AMD Radeon HD 8870M vs AMD Radeon Vega 8 Comparison
AMD Radeon HD 8870M
Radeon Vega 8
PERFORMANCE BENCHMARKS
Analysis: AMD Radeon HD 8870M vs AMD Radeon Vega 8
FAQ
Q: Which GPU has the higher average benchmark score?
A: The AMD Radeon Vega 8 leads with an average benchmark score of 9,221, while the AMD Radeon HD 8870M trails at 8,462. This places the Vega 8 in the 45th percentile of all GPUs, versus the 43rd percentile for the HD 8870M.
Q: How do the two compare in the Geekbench OpenCL test?
A: The Vega 8 scores 8,822 in Geekbench OpenCL, which is 4.3% ahead of the HD 8870M's 8,462. This is the only head-to-head benchmark available, and the Vega 8 wins it.
Q: What are the closest rivals to the Vega 8?
A: The nearest rivals to the Vega 8 include the AMD Radeon 890M (0.1% higher average score), the NVIDIA GeForce GTX 960 (0.6% lower), the NVIDIA GeForce GTX 465 (0.8% lower), and the NVIDIA GeForce GTX 850M (0.9% lower).
Q: What are the closest rivals to the HD 8870M?
A: The nearest rivals to the HD 8870M include the NVIDIA GeForce MX330 (0.1% higher average score), the AMD Radeon 880M (0.3% higher), the NVIDIA GeForce GTX 675MX (0.4% higher), and the Intel Arc A380 (1.1% lower).
Q: Which GPU has a greater number of shading units?
A: The HD 8870M has 640 shading units, which is more than the Vega 8's 512 shading units. However, the Vega 8 achieves a higher FP32 throughput of 1,126.4 GFLOPS versus 992.0 GFLOPS for the HD 8870M.
Q: Are both GPUs end-of-life products?
A: Yes, both the Radeon Vega 8 and the Radeon HD 8870M are marked as end-of-life. The Vega 8 was released on February 11, 2018, while the HD 8870M was released on March 31, 2013.
Architecture Differences
The AMD Radeon Vega 8 and AMD Radeon HD 8870M represent two distinct generations of AMD graphics architecture. The Vega 8 is built on GCN 5.0, using the Raven chip, and belongs to the Vega IGP (Raven Ridge) generation. In contrast, the HD 8870M is based on the older GCN 1.0 architecture, using the Venus chip, and belongs to the Solar System (HD 8800M) generation.
The manufacturing process differs significantly between the two. The Vega 8 is fabricated on a 14 nm process at GlobalFoundries, while the HD 8870M uses a 28 nm process at TSMC. This process advantage contributes to the Vega 8's higher transistor density of 23.5M per mm², compared to 12.2M per mm² for the HD 8870M. The Vega 8 packs 4,940 million transistors on a 210 mm² die, whereas the HD 8870M has 1,500 million transistors on a 123 mm² die.
Memory architecture is a major dividing line. The Vega 8 uses system-shared memory, with its memory type, bus width, and bandwidth all described as "System Shared" or "System Dependent." The HD 8870M, by contrast, comes with 2 GB of dedicated GDDR5 memory on a 128-bit bus, delivering 72.00 GB/s of bandwidth. This dedicated memory gives the HD 8870M a fixed memory performance profile, whereas the Vega 8's memory performance depends entirely on the host system.
The Vega 8 has 512 shading units, 32 texture mapping units, and 8 ROPs. The HD 8870M has more of each: 640 shading units, 40 TMUs, and 16 ROPs. Despite having fewer units, the Vega 8 achieves a higher pixel rate of 8.800 GPixel/s and a higher texture rate of 35.20 GTexel/s compared to the HD 8870M's 12.40 GPixel/s and 31.00 GTexel/s. The Vega 8 also leads in FP32 performance at 1,126.4 GFLOPS, and it supports FP16 at 2.253 TFLOPS (2:1), while the HD 8870M has no listed FP16 capability.
