AMD Radeon HD 8970M vs NVIDIA GeForce RTX 2080 Comparison

The NVIDIA GeForce RTX 2080 and AMD Radeon HD 8970M represent two vastly different eras of GPU design, separated by five years of architectural evolution and a fundamental shift in performance targets. The data available for direct comparison is limited to a single shared benchmark, but that result is decisive. In the Geekbench OpenCL test, the RTX 2080 scores 91,313, while the HD 8970M scores 21,237. This yields a delta of 330%, meaning the RTX 2080 delivers more than four times the compute performance of the older mobile part. The RTX 2080 wins the only head-to-head benchmark, securing a 1-0 victory. While the HD 8970M holds a respectable 66th percentile ranking among all GPUs, the RTX 2080 sits at the 68th percentile, a gap that understates the massive raw performance difference because the mobile GPU’s percentile is buoyed by its historical context.

Head-to-Head Benchmarks

The sole direct comparison available is the Geekbench OpenCL test, and it paints a stark picture of generational progress. The RTX 2080 achieves a score of 91,313, which is 330% higher than the HD 8970M’s 21,237. This is not a marginal improvement; it is a categorical leap. The RTX 2080’s average benchmark score across all tests is 22,895, which is actually higher than the HD 8970M’s average of 21,237, despite the RTX 2080 being tested across a broader suite of workloads. The HD 8970M’s nearest rivals in the database include the AMD Radeon RX Vega M GL, which scores 21,153 (a 0.4% delta), and the NVIDIA RTX A4000 Mobile at 21,379 (a -0.7% delta). This places the HD 8970M in a performance tier that is roughly equivalent to low-end mobile discrete GPUs from nearly a decade later. Conversely, the RTX 2080’s nearest rivals include the Intel Arc B580 at 23,021 (-0.5% delta) and the NVIDIA GeForce RTX 3080 at 23,172 (-1.2% delta), showing it competes with modern mid-range and high-end desktop parts. The 330% delta in the head-to-head test is consistent with the massive differences in raw hardware specifications, such as shading units and memory bandwidth, which will be explored in the architecture section.

Architecture Differences

The architectural chasm between these two GPUs is immense. The RTX 2080 is built on NVIDIA’s Turing architecture, fabricated on a 12 nm process at TSMC, while the HD 8970M uses AMD’s GCN 1.0 architecture on a 28 nm process, also from TSMC. The transistor counts reflect this generational gap: the RTX 2080 packs 13,600 million transistors on a 545 mm² die, resulting in a density of 25.0M transistors per mm². The HD 8970M, in contrast, has only 2,800 million transistors on a 212 mm² die, with a density of 13.2M/mm². This means the RTX 2080 has nearly five times the transistor budget, enabling far more complex compute units. The RTX 2080 features 2,944 shading units, 184 texture mapping units (TMUs), and 64 render output units (ROPs). The HD 8970M has 1,280 shading units, 80 TMUs, and 32 ROPs. The RTX 2080 also introduces dedicated hardware that the HD 8970M entirely lacks: 46 ray tracing cores and 368 tensor cores. These are purpose-built for real-time ray tracing and AI-accelerated workloads, respectively, and their absence in the GCN 1.0 architecture means the HD 8970M cannot perform these tasks at any meaningful speed. Clock speeds also differ, with the RTX 2080 boosting to 1710 MHz versus the HD 8970M’s 900 MHz boost, though the older card’s lower core count makes this comparison less relevant.

Memory subsystems are equally divergent. The RTX 2080 uses 8 GB of GDDR6 memory on a 256-bit bus, delivering 448.0 GB/s of bandwidth. The HD 8970M uses 4 GB of GDDR5 on the same 256-bit bus, but only achieves 153.6 GB/s. This 2.9x bandwidth advantage is critical for modern games and compute tasks that are memory-bound. The RTX 2080’s memory clock runs at 1750 MHz (14 Gbps effective), while the HD 8970M’s memory is at 1200 MHz (4.8 Gbps effective). Pixel and texture fill rates tell a similar story: the RTX 2080 outputs 109.4 GPixel/s and 314.6 GTexel/s, versus 28.80 GPixel/s and 72.00 GTexel/s for the HD 8970M. Floating-point performance is where the gap widens most dramatically: the RTX 2080 delivers 10.07 TFLOPS FP32, while the HD 8970M manages just 2.304 TFLOPS FP32. The RTX 2080 also supports FP16 at 20.14 TFLOPS (2:1 ratio), a feature the HD 8970M does not expose. API support differs as well: the RTX 2080 supports DirectX 12 Ultimate (12_2) and Vulkan 1.4, whereas the HD 8970M is limited to DirectX 12 (11_1) and Vulkan 1.2.170. Both support OpenGL 4.6. The RTX 2080 also has a much higher power envelope at 215 W TDP, versus 100 W for the HD 8970M, reflecting its desktