AMD A8-6600K vs Intel Pentium Silver J5005 Comparison

Where Each One Wins

The recorded data splits these two processors along clear lines of intended use. The Intel Pentium Silver J5005, a 14 nm Goldmont Plus part from the Gemini Lake family, dominates the modern rendering workloads. In Cinebench R15 multi-core, it posts 268 points, while the AMD A8-6600K has no recorded result for that test. The same pattern appears in Cinebench R20 and R23, where the J5005 scores 1117 and 2660 respectively, again with no comparable AMD entries. Single-threaded Cinebench results are similarly one-sided: the J5005 reaches 78.3 in R15, 157 in R20, and 375 in R23, all without an AMD counterpart in the database.

The AMD A8-6600K, a 32 nm Piledriver part from the Richland line, wins exclusively in the Geekbench tests that were recorded for it. It scores 1076 in Geekbench multi-core and 451 in single-core. The J5005 has no Geekbench entries in the database, so the AMD part holds those two victories uncontested. This asymmetry is not a fluke of benchmark selection; it reflects the different eras and design goals of the two chips. The J5005 was released in December 2017 as a mobile-oriented, low-power part. The A8-6600K arrived in June 2013 as a desktop chip with an unlocked multiplier and a 100 W thermal envelope.

The use-case split is therefore straightforward. For anyone running modern Cinebench-based workloads, the J5005 is the only option with recorded data, and its scores are respectable for a 10 W part. For legacy Geekbench testing, the A8-6600K holds the field. The database shows zero head-to-head benchmark entries, meaning no test was run on both chips. That absence forces the analysis to rely on the separate benchmark suites, but the direction of the wins is unambiguous: Cinebench belongs to Intel, Geekbench belongs to AMD.

Architecture Differences

The two processors share a core count of four and a thread count of four, but nearly every other architectural detail diverges. The J5005 uses Intel's Goldmont Plus microarchitecture on a 14 nm process, fabricated by Intel itself. Its die size is 93 mm². The A8-6600K uses AMD's Piledriver architecture on a 32 nm process from GlobalFoundries, with a much larger die at 246 mm² and 1,303 million transistors. The process node difference alone explains a substantial portion of the efficiency gap: 14 nm versus 32 nm is a two-generation jump in lithography.

Clock speeds tell the opposite story. The J5005 runs at a base of 1500 MHz and boosts to 2.80 GHz. The A8-6600K starts at 3.90 GHz and boosts to 4.20 GHz. That is more than double the base clock and a 50% higher boost clock for the AMD part. Yet the J5005 compensates with its architecture and process efficiency, as the Cinebench scores demonstrate. The thermal design power difference is stark: 10 W for the J5005 versus 100 W for the A8-6600K. That is a 90 W gap, which the database records without comment, but the implication for sustained workloads and cooling requirements is obvious.

Cache layouts also differ. The J5005 has 56 KB of L1 per core and 4 MB of shared L2. The A8-6600K has 192 KB of L1 total and 4 MB of L2. The Intel part's L1 is described per core, while the AMD part's is a single aggregate figure, so direct comparison is imprecise, but the total L2 capacity matches at 4 MB. Neither chip has L3 cache. Memory support separates them further: the J5005 uses DDR4 with dual-channel bandwidth of 38.4 GB/s, while the A8-6600K uses DDR3 with 29.9 GB/s. The Intel part also carries PCIe Gen 2 with 6 lanes (CPU only), whereas the AMD part lists PCIe Gen 2 without a lane count. Integrated graphics differ as well: UHD Graphics 605 on the Intel side versus Radeon HD 8570D on the AMD side. The J5005 is socketed as Intel BGA 1090, meaning it is soldered, while the A8-6600K uses AMD Socket FM2, a replaceable desktop socket. The AMD part has an unlocked multiplier; the Intel part does not.

Head-to-Head Benchmarks

The database contains no head-to-head benchmark entries for these two processors. That means no single test was run on both chips under identical conditions. The available data comes from separate benchmark suites: Cinebench R15, R20, and R23 for the J5005, and Geekbench multi-core and single-core for the A8-6600K. This lack of overlap prevents a direct score-to-score comparison on the same workload. However, the average benchmark scores provide a cross-suite summary. The J5005 has an average score of 776, and the A8-6600K sits at 764. That 12-point gap is small, roughly 1.6% in favor of Intel, but it masks the divergent workload characteristics.

Looking at the nearest rivals clarifies the positioning. The J5005's closest competitor in the database is the AMD Athlon X4 760K with an average score of 777, a delta of -0.1% from the J5005. The Intel Core i3-3250T and Intel Celeron G5920 both score 775, each 0.2% behind. The Intel Pentium G4520 scores 778, 0.2% ahead. These rivals are all within a narrow band, suggesting the J5005 sits in a crowded performance tier despite its architectural differences. The A8-6600K's nearest rivals include the Intel Core i7-3537U at 766 (0.3% behind), the Intel Core i3-3240 at 761 (0.5% ahead of the AMD part), the Intel Core i5-3210M at 758 (0.8% ahead), and the Intel Core i3-5020U at 757 (0.9% ahead). The AMD part is the weakest of that group, trailing all four rivals by small margins.

