadwell-DE) |

| Transistors | 3,400 million | 3,200 million |

| L3 Cache | 20 MB (shared) | 1.5 MB (per core) |

| Memory Support | DDR4 | DDR3, DDR4 |

| Memory Bus | Quad-channel | Dual-channel |

| Memory Bandwidth | 76.8 GB/s | Not listed |

| PCIe | Gen 3, 40 Lanes (CPU only) | Gen 3, 24 Lanes (CPU only) |

| Market Segment | Desktop | Server/Workstation |

| Production Status | End-of-life | Active |

| Release Date | 2016-06-19 | 2016-02-23 |

| Launch MSRP | $1113 | Not listed |

| Part Number | SR2PK | SR2LZ |