AMD FX-9800P vs Intel Core i3-5020U Comparison

Where Each One Wins

The benchmark split between these two mobile processors is unusually clear-cut, and it comes down to which benchmark family you trust. The AMD FX-9800P takes five of the seven head-to-head tests, all of them from the Cinebench suite. The Intel Core i3-5020U wins the remaining two, both from Geekbench. That is not a tie, but the margins tell a more nuanced story than the raw win count.

In Cinebench R15 multicore, the AMD scores 210 against the Intel's 188, a 10.5% advantage. The pattern repeats in Cinebench R20 multicore (876 vs 787, a 10.2% gap) and in Cinebench R23 multicore (2088 vs 1876, again 10.2%). The single-core Cinebench results follow the same direction: R20 single-core shows 123 vs 110 (10.6% for AMD), and R23 single-core shows 294 vs 264 (10.2% for AMD). So across every Cinebench iteration, the AMD FX-9800P leads by roughly 10% in both single-threaded and multi-threaded workloads.

The Intel part fights back decisively in Geekbench. In Geekbench multicore, the Core i3-5020U scores 1379 against the FX-9800P's 1127, a 22.4% lead. The single-core Geekbench result is even more lopsided: 696 vs 498, a 39.8% advantage for Intel. These are not small margins, they are dominant wins that flip the overall picture.

So where does each one win? If your workload is rendered in Cinebench, the AMD FX-9800P is consistently ahead by a tenth. If your workload is measured by Geekbench, the Intel Core i3-5020U is faster by a fifth to two-fifths. The database shows a processor that wins decisively in one benchmark ecosystem and loses decisively in another. Neither chip is universally superior, the choice depends entirely on which software you prioritize.

Architecture Differences

The two chips come from fundamentally different design philosophies. The Intel Core i3-5020U is built on the Broadwell architecture, specifically the Broadwell-U codename, fabricated on Intel's 14 nm process node. It packs 1,300 million transistors into an 82 mm² die. The AMD FX-9800P uses the Excavator architecture under the Bristol Ridge codename, built on GlobalFoundries' 28 nm process. That older node requires far more silicon: 3,100 million transistors spread across a 250 mm² die. The die size difference is stark, AMD uses roughly three times the area to house more than double the transistor count.

Core counts differ as expected. The Intel chip has 2 cores and 4 threads, relying on Hyper-Threading to reach four threads. The AMD chip has 4 physical cores and 4 threads, no SMT. Cache layouts are entirely different. Intel provides 64 KB L1 per core, 256 KB L2 per core, and a shared 3 MB L3 cache. AMD lists 320 KB L1 total and 1 MB L2 per module, with no L3 cache at all. That missing L3 is a significant architectural gap, though the AMD chip's higher base clock partially compensates.

Clock speeds favor AMD on paper. The FX-9800P runs at a 2.70 GHz base clock with a 3.60 GHz boost clock. The Core i3-5020U has a fixed 2.20 GHz base clock with no boost capability, a typical limitation for the Core i3 tier in that generation. The AMD chip also supports a higher memory bandwidth, 29.9 GB/s with DDR4, while Intel is limited to 25.6 GB/s with DDR3. Both use dual-channel memory buses.

PCIe support differs too. The Intel chip provides Gen 2 with 12 lanes from the CPU, while AMD offers Gen 3 with 8 lanes. That means AMD has faster per-lane bandwidth but fewer total lanes. Integrated graphics also diverge: Intel uses HD 5500, AMD uses Radeon R7 with 8 compute units. Both are mobile parts with a 15 W TDP, both are end-of-life, and neither has an unlocked multiplier. The Intel chip launched with a $281 MSRP, while the AMD chip has no recorded launch price.

Head-to-Head Benchmarks

The head-to-head data reveals a consistent pattern with one major outlier. Let's walk through each test in order.

Cinebench R15 multicore: AMD FX-9800P wins 210 to 188, a 10.5% margin. This is the oldest Cinebench test in the set, and AMD's lead is already established.

Cinebench R20 multicore: AMD wins 876 to 787, a 10.2% margin. The gap is nearly identical to R15, suggesting the AMD advantage scales consistently across Cinebench versions.

Cinebench R20 single-core: AMD wins 123 to 110, a 10.6% margin. Even in a single-threaded test, the AMD chip's higher clock speed (2.70 GHz base, 3.60 GHz boost) overcomes Intel's architectural efficiency.

Cinebench R23 multicore: AMD wins 2088 to 1876, a 10.2% margin. The pattern holds exactly.

Cinebench R23 single-core: AMD wins 294 to 264, a 10.2% margin. Again, consistent.

Then Geekbench flips everything. Geekbench multicore: Intel wins 1379 to 1127, a 22.4% margin. This is more than double the AMD lead in any Cinebench test. Geekbench single-core: Intel wins 696 to 498, a 39.8% margin. That is the largest delta in the entire comparison, nearly four times the AMD advantage in any single test.

The interpretation is straightforward. The AMD FX-9800P beats the Intel Core i3-5020U by about 10% across all Cinebench workloads, both single and multi-threaded. The Intel chip beats AMD by 22% to 40% in Geekbench. The database shows two different benchmark philosophies producing opposite verdicts. Cinebench is a rendering workload that appears to favor AMD's higher clocks and four physical cores. Geekbench is a broader system test that appears to favor Intel's architecture, particularly in single-threaded scenarios where the 39.8% gap suggests a major IPC advantage for Broadwell.

Specification Differences

The recorded specifications show where these two processors diverge. The table below lists only the fields where the two differ.

| Specification | Intel Core i3-5020U | AMD FX-9800P |

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

| Cores | 2 | 4 |

| Threads | 4 | 4 |

| Base Clock | 2.20 GHz | 2.70 GHz |

| Boost Clock | None | 3.60 GHz |

| Socket | Intel BGA 1168 | AMD Socket FP4 |

| Architecture | Broadwell | Excavator |

| Codename | Broadwell-U | Bristol Ridge |

| Process Node | 14 nm | 28 nm |

| Foundry | Intel | GlobalFoundries |

| Transistors | 1,300 million | 3,100 million |

| Die Size | 82 mm² | 250 mm² |

| L1 Cache | 64 KB (per core) | 320 KB |

| L2 Cache | 256 KB (per core) | 1 MB (per module) |

| L3 Cache | 3 MB (shared) | None |

| Memory Support | DDR3 | DDR4 |

| Memory Bandwidth | 25.6 GB/s | 29.9 GB/s |

| PCIe | Gen 2, 12 Lanes | Gen 3, 8 Lanes |

| Integrated Graphics | Intel HD 5500 | Radeon R7 8CU |

| Release Date | 2015-02-28 | 2016-05-30 |

| Launch MSRP | $281 | None |

| Part Number | SR240 | FM980PADY44AB |

Shared specifications include a 15 W TDP, dual-channel memory bus, no ECC support, mobile market segment, end-of-life production status, and a locked multiplier. Both sit at the 20th percentile among all CPUs in the database, meaning neither is a standout performer in the broader CPU landscape.

FAQ

Q: Which processor is faster in Cinebench R23 multicore?

A: The AMD FX-9800P scores 2088 versus the Intel Core i3-5020U's 1876, giving AMD a 10.2% lead in that test.

Q: How large is the Intel lead in Geekbench single-core?

A: The Intel Core i3-5020U scores 696 against the AMD FX-9800P's 498, a 39.8% advantage. This is the largest margin in any head-to-head benchmark between the two.

Q: Do both processors have the same number of threads?

A: Yes, both have 4 threads. The Intel chip achieves this with 2 cores and Hyper-Threading, while the AMD chip uses 4 physical cores without SMT.

Q: Does the AMD FX-9800P have a boost clock?

A: Yes, the AMD chip has a 2.70 GHz base clock and a 3.60 GHz boost clock. The Intel Core i3-5020U has no boost capability, running at a fixed 2.20 GHz.

Q: Which processor supports DDR4 memory?

A: Only the AMD FX-9800P supports DDR4, with 29.9 GB/s bandwidth. The Intel Core i3-5020U is limited to DDR3 with 25.6 GB/s bandwidth.

Q: What are the average benchmark scores for each chip?

A: The Intel Core i3-5020U has an average benchmark score of 757, while the AMD FX-9800P scores 745. The Intel chip sits 0.1% above the Intel Xeon E5450 in the nearest rivals list, while the AMD chip sits 0.1% below the AMD A10 PRO-7800B.

The Verdict

The data supports a clear split decision. If your primary benchmark is Cinebench, the AMD FX-9800P is the better processor. It wins every Cinebench test, single-core and multi-core, by a consistent 10.2% to 10.6% margin. The four physical cores and higher clock speeds (2.70 GHz base, 3.60 GHz boost) give it a real advantage in rendering workloads. The AMD chip also brings DDR4 support with higher bandwidth (29.9 GB/s vs 25.6 GB/s) and PCIe Gen 3, though with fewer lanes.

If your primary benchmark is Geekbench, the Intel Core i3-5020U is the clear winner. It beats AMD by 22.4% in multicore and 39.8% in single-core. That single-core margin is enormous and points to a substantial IPC advantage from the Broadwell architecture on a 14 nm process. The Intel chip also has a shared 3 MB L3 cache, which the AMD chip lacks entirely, and a much smaller die (82 mm² vs 250 mm²) with lower transistor count (1,300 million vs 3,100 million), indicating better power efficiency per unit of work.

The average benchmark scores are nearly identical: 757 for Intel, 745 for AMD. Both sit at the 20th percentile among all CPUs. Neither chip is a performance leader, they are entry-level mobile parts from different generations and different design schools. The Intel chip launched earlier (February 2015) with a $281 MSRP, while the AMD chip arrived later (May 2016) with no recorded launch price.

For a buyer choosing between these two today, the decision hinges on software compatibility and workload. If you run Cinebench-style rendering tasks, the AMD FX-9800P offers a consistent 10% advantage. If you run Geekbench-style general-purpose applications, the Intel Core i3-5020U offers a 22% to 40% advantage. The Intel chip is also the safer bet for single-threaded responsiveness, where its 39.8% lead in Geekbench single-core suggests noticeably snappier everyday performance. The AMD chip is the safer bet for multi-threaded rendering, where its 10.2% lead in Cinebench R23 multicore shows real strength.

Both processors are end-of-life, so this is a used-market or legacy-system decision. The database shows two chips that are statistically tied on average (757 vs 745) but diametrically opposed in benchmark-specific results. Choose Intel for Geekbench workloads, choose AMD for Cinebench workloads. There is no universal winner here, only the right tool for the right test.