Intel Xeon E3-1275 v5 vs Intel Xeon E5-2628 v3 Comparison

The Intel Xeon E5-2628 v3 and the Intel Xeon E3-1275 v5 represent two different eras and market positions within Intel’s server and workstation lineup. The E5-2628 v3 is an 8-core, 16-thread Haswell-EP part designed for dual-socket platforms, while the E3-1275 v5 is a 4-core, 8-thread Skylake part aimed at single-socket workstations. The recorded benchmark data shows a consistent, albeit narrow, advantage for the older E5 processor across every Cinebench test, despite the E3’s higher clock speeds and newer architecture. This outcome highlights how core count and memory bandwidth can outweigh raw clock frequency in sustained workloads.

Head-to-Head Benchmarks

The head-to-head measurements cover six Cinebench tests, spanning R15, R20, and R23, each in both single-core and multi-core configurations. In every single test, the Intel Xeon E5-2628 v3 finishes ahead. The margin is remarkably consistent, ranging from 1.5% to 2.0%. The largest single win for the E5-2628 v3 comes in the Cinebench R15 single-core test, where it scores 102 against the E3-1275 v5’s 100, a 2% difference. That is a notable result because the E3-1275 v5 runs at a 3.60 GHz base and 4.00 GHz boost, while the E5-2628 v3 operates at just 2.50 GHz base and 3.00 GHz boost. The newer Skylake core with higher clocks still loses to the older Haswell part in single-threaded performance, which suggests the E5’s memory subsystem or other platform advantages play a role.

In multi-core workloads, the story is similar. The E5-2628 v3 posts 723 in Cinebench R15 multi-core versus 712 for the E3-1275 v5, a 1.5% lead. In Cinebench R20 multi-core, the scores are 3015 and 2967, respectively, a 1.6% gap. Cinebench R23 multi-core shows 7179 for the E5 and 7065 for the E3, again a 1.6% difference. The E5’s advantage in multi-core is easier to explain: it has twice as many cores and threads. However, the E3’s four cores clocked up to 4.00 GHz nearly close that gap, which shows the efficiency of the Skylake design. Still, the data is unambiguous: the E5-2628 v3 wins all six tests, with the E3-1275 v5 never taking a single benchmark.

The average benchmark scores from the database reinforce this pattern. The E5-2628 v3 has an average score of 2076, while the E3-1275 v5 sits at 2043. That is a difference of about 1.6%, which aligns with the head-to-head deltas. In terms of overall percentile ranking, the E5-2628 v3 lands at the 46th percentile among all tested CPUs, while the E3-1275 v5 is at the 45th. These are very close positions, meaning both processors sit in the lower-middle range of the performance distribution. The nearest rivals for the E5-2628 v3 include the Intel Xeon W-2123 (average score 2079, a 0.1% gap), the Intel Core i3-9350KF (2082, 0.3% gap), and the Intel Xeon E3-1585 v5 (2067, 0.5% gap). For the E3-1275 v5, the closest competitors are the AMD Ryzen 5 7520U (2046, 0.1% gap), the Intel Core i7-7700T (2039, 0.2% gap), and the Intel Xeon E3-1270 v5 (2048, 0.2% gap). These rival comparisons show that both processors are tightly clustered in performance, and the differences between them are within the range of normal run-to-run variation.

FAQ

Q: Which processor wins more benchmarks in the recorded data?

A: The Intel Xeon E5-2628 v3 wins all six head-to-head tests. It leads in Cinebench R15 multi-core (723 vs 712), R15 single-core (102 vs 100), R20 multi-core (3015 vs 2967), R20 single-core (425 vs 418), R23 multi-core (7179 vs 7065), and R23 single-core (1013 vs 997).

Q: How big is the performance gap between the two CPUs?

A: The largest margin is 2% in Cinebench R15 single-core. All other tests show a 1.5% to 1.7% lead for the E5-2628 v3. The average benchmark scores are 2076 for the E5 and 2043 for the E3, a difference of about 1.6%.

Q: Does the E3-1275 v5’s higher clock speed help it win any tests?

A: No. Despite a 3.60 GHz base and 4.00 GHz boost clock, the E3-1275 v5 loses every single-core test. The E5-2628 v3, running at 2.50 GHz base and 3.00 GHz boost, scores higher in both Cinebench R15 single-core (102 vs 100) and R23 single-core (1013 vs 997).

Q: What are the core and thread counts for each processor?

