CPU Comparison

# AMD Opteron 6328 vs AMD Opteron 6272

The AMD Opteron 6328 and AMD Opteron 6272 are both 32 nm server processors from the Opteron 6000 series, sharing the same G34 socket and 115 W TDP, but they represent two distinct generations of AMD's modular architecture. The 6328, built on the Piledriver design (codename Abu Dhabi), ships with 8 cores and 8 threads, clocked at 3.20 GHz base and 3.80 GHz boost. The 6272, built on the older Bulldozer design (codename Interlagos), doubles the core count to 16 cores and 16 threads but runs at a much lower 2.10 GHz base and 3.00 GHz boost. Despite having half the cores, the 6328 wins every single benchmark in the head-to-head comparison, albeit by narrow margins of 1.3% to 1.6%. The data reveals a fascinating trade-off: the newer architecture's higher clocks and improved IPC fully compensate for the 6272's double core count, making the 6328 the better performer across the board.

The Verdict

The benchmark data is unambiguous: the AMD Opteron 6328 wins all six head-to-head tests against the AMD Opteron 6272. In Cinebench R15 multicore, the 6328 scores 454 versus 448 for the 6272, a 1.3% advantage. That pattern holds across every test, R15 singlecore (64 vs 63, +1.6%), R20 multicore (1894 vs 1870, +1.3%), R20 singlecore (267 vs 263, +1.5%), R23 multicore (4510 vs 4454, +1.3%), and R23 singlecore (636 vs 628, +1.3%). The margins are modest but consistent, and the 6328 never loses a single benchmark.

Who should pick which? If the choice is purely about raw Cinebench scores, the 6328 is the clear winner, it achieves higher performance in every workload measured despite using fewer physical cores. The 6272's dual 8-core modules simply cannot overcome the 6328's 1.10 GHz base-clock advantage and newer Piledriver architecture. However, the 6272 does offer double the core count (16 vs 8) at a lower launch MSRP ($523 vs $575), which might matter in workloads that scale with core count rather than clock speed, though the benchmark data here does not show any such advantage for the 6272. The average benchmark scores reinforce this: the 6328 averages 1304 across all tests, while the 6272 averages 1288, placing both at the 35th percentile of all CPUs and making them effectively equivalent in overall standing.

For a server/workstation buyer deciding strictly from this data, the 6328 is the superior part. It delivers better performance in every measured metric, and its 115 W TDP matches the 6272 despite running at significantly higher clocks. The 6272's only advantages are its lower launch price and higher core count, but neither translates into a benchmark win. The data suggests that AMD's architectural improvements in Piledriver over Bulldozer were substantial enough to overcome a 2x core deficit.

Architecture Differences

The architectural gap between these two Opterons is the heart of the performance difference. The 6328 uses the Piledriver architecture, while the 6272 uses the original Bulldozer design. Both are built on GlobalFoundries' 32 nm process with 2,400 million transistors and a dual-die layout of 2x 315 mm², but the internal microarchitecture differs significantly.

Piledriver, introduced with the Abu Dhabi generation, was AMD's second iteration of its modular core design. The 6328 implements this as 8 cores in what the cache layout reveals: 384 KB of L1 cache and 2 MB of L2 per module, with 8 MB of L3 per die. The 6272, using Bulldozer, has 768 KB of L1 cache, exactly double, because its 16 cores require more L1 capacity. Both share the same L2 (2 MB per module) and L3 (8 MB per die) configurations, and both have the same total transistor count and die size, indicating that the core count difference is managed through the module design rather than silicon area.

The clock speed gap is the most striking architectural consequence. The 6328 operates at 3.20 GHz base and 3.80 GHz boost, while the 6272 limps along at 2.10 GHz base and 3.00 GHz boost. This 1.10 GHz base-clock difference is enormous, over 52% higher base clock for the 6328. Piledriver also brought IPC improvements over Bulldozer, which explains why the 6328 wins even the single-core tests (where core count is irrelevant) by 1.3% to 1.6%.

Memory support also differs: both use quad-channel DDR3, but the 6328 delivers 59.7 GB/s of memory bandwidth versus 51.2 GB/s for the 6272, an 8.5 GB/s gap that favors the newer part. Both support ECC memory, both use PCIe Gen 2, and neither has integrated graphics. The 6328 was released on 2012-11-04, while the 6272 came earlier on 2011-11-13, meaning the 6328 benefits from roughly a year of additional architectural refinement.

FAQ

Q: Which processor has more cores, and does it matter?

A: The 6272 has 16 cores and 16 threads, double the 6328's 8 cores and 8 threads. However, the benchmark data shows the 6328 wins every test despite the core deficit, indicating that the 6272's extra cores do not translate into better Cinebench performance.

Q: What is the clock speed difference between the two?

A: The 6328 runs at 3.20 GHz base and 3.80 GHz boost, while the 6272 runs at 2.10 GHz base and 3.00 GHz boost. That's a 1.10 GHz base-clock advantage for the 6328, which is the primary driver of its benchmark wins.

Q: Do both processors use the same socket and memory?

A: Yes, both use AMD Socket G34 and support quad-channel DDR3 memory. However, the 6328 achieves 59.7 GB/s memory bandwidth versus 51.2 GB/s for the 6272, a difference that favors the newer part.

Q: Which processor has better single-core performance?