The biggest wins each way come from the benchmark suites themselves. The J5005's Cinebench R23 multi-core score of 2660 is its strongest recorded result, while its R23 single-core score of 375 shows a 7.1x ratio between multi and single-threaded performance in that test. The A8-6600K's Geekbench multi-core score of 1076 versus its single-core score of 451 yields a ratio of 2.4x. These ratios indicate that the Intel part scales much better across cores in Cinebench, likely due to its shared L2 and efficient Goldmont Plus cores, while the AMD part shows more modest scaling in Geekbench.

The Verdict

The data does not declare a single winner; it declares a context-dependent choice. For modern rendering workloads as captured by Cinebench, the Intel Pentium Silver J5005 is the only chip with recorded scores, and those scores are internally consistent and respectable. Its 2660 in R23 multi-core places it in the 21st percentile of all CPUs in the database, which is low overall but reasonable for a 10 W mobile part. The A8-6600K has no Cinebench data, so any claim about its rendering performance would be unsupported. For Geekbench, the AMD A8-6600K holds the only scores, with 1076 multi-core and 451 single-core, placing it in the 20th percentile. The J5005 has no Geekbench entries.

The average benchmark scores nearly tie, with the J5005 at 776 and the A8-6600K at 764, a gap of 1.6%. That near-tie, combined with the lack of overlapping tests, suggests the two chips are broadly comparable in overall performance but achieve it through very different means. The J5005 does more with far less power: 10 W versus 100 W is a tenfold difference in thermal design power. The A8-6600K compensates with much higher clocks and an unlocked multiplier for overclocking, but it draws ten times the power and uses an older, larger process node.

Who should pick which? A user constrained by power limits, small form factors, or soldered mobile platforms should choose the J5005, as its data shows it can deliver competitive Cinebench scores at a fraction of the power draw. A user with a desktop FM2 motherboard who prioritizes Geekbench performance and overclocking flexibility should choose the A8-6600K, since its unlocked multiplier and higher clocks are the only recorded advantages it holds. The J5005 is end-of-life, as is the A8-6600K, so neither offers a future upgrade path. The choice comes down to which benchmark suite matters more and which platform constraints apply.

FAQ

Q: Which processor has better Cinebench R23 multi-core performance?

A: The Intel Pentium Silver J5005 is the only chip with a recorded Cinebench R23 multi-core score, posting 2660. The AMD A8-6600K has no Cinebench R23 data in the database.

Q: How do the average benchmark scores compare?

A: The J5005 averages 776 across its recorded benchmarks, while the A8-6600K averages 764. The J5005 leads by 12 points, or about 1.6%.

Q: What is the power consumption difference?

A: The J5005 has a thermal design power of 10 W, while the A8-6600K is rated at 100 W. That is a tenfold difference in favor of the Intel part.

Q: Can the AMD A8-6600K be overclocked?

A: Yes. The A8-6600K has an unlocked multiplier, which the database records as true. The Intel J5005 does not have an unlocked multiplier.

Q: What memory types do the two processors support?

A: The J5005 supports DDR4 with dual-channel bandwidth of 38.4 GB/s. The A8-6600K supports DDR3 with dual-channel bandwidth of 29.9 GB/s.

Q: Which processor has a higher base clock?

A: The A8-6600K runs at a base clock of 3.90 GHz and boosts to 4.20 GHz. The J5005 runs at 1.50 GHz base and boosts to 2.80 GHz.

Specification Differences

| Specification | Intel Pentium Silver J5005 | AMD A8-6600K |

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

| Base clock | 1500 MHz | 3900 MHz |

| Boost clock | 2.80 GHz | 4.20 GHz |

| Thermal design power | 10 W | 100 W |

| Socket | Intel BGA 1090 | AMD Socket FM2 |

| Architecture | Goldmont Plus | Piledriver |

| Codename | Gemini Lake | Richland |

| Process node | 14 nm | 32 nm |

| Foundry | Intel | GlobalFoundries |

| Transistors | Not recorded | 1,303 million |

| Die size | 93 mm² | 246 mm² |

| L1 cache | 56 KB per core | 192 KB total |

| L2 cache | 4 MB shared | 4 MB |

| Memory support | DDR4 | DDR3 |

| Memory bandwidth | 38.4 GB/s | 29.9 GB/s |

| PCIe | Gen 2, 6 lanes (CPU only) | Gen 2 |

| Integrated graphics | UHD Graphics 605 | Radeon HD 8570D |

| Market segment | Mobile | Desktop |

| Release date | December 2017 | June 2013 |

| Multiplier unlocked | No | Yes |

| Launch MSRP | $161 | Not recorded |

| Part number | SR3S3 | AD660KWOHLBOXAD660KWOA44HL |