A: The E5-2628 v3 has 8 cores and 16 threads. The E3-1275 v5 has 4 cores and 8 threads. The E5’s advantage in core count helps it in multi-core tests, though the E3’s higher clocks narrow the gap.

Q: How do these processors compare to their nearest rivals in the database?

A: The E5-2628 v3 is within 0.5% of the Intel Xeon W-2123, Intel Core i3-9350KF, and Intel Xeon E3-1585 v5. The E3-1275 v5 is within 0.3% of the AMD Ryzen 5 7520U, Intel Core i7-7700T, and Intel Xeon E3-1270 v5. Both CPUs are in a dense performance cluster.

Q: What memory bandwidth do these platforms support?

A: The E5-2628 v3 supports quad-channel DDR4 memory with a bandwidth of 68.3 GB/s. The E3-1275 v5 supports dual-channel DDR3 or DDR4 memory with a bandwidth of 34.1 GB/s. The E5’s doubled memory channels likely contribute to its multi-core performance.

Architecture Differences

The two processors come from different architectural generations and serve different platform roles. The Intel Xeon E5-2628 v3 is built on the Haswell-EP architecture, using a 22 nm process node. It is fabricated with 2,600 million transistors on a 356 mm² die. The E5-2628 v3 uses the Intel Socket 2011-3, which supports quad-channel memory and 40 PCIe Gen 3 lanes from the CPU. This is a high-end server platform designed for dual-socket configurations, though the specific model here operates as a standalone part in the database.

The Intel Xeon E3-1275 v5 is based on the Skylake-DT architecture, using a 14 nm process node. It contains 1,750 million transistors on a 122 mm² die, which is significantly smaller and more power-efficient per transistor. The E3-1275 v5 uses the Intel Socket 1151, which only supports dual-channel memory and 16 PCIe Gen 3 lanes. This platform is aimed at single-socket workstations and entry-level servers. The E3 also includes integrated graphics in the form of HD Graphics P530, a feature the E5-2628 v3 lacks entirely.

Cache configurations differ as well. Both processors have 64 KB of L1 cache per core and 256 KB of L2 cache per core. The shared L3 cache, however, is much larger on the E5-2628 v3: 20 MB shared, compared to 8 MB shared on the E3-1275 v5. This larger L3 cache helps the E5 feed its 8 cores, especially in multi-threaded workloads. The E3’s smaller 8 MB cache is still adequate for 4 cores, but it cannot match the E5’s capacity.

Memory support also marks a clear architectural split. The E5-2628 v3 supports DDR4 memory in a quad-channel configuration, yielding a theoretical bandwidth of 68.3 GB/s. The E3-1275 v5 supports both DDR3 and DDR4, but only in a dual-channel configuration, capping bandwidth at 34.1 GB/s. That is exactly half the E5’s bandwidth, which directly impacts multi-core performance in memory-sensitive applications.

Specification Differences

The core specifications show a clear trade-off between core count and clock speed. The E5-2628 v3 has 8 cores and 16 threads, with a base clock of 2.50 GHz and a boost clock of 3.00 GHz. The E3-1275 v5 has 4 cores and 8 threads, with a base clock of 3.60 GHz and a boost clock of 4.00 GHz. The E3 runs 44% higher at base and 33% higher at boost, yet it still loses every benchmark.

Thermal design power differs slightly: the E5-2628 v3 is rated at 85 watts, while the E3-1275 v5 is rated at 80 watts. Despite having twice the cores, the E5 only draws 5 watts more, which reflects the older 22 nm process versus the newer 14 nm process. The E5’s larger die area and higher transistor count do not translate into a proportionally higher TDP.

Sockets are completely different. The E5-2628 v3 uses Intel Socket 2011-3, which is a large server socket supporting quad-channel memory and 40 PCIe Gen 3 lanes. The E3-1275 v5 uses Intel Socket 1151, a mainstream desktop-style socket with dual-channel memory and 16 PCIe Gen 3 lanes. This means the two processors are not interchangeable, and any platform decision must account for the motherboard and memory constraints.

Memory support differs in both type and width. The E5-2628 v3 supports DDR4 only, in quad-channel mode. The E3-1275 v5 supports both DDR3 and DDR4, but only in dual-channel mode. The E5’s memory bandwidth of 68.3 GB/s is exactly double the E3’s 34.1 GB/s. Both processors support ECC memory, which is a requirement for server and workstation reliability.

