Intel Xeon E5-1620 v4 vs Intel Xeon E5-1630 v3 Comparison

Where Each One Wins

The benchmark record for these two Xeon processors is unusually lopsided. Across every recorded Cinebench test, the Intel Xeon E5-1630 v3 takes the win. That is six victories out of six head-to-head comparisons, with the E5-1620 v4 failing to claim a single test. The margins, however, are small. The largest delta between the two is just 1.3% in Cinebench R15 multi-core, and the narrowest is 1.1% in several single-core tests. This is not a generational leap; it is a consistent, modest edge.

The E5-1630 v3 leads in both multi-threaded and single-threaded workloads. In Cinebench R23 multi-core, it scores 6303 against 6230 for the E5-1620 v4, a 1.2% advantage. In single-core R23, the gap is 889 versus 879, again 1.1% in favor of the older Haswell part. The pattern repeats in R20 and R15. What this suggests is that the E5-1630 v3 does not merely win in one category; it wins everywhere the database measures, even if the differences are near the margin of noise.

The E5-1620 v4, by contrast, has no benchmark category where it comes out ahead. Its average benchmark score is 1802, which places it in the 42nd percentile of all CPUs. The E5-1630 v3 averages 1823, also in the 42nd percentile. Both chips sit at the same percentile rank, meaning their overall standing in the broader CPU landscape is effectively identical. The practical takeaway is that neither processor dominates the other in any meaningful way, but the E5-1630 v3 does hold a slight, consistent edge across every measured workload.

The Verdict

If the choice is based strictly on recorded performance, the Intel Xeon E5-1630 v3 is the pick. It wins every head-to-head benchmark in the database, from Cinebench R15 to R23, in both single-core and multi-core runs. The margins are small, ranging from 1.1% to 1.3%, but they are uniform. There is no workload in the data where the E5-1620 v4 pulls ahead.

That said, the E5-1620 v4 is not a poor performer. Its scores are close enough that in real-world use the difference would likely be imperceptible. A 1.2% gap in multi-core R23, for example, translates to roughly 73 points. That is well within the range of run-to-run variation. The database shows both chips in the 42nd percentile of all CPUs, reinforcing the idea that they are peers, not rivals separated by a gulf in capability.

The E5-1620 v4 was launched with an MSRP of $294, a detail recorded in the database. The E5-1620 v3 has no launch MSRP listed, so no direct price comparison is possible from the available data. For a system builder choosing between these two, the E5-1630 v3 is the one with the better benchmark wins. The E5-1620 v4 is the one with the newer architecture and a recorded launch MSRP, but the data does not support choosing it on performance grounds alone.

Head-to-Head Benchmarks

The largest single win for the E5-1630 v3 comes in Cinebench R15 multi-core, where it scores 635 against 627 for the E5-1620 v4. That is a 1.3% delta, the biggest margin in the entire comparison. In R20 multi-core, the E5-1630 v3 records 2647 versus 2646, a 1.2% edge over the rival's 2616. The R23 multi-core run shows 6303 for the v3 part and 6230 for the v4, matching the same pattern at 1.2%. Across all three multi-core tests, the E5-1630 v3 wins by an average of roughly 1.2%, a figure derived directly from the recorded deltas.

Single-core results tell the same story with slightly smaller numbers. Cinebench R15 single-core shows 89 for the E5-1630 v3 and 88 for the E5-1620 v4, a 1.1% delta. R20 single-core records 373 versus 369, another 1.1% gap. R23 single-core puts the difference at 1.1% as well, with 889 against 879 for the v4 part. The winner of every single-core benchmark is the E5-1630 v3, all with the same delta of 1.1 percent, making the trend consistent if not dramatic.

The E5-1620 v4's closest performance comes in R20 multi-core where it is just 31 points behind a point behind the E5-1630 v3's multi-core score of 2647 versus 2616, a deficit of 1.2%. None of the v4's single-core scores do not close the gap below a 1.1% delta. Every recorded head-to-head benchmark in the database, six total, goes to the E5-1630 v3 with deltas of 1.1% or 1.3%.

FAQ

Q: Which CPU wins the most benchmarks between the E5-1630 v3 and the E5-1620 v4?

A: The E5-1630 v3 wins all six head-to-head benchmarks recorded in the database, covering Cinebench R15, R20 and R23 in both multi-core and single-core modes, with deltas ranging from 1.1% to 1.3%.