Clock speeds differ substantially. The Vega 8 has a base clock of 300 MHz and a boost clock of 1,100 MHz. The HD 8870M runs at a higher base of 725 MHz but a lower boost of 775 MHz. The Vega 8's wide clock range reflects its IGP nature, where power management is more flexible. The HD 8870M's memory clock is listed at 1,125 MHz with 4.5 Gbps effective, while the Vega 8's memory clock is "System Shared."
The bus interface also differs: the Vega 8 is an IGP with no power connectors and a 25 W TDP, while the HD 8870M uses PCIe 3.0 x16. The Vega 8 supports DirectX 12 (12_1), OpenGL 4.6, and Vulkan 1.3. The HD 8870M supports DirectX 12 (11_1), OpenGL 4.6, and Vulkan 1.2.170. The Vega 8's display outputs are motherboard dependent, while the HD 8870M has no display output information listed.
The Verdict
Based strictly on benchmark data, the AMD Radeon Vega 8 is the stronger performer. Its average benchmark score of 9,221 is 9.0% higher than the HD 8870M's 8,462. The Vega 8 also wins the only direct head-to-head comparison, the Geekbench OpenCL test, by 4.3%. In terms of overall GPU ranking, the Vega 8 sits at the 45th percentile, two points above the HD 8870M's 43rd percentile.
The Vega 8's closest rival, the AMD Radeon 890M, scores 9,210, which is only 0.1% lower than the Vega 8's average. The HD 8870M's closest rival, the NVIDIA GeForce MX330, scores 8,458, which is 0.1% higher than the HD 8870M. This suggests the Vega 8 competes in a slightly higher performance tier than the HD 8870M.
For users choosing between the two, the data clearly favors the Vega 8. It wins the available benchmark, has a higher average score, and supports newer API versions including Vulkan 1.3 versus 1.2.170. The HD 8870M does offer dedicated GDDR5 memory with 72.00 GB/s bandwidth, which may benefit workloads that rely on consistent memory performance, but this advantage does not translate into a benchmark win in the data provided.
The Vega 8 is the better choice for anyone prioritizing raw compute performance and modern API support. The HD 8870M remains a viable option only if dedicated memory is a strict requirement, but the benchmark results show it lagging behind in overall performance.
Specification Differences
| Specification | AMD Radeon Vega 8 | AMD Radeon HD 8870M |
|---|---|---|
| Chip | Raven | Venus |
| Architecture | GCN 5.0 | GCN 1.0 |
| Generation | Vega IGP (Raven Ridge) | Solar System (HD 8800M) |
| Process Node | 14 nm | 28 nm |
| Foundry | GlobalFoundries | TSMC |
| Transistors | 4,940 million | 1,500 million |
| Die Size | 210 mm² | 123 mm² |
| Transistor Density | 23.5M / mm² | 12.2M / mm² |
| Base Clock | 300 MHz | 725 MHz |
| Boost Clock | 1,100 MHz | 775 MHz |
| Memory Size | System Shared | 2 GB |
| Memory Type | System Shared | GDDR5 |
| Memory Bus Width | System Shared | 128 bit |
| Memory Bandwidth | System Dependent | 72.00 GB/s |
| Shading Units | 512 | 640 |
| TMUs | 32 | 40 |
| ROPs | 8 | 16 |
| Pixel Rate | 8.800 GPixel/s | 12.40 GPixel/s |
| Texture Rate | 35.20 GTexel/s | 31.00 GTexel/s |
| FP32 Performance | 1,126.4 GFLOPS | 992.0 GFLOPS |
| FP16 Performance | 2.253 TFLOPS (2:1) | Not listed |
| TDP | 25 W | Not listed |
| Bus Interface | IGP | PCIe 3.0 x16 |
| Power Connectors | None | Not listed |
| DirectX Support | 12 (12_1) | 12 (11_1) |
| Vulkan Support | 1.3 | 1.2.170 |
| Display Outputs | Motherboard Dependent | Not listed |
| Predecessor | GCN 3.0 IGP | London |
| Successor | Vega II IGP | Gem System |
| Release Date | 2018-02-11 | 2013-03-31 |
Head-to-Head Benchmarks
The only direct benchmark comparison between the two GPUs is the Geekbench OpenCL test. In this test, the AMD Radeon Vega 8 scores 8,822, while the AMD Radeon HD 8870M scores 8,462. The Vega 8 wins by 4.3%, a clear but not overwhelming margin. This result aligns with the overall average benchmark scores, where the Vega 8's 9,221 exceeds the HD 8870M's 8,462 by a larger 9.0% gap.