The E3-1275 v5 includes integrated graphics (HD Graphics P530), while the E5-2628 v3 has none. This makes the E3 a better option for systems that need a display output without a discrete GPU, though the E5 is typically paired with a dedicated graphics card in a server or workstation context.

Release dates also differ. The E5-2628 v3 was released on September 7, 2014, while the E3-1275 v5 came out on October 18, 2015. Both are now end-of-life products. The E3-1275 v5 has a launch MSRP of $350, which can be stated once as a reference point, though the E5-2628 v3 has no recorded launch price.

The Verdict

Based strictly on the recorded benchmark data, the Intel Xeon E5-2628 v3 is the faster processor in every measured test. It wins all six Cinebench tests, with margins between 1.5% and 2.0%. Its average benchmark score of 2076 is higher than the E3-1275 v5’s 2043. The E5-2628 v3 also holds a higher percentile ranking at 46 versus 45. If the only concern is raw CPU performance, the E5-2628 v3 is the clear choice.

However, the decision is not just about raw speed. The E3-1275 v5 offers a lower TDP of 80 watts versus 85 watts, integrated graphics, and a much smaller die (122 mm² vs 356 mm²). It also supports both DDR3 and DDR4, which gives platform flexibility. The E3’s higher clock speeds (up to 4.00 GHz) do not translate into benchmark wins, but they may benefit older software that is not well-threaded, even though the recorded tests do not show such an advantage.

For a new system build, the E5-2628 v3 requires a Socket 2011-3 motherboard with quad-channel memory support, which is typically more expensive and geared toward dual-socket servers. The E3-1275 v5 uses Socket 1151, which is more common and often cheaper to implement. The E5’s 40 PCIe Gen 3 lanes versus 16 on the E3 also makes it better for heavy expansion card usage.

The verdict from the data is straightforward: pick the E5-2628 v3 if you need the extra cores, threads, and memory bandwidth, and if the platform cost is acceptable. Pick the E3-1275 v5 if you want a lower power draw, integrated graphics, and a simpler, cheaper motherboard, and if you can accept a small performance deficit in every benchmark.

Where Each One Wins

The E5-2628 v3 wins in all multi-core scenarios. Its 8 cores and 16 threads, combined with 20 MB of L3 cache and 68.3 GB/s of memory bandwidth, make it the superior choice for heavily threaded workloads like video rendering, 3D modeling, scientific simulations, and database processing. The Cinebench R23 multi-core score of 7179 versus 7065 shows a 1.6% lead, which grows with larger data sets that stress memory.

The E5-2628 v3 also wins in single-core tests, which is surprising given the E3’s clock advantage. The E5’s 102 versus 100 in Cinebench R15 single-core and 1013 versus 997 in R23 single-core suggest that the E5’s platform, including its quad-channel memory controller and larger cache, compensates for lower clocks. This makes the E5 a better choice for mixed workloads that combine single-threaded and multi-threaded tasks.

The E3-1275 v5 does not win any recorded benchmark, but it has advantages outside of raw performance. Its 80 watt TDP makes it easier to cool in smaller chassis. Its integrated HD Graphics P530 means it can run a display without a discrete GPU, which is useful for basic server tasks or workstations with low graphical demands. Its support for DDR3 memory allows builders to reuse older RAM, reducing platform cost. Its dual-channel memory bandwidth of 34.1 GB/s is half the E5’s, but for many light server workloads, that is sufficient.

The E3-1275 v5 also has a smaller physical footprint and less complex power delivery requirements, which makes it a better fit for compact workstation builds. Its 16 PCIe Gen 3 lanes are enough for a single high-end GPU or a couple of NVMe drives. The E5’s 40 lanes are overkill unless you need multiple GPUs or many expansion cards.

In terms of production status, both are end-of-life, so neither has a future upgrade path. The E5-2628 v3’s Socket 2011-3 platform supports other Xeon E5 v3 and v4 parts, which could offer more cores for an upgrade. The E3-1275 v5’s Socket 1151 platform is limited to Skylake and Kaby Lake Xeon parts, which are also end-of-life.

The practical takeaway is that the E5-2628 v3 is the better CPU for performance, but the E3-1275 v5 is the better platform for simplicity and low-power operation. The recorded data cannot tell you which is more important for your specific use case, but it does show that the E5-2628 v3 is consistently faster in every benchmark. If you prioritize benchmark scores, choose the E5. If you prioritize platform flexibility, integrated graphics, and lower power, the E3 is a reasonable alternative despite losing all head-to-head tests.