A: The 6328 wins all three single-core tests (R15, R20, R23) with scores of 64, 267, and 636 respectively, versus 63, 263, and 628 for the 6272. The margins range from 1.3% to 1.6%, all in favor of the 6328.

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

A: The 6328's average benchmark score of 1304 places it within 0.7% of the Intel Core i5-6400 (1307, -0.2% delta), Core i7-3610QE (1297, +0.5%), Athlon Gold 7220U (1311, -0.6%), and Core i7-3770T (1314, -0.7%). The 6272's 1288 average is similarly close to the Core i7-3630QM (1289, -0.1%), Core i5-4670R (1286, +0.1%), Core i5-3570 (1286, +0.2%), and Core i5-3570K (1293, -0.4%).

Q: What is the release date and production status of each?

A: The 6328 was released on 2012-11-04, and the 6272 was released on 2011-11-13. Both are marked as end-of-life production status.

Specification Differences

| Specification | AMD Opteron 6328 | AMD Opteron 6272 |

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

| Architecture | Piledriver | Bulldozer |

| Codename | Abu Dhabi | Interlagos |

| Cores | 8 | 16 |

| Threads | 8 | 16 |

| Base Clock | 3.20 GHz | 2.10 GHz |

| Boost Clock | 3.80 GHz | 3.00 GHz |

| L1 Cache | 384 KB | 768 KB |

| Memory Bandwidth | 59.7 GB/s | 51.2 GB/s |

| Release Date | 2012-11-04 | 2011-11-13 |

| Launch MSRP | $575 | $523 |

Both processors share the same 115 W TDP, AMD Socket G34, 32 nm process node, GlobalFoundries foundry, 2,400 million transistors, 2x 315 mm² die size, 2 MB L2 per module, 8 MB L3 per die, DDR3 memory support, quad-channel memory bus, ECC memory support, PCIe Gen 2, no integrated graphics, server/workstation market segment, and locked multiplier. The part numbers differ (OS6328WKT8GHK vs OS6272WKTGGGU), and the 6328's generation is listed as "Opteron (Abu Dhabi)" while the 6272's is "Opteron (Interlagos)".

Head-to-Head Benchmarks

The head-to-head results are a clean sweep for the 6328, with all six Cinebench tests going its way. The largest margin comes in the R15 singlecore test, where the 6328's 64 points beat the 6272's 63 by 1.6%. That's the biggest win in percentage terms, telling us the 6328's per-core advantage is most pronounced in lightly-threaded workloads.

In the multicore tests, the margins are slightly tighter but still consistent. R15 multicore shows 454 vs 448 (+1.3%), R20 multicore shows 1894 vs 1870 (+1.3%), and R23 multicore shows 4510 vs 4454 (+1.3%). The fact that the 6328 wins these multicore tests despite having half the cores is remarkable, it means each of its 8 cores is delivering roughly double the performance per core of the 6272's 16 cores. In R23 multicore, for example, the 6328 averages 563.75 points per core (4510/8), while the 6272 averages just 278.375 points per core (4454/16).

Single-core tests also favor the 6328: R20 singlecore shows 267 vs 263 (+1.5%), and R23 singlecore shows 636 vs 628 (+1.3%). These margins, while small in absolute terms, are consistent across all three Cinebench generations, suggesting a stable architectural advantage rather than a workload-specific quirk.

The average benchmark scores reinforce the picture: the 6328 averages 1304 across all tests, versus 1288 for the 6272, a 16-point gap that places the 6328 slightly ahead of the 6272's nearest rivals like the Core i7-3630QM (1289) but slightly behind the 6328's own nearest rival, the Athlon Gold 7220U (1311). Both processors sit at the 35th percentile of all CPUs, indicating they are mid-pack performers in the broader market.

Where Each One Wins

The 6328 wins everywhere in this benchmark suite, so the "where each one wins" analysis must focus on the magnitude and nature of its victories. The 6328's biggest wins are in single-core tests, where its clock speed and Piledriver IPC shine. The 1.6% win in R15 singlecore and 1.5% win in R20 singlecore show that the 6328's per-thread performance is its strongest asset. This makes it the better choice for workloads that are latency-sensitive or have limited parallelism, such as database transactions or lightly-threaded server applications.

The 6328's multicore wins, while still present, are slightly narrower at 1.3% across all three Cinebench multicore tests. This suggests that in heavily threaded workloads, the 6272's 16 cores do close the gap somewhat, but not enough to overtake the 6328. The 6328's higher memory bandwidth (59.7 GB/s vs 51.2 GB/s) likely contributes to its multicore advantage, as memory-bound workloads benefit from the extra throughput.

The 6272's only theoretical wins come from its specifications rather than benchmark results. It has double the cores (16 vs 8) and threads, which could matter in workloads that the Cinebench suite doesn't measure, for example, highly parallel server virtualization or scientific computing that scales perfectly with core count. The 6272 also had a lower launch MSRP ($523 vs $575), which might have made it attractive for cost-conscious buyers at launch, though the benchmark data shows no performance advantage to justify that trade-off.

The data ultimately tells a clear story: the 6328 is the better processor in every measured dimension. Its newer Piledriver architecture, higher clocks, and superior memory bandwidth overcome the 6272's core-count advantage. For anyone choosing between these two end-of-life server parts, the 6328 is the performance pick, and the 6272's only remaining argument is its higher core count and lower launch price, neither of which shows up as a win in the benchmark results.