Intel Xeon D-1567 vs Intel Xeon E5-2640 v3 Comparison

FAQ

Q: Which processor has more cores and threads?

A: The Intel Xeon D-1567 has 12 cores and 24 threads, while the Intel Xeon E5-2640 v3 has 8 cores and 16 threads.

Q: What is the difference in base and boost clock speeds?

A: The Intel Xeon E5-2640 v3 runs at a base clock of 2.60 GHz and boosts to 3.40 GHz. The Intel Xeon D-1567 runs at 2.10 GHz base and 2.70 GHz boost.

Q: Which CPU has a larger L3 cache?

A: The Intel Xeon E5-2640 v3 has 20 MB of shared L3 cache. The Intel Xeon D-1567 has 1.5 MB of L3 cache per core, with no total L3 value recorded in the database.

Q: What is the launch MSRP of each processor?

A: The Intel Xeon D-1567 has a launch MSRP of 1199, and the Intel Xeon E5-2640 v3 has a launch MSRP of $939.

Q: Which chip supports more memory channels?

A: The Intel Xeon E5-2640 v3 supports quad-channel memory, while the Intel Xeon D-1567 supports dual-channel memory.

Q: How do the two compare in average benchmark score?

A: The Intel Xeon D-1567 averages 2770 across all benchmarks, placing in the 51st percentile of all CPUs. The Intel Xeon E5-2640 v3 averages 2703, placing in the 50th percentile.

Architecture Differences

The two processors come from different Intel microarchitectures and are built for different physical contexts. The Intel Xeon D-1567 is a Broadwell-DE part on a 14 nm process, fabricated by Intel with 3,200 million transistors on a 246 mm² die. It uses the Intel BGA 1667 socket, which means it is a soldered, embedded-style processor rather than a socketed upgrade part. The Intel Xeon E5-2640 v3 is a Haswell-EP part on a 22 nm process, with 2,600 million transistors on a larger 356 mm² die. It uses the Intel Socket 2011-3, a standard server/workstation socket with replaceable processors.

The memory controllers differ significantly. The Xeon D-1567 supports both DDR3 and DDR4 across a dual-channel memory bus. The Xeon E5-2640 v3 supports only DDR4, but across a quad-channel memory bus with a recorded memory bandwidth of 59.7 GB/s. The database does not record a memory bandwidth figure for the Xeon D-1567, so the E5-2640 v3 has the only quantified bandwidth advantage here.

PCIe connectivity also separates the two. The Xeon D-1567 provides Gen 3, 24 lanes (CPU only). The Xeon E5-2640 v3 provides Gen 3, 40 lanes (CPU only). That is a substantial difference in available expansion capability for storage controllers, network cards, or accelerators.

Cache organization is another architectural split. The Xeon D-1567 uses per-core L1 of 64 KB, per-core L2 of 256 KB, and per-core L3 of 1.5 MB. The Xeon E5-2640 v3 uses the same per-core L1 and L2 sizes, but its L3 is 20 MB shared across all cores. The shared pool on the E5-2640 v3 is a classic Haswell-EP design, while the per-core L3 on the D-1567 reflects the Broadwell-DE compact SoC approach.

Both processors support ECC memory, and neither includes integrated graphics. Both have locked multipliers. The production status differs: the Xeon D-1567 is marked Active, while the Xeon E5-2640 v3 is End-of-life. The release dates are also different, with the E5-2640 v3 launching in September 2014 and the D-1567 launching in February 2016.

The Verdict

The benchmark data shows a clear, if narrow, overall winner in the Intel Xeon D-1567. It wins all six recorded head-to-head Cinebench tests, with deltas between 2.4% and 3.0%. The average benchmark score favors the D-1567 at 2770 versus 2703 for the E5-2640 v3, and the percentile rank is 51 versus 50.

The Xeon D-1567 is the better choice if you want more cores and threads in a lower-TDP package. It offers 12 cores and 24 threads at 65 W TDP, compared to 8 cores and 16 threads at 90 W TDP for the E5-2640 v3. For dense, multicore workloads in a power-constrained chassis, the data points squarely to the D-1567.

The Xeon E5-2640 v3 still has arguments in its favor. It carries a higher base clock and boost clock, a larger shared L3 cache of 20 MB, quad-channel DDR4 memory with 59.7 GB/s bandwidth, and 40 PCIe Gen 3 lanes. If your workload depends on memory bandwidth or many expansion devices, the E5-2640 v3 is the more capable platform despite losing every Cinebench race.

The launch MSRP of the E5-2640 v3 is $939, while the D-1567 is 1199. The database records these as launch prices only, not current market pricing, so any purchase decision should be based on platform fit and benchmark behavior rather than price.

For a straightforward compute comparison, the D-1567 wins. For memory-bandwidth-heavy or expansion-heavy server roles, the E5-2640 v3 remains relevant. The recorded data does not support calling either chip a clear overall platform winner; it supports calling the D-1567 the faster CPU in these specific rendering benchmarks.

Specification Differences

The two CPUs differ in nearly every major specification field.