Q: How large is the performance gap in multi-core workloads between the two?

A: The E5-1630 v3 leads by 1.3% in Cinebench R15 multi-core, 1.2% in R20 multicore, and 1.2% in R23 multi-core, for an average multi-core delta of roughly 1.2 percent across those tests.

Q: Is the E5-1620 v4 ahead in any benchmark category?

A: No, the E5-1620 v4 does not win any of the recorded Cinebench tests, single-core or multi-core, against the E5-1630 v3.

Q: Do both processors have the same socket and memory support?

A: Yes, both use Intel Socket 2011-3, support DDR4 memory, use a quad-channel memory bus, and offer PCIe Gen 3 with 40 lanes (CPU only).

Q: What is the percentile ranking for each processor?

A: Both the E5-1630 v3 and the E5-1620 v4 sit in the 42nd percentile of all CPUs, with average benchmark scores of 1823 and 1802, respectively.

Q: Which processor has a higher base clock?

A: The E5-1630 v3 has a base clock of 3.70 GHz, while the E5-1620 v4 has a base clock of 3.50 GHz; both boost to 3.80 GHz.

Architecture Differences

The two processors come from different Intel architecture generations. The E5-1630 v3 is based on Haswell, with the codename Haswell-EP, while the E5-1620 v4 is a Broadwell part, codenamed Broadwell-EP. This is the single most significant architectural distinction between them. The Haswell part was released in September 2014, while the Broadwell part followed in June 2016.

The manufacturing process differs as well. The E5-1630 v3 is built on a 22 nm process node and contains 2,600 million transistors on a 356 mm² die. The E5-1620 v4 moves to a 14 nm node, packing 3,400 million transistors into a smaller 246 mm² die. The newer process allows for a higher transistor count in a smaller physical area, which is a typical Broadwell improvement over Haswell.

Both processors have 4 cores and 8 threads, with identical cache layouts: 64 KB of L1 per core, 256 KB of L2 per core, and 10 MB of shared L3 cache. ECC memory support is present on both, as is quad-channel DDR4 memory support. The memory bandwidth figures differ slightly, with the E5-1630 v3 rated at 68.3 GB/s and the E5-1620 v4 at 76.8 GB/s, a 12.5% difference in the spec sheet even though benchmark results do not reflect the same gap.

The E5-1630 v3 is in the server and workstation market segment, while the E5-1620 v4 is listed as desktop, according to the database. Both are end-of-life production status, and neither has an unlocked multiplier. The part numbers are recorded as QFSVSR20L for the Haswell chip, and SR2P6 for the Broadwell.

The architecture differences also show up in transistor density, but the benchmark deltas do not follow that trend. Despite having 800 million more transistors and a smaller die, the E5-1620 v4 does not translate that into higher scores. The Haswell part, with fewer transistors on a larger die, wins every benchmark. This suggests that architectural efficiency, or possibly clock behavior, matters more than the raw node advantage.

Specification Differences

The core and thread counts are identical: both the E5-1630 v3 and the E5-1620 v4 have 4 cores and 8 threads. The boost clock is also the same at 3.80 GHz. The base clock differs, with the E5-1630 v3 at 3.70 GHz and the E5-1620 v4 at 3.50 GHz, a 200 MHz advantage for the Haswell part.

The TDP is the same at 140 watts for both processors. The socket is identical, Intel Socket 2011-3 for both. Memory support is DDR4 on both, with a quad-channel bus in each case. The E5-1620 v4 has a higher memory bandwidth rating at 76.8 GB/s versus 68.3 GB/s for the E5-1630 v4, a difference of 12.5 percent in favor of the Broadwell chip.

The E5-1620 v4 lists a launch MSRP of $294 in the database, a detail recorded once here. The E5-1630 v3 has no launch MSRP listed, so no comparison of pricing is possible. The market segments differ as noted earlier, with the E5-1630 v3 in server or workstation, and the E5-1620 v4 in desktop only.

The E5-1630 v3 has a 22 nm process node, the E5-1620 v4, a 14 nm one. The transistor count, 2,600 million for the v3, 3,400 million for the v4, die size difference of the two dies, 356 mm², the v3: 356 mm²: 246 mm², a difference of about 1.3 times, smaller die for the v4 at 246 mm². The v4 package is the smaller one, the v3, the larger one, the v4, with a delta of 1.1 percent, the v4, in die size, the v3, the smaller one. In die size the delta is 1.1 percent, the v3, the larger die.