The Vega 8 also has additional benchmark results that the HD 8870M lacks. The Vega 8 scores 10,706 in Geekbench Metal and 8,134 in Geekbench Vulkan. These results are not available for the HD 8870M, so they cannot be directly compared. However, the Vega 8's Vulkan score of 8,134 is notably lower than its OpenCL score of 8,822, suggesting the Vega 8 performs better under OpenCL than under Vulkan.
The HD 8870M's single benchmark score of 8,462 places it close to its nearest rivals. The NVIDIA GeForce MX330 scores 8,458 (0.1% higher), the AMD Radeon 880M scores 8,436 (0.3% higher), and the NVIDIA GeForce GTX 675MX scores 8,427 (0.4% higher). Only the Intel Arc A380 scores lower at 8,558 (1.1% lower), but that comparison is inverted: the Arc A380 actually scores higher than the HD 8870M.
The Vega 8's average score of 9,221 sits in a tighter competitive cluster. The AMD Radeon 890M scores 9,210 (0.1% lower), the NVIDIA GeForce GTX 960 scores 9,273 (0.6% higher), the NVIDIA GeForce GTX 465 scores 9,294 (0.8% higher), and the NVIDIA GeForce GTX 850M scores 9,302 (0.9% higher). This places the Vega 8 just below a group of mid-range discrete GPUs, while the HD 8870M sits just above a group of lower-tier mobile GPUs.
Where Each One Wins
The AMD Radeon Vega 8 wins in overall compute performance. Its average benchmark score of 9,221 is higher, and it wins the only head-to-head test by 4.3%. The Vega 8 also has a higher FP32 throughput of 1,126.4 GFLOPS and supports FP16 at 2.253 TFLOPS, which the HD 8870M does not list. For workloads that rely on raw compute, such as OpenCL-based applications, the Vega 8 is the clear choice.
The Vega 8 also wins on modern API support. It supports DirectX 12 (12_1) and Vulkan 1.3, while the HD 8870M only supports DirectX 12 (11_1) and Vulkan 1.2.170. This means the Vega 8 is better suited for newer games and applications that leverage the latest graphics features. The Vega 8's 14 nm process also gives it a transistor density advantage of 23.5M per mm² versus 12.2M per mm², which contributes to its efficiency.
The AMD Radeon HD 8870M wins on dedicated memory. It has 2 GB of GDDR5 memory on a 128-bit bus with 72.00 GB/s bandwidth, whereas the Vega 8 relies on system-shared memory with bandwidth that is system dependent. For workloads that require consistent memory performance without relying on the host system's RAM, the HD 8870M has an advantage. The HD 8870M also has more shading units (640 vs 512), more TMUs (40 vs 32), and more ROPs (16 vs 8), which gives it a higher pixel rate of 12.40 GPixel/s compared to the Vega 8's 8.800 GPixel/s.
The HD 8870M's higher base clock of 725 MHz (versus 300 MHz for the Vega 8) may also benefit workloads that operate at sustained frequencies without boosting. However, the Vega 8's boost clock of 1,100 MHz is substantially higher than the HD 8870M's 775 MHz, so the Vega 8 can reach much higher peak performance when power and thermal headroom allow.
For gaming, the data suggests the Vega 8 is the stronger choice due to its higher average score and newer API support. For compute tasks that emphasize pixel throughput or require dedicated memory, the HD 8870M has specific advantages. Ultimately, the Vega 8 wins on overall performance and modern feature support, while the HD 8870M offers dedicated memory and higher raw pixel rates.