• Cores: 12 (D-1567) vs 8 (E5-2640 v3)
• Threads: 24 (D-1567) vs 16 (E5-2640 v3)
• Base clock: 2.10 GHz (D-1567) vs 2.60 GHz (E5-2640 v3)
• Boost clock: 2.70 GHz (D-1567) vs 3.40 GHz (E5-2640 v3)
• TDP: 65 W (D-1567) vs 90 W (E5-2640 v3)
• Socket: Intel BGA 1667 (D-1567) vs Intel Socket 2011-3 (E5-2640 v3)
• Architecture: Broadwell (D-1567) vs Haswell (E5-2640 v3)
• Process node: 14 nm (D-1567) vs 22 nm (E5-2640 v3)
• Transistors: 3,200 million (D-1567) vs 2,600 million (E5-2640 v3)
• Die size: 246 mm² (D-1567) vs 356 mm² (E5-2640 v3)
• L3 cache: 1.5 MB per core (D-1567) vs 20 MB shared (E5-2640 v3)
• Memory support: DDR3, DDR4 (D-1567) vs DDR4 only (E5-2640 v3)
• Memory bus: Dual-channel (D-1567) vs Quad-channel (E5-2640 v3)
• Memory bandwidth: not recorded (D-1567) vs 59.7 GB/s (E5-2640 v3)
• PCIe: Gen 3, 24 lanes (D-1567) vs Gen 3, 40 lanes (E5-2640 v3)
• Production status: Active (D-1567) vs End-of-life (E5-2640 v3)
• Release date: February 2016 (D-1567) vs September 2014 (E5-2640 v3)
• Part number: SR2M3 (D-1567) vs SR205 (E5-2640 v3)

Both support ECC memory, both lack integrated graphics, and both have locked multipliers. The L1 and L2 cache sizes are identical per core: 64 KB and 256 KB respectively.

Head-to-Head Benchmarks

The head-to-head results are consistent across all six Cinebench tests. The Intel Xeon D-1567 wins every single one, but by margins that are small rather than decisive.

In Cinebench R15 multicore, the D-1567 scores 965 against 942 for the E5-2640 v3, a 2.4% advantage. In Cinebench R15 singlecore, the D-1567 scores 136 against 132, a 3.0% edge. That single-core margin is the largest recorded in the entire comparison.

Cinebench R20 multicore gives the D-1567 a 4022 score versus 3926, again a 2.4% delta. Single-core R20 is 567 versus 553, a 2.5% margin. Cinebench R23 multicore shows 9578 for the D-1567 and 9348 for the E5-2640 v3, a 2.5% gap. Single-core R23 is 1352 versus 1319, also 2.5%.

The pattern is remarkably uniform. The D-1567 leads by roughly two and a half percent in almost every test, with the single exception of the R15 single-core test at 3.0%. The E5-2640 v3 never closes the gap to under 2.4% in any recorded test.

The average benchmark score reinforces the same story: 2770 for the D-1567 versus 2703 for the E5-2640 v3. The D-1567 sits at the 51st percentile of all CPUs, while the E5-2640 v3 sits at the 50th percentile. The nearest rivals in the database for the D-1567 include the AMD Ryzen 3 PRO 4350GE at 2767, the AMD Ryzen 3 PRO 4350G at 2774, the Intel Core i7-5930K at 2775, and the AMD Ryzen 3 5400U at 2765. The E5-2640 v3 sits alongside the Intel Core i7-1185G7E at 2703, the Intel Core i7-8700T at 2701, the Intel Core i5-1345UE at 2698, and the Intel Xeon E-2226G at 2709.

What these numbers mean in practice is that the D-1567 extracts its performance from 12 cores and 24 threads at a lower TDP and a slower clock speed. The E5-2640 v3 counters with higher clocks and a larger shared cache, but the extra four cores of the D-1567 outweigh those advantages in every Cinebench workload.

Where Each One Wins

The Intel Xeon D-1567 wins in raw CPU rendering performance. All six Cinebench tests go to it, with multicore deltas of 2.4% to 2.5% and a single-core delta of 3.0%. It also wins on core count, thread count, TDP efficiency, process node, and production status. For any workload that scales with cores, such as software compilation, virtualization density, or batch rendering, the recorded data favors the D-1567.

The D-1567 also wins on power efficiency. It delivers more multicore performance while consuming a 65 W TDP, which is 25 W lower than the E5-2640 v3's 90 W TDP. In a dense server environment where thermal headroom is scarce, that difference can matter more than the benchmark delta.

The Intel Xeon E5-2640 v3 wins on platform-level capabilities. It has a 20 MB shared L3 cache, which can be beneficial for workloads with shared data sets. It supports quad-channel DDR4 memory with a recorded 59.7 GB/s of bandwidth, a major advantage for memory-bound tasks like database analytics or high-performance computing simulations. The database does not list a memory bandwidth figure for the D-1567, so the E5-2640 v3 is the only one with a quantified bandwidth advantage.

The E5-2640 v3 also wins on PCIe expansion with 40 Gen 3 lanes versus 24 lanes on the D-1567. That makes it the better fit for systems with multiple GPUs, NVMe storage arrays, or high-speed network adapters. It also uses a socketed platform, Intel Socket 2011-3, which allows for CPU replacement, whereas the D-1567 is soldered to a BGA 1667 board.

The E5-2640 v3 wins on clock speed as well, with a 2.60 GHz base and 3.40 GHz boost compared to 2.10 GHz and 2.70 GHz on the D-1567. For lightly threaded workloads that do not scale with cores, the E5-2640 v3 has a theoretical frequency advantage, although the recorded single-core Cinebench results still go to the D-1567 by 2.5% to 3.0%.

In summary, the D-1567 is the compute winner, and the E5-2640 v3 is the platform winner. If the workload is core-scaled and power-sensitive, the D-1567 is the data-backed choice. If the workload needs memory bandwidth, cache capacity, or PCIe lanes, the E5-2640 v3 offers capabilities the D-1567 cannot match in the recorded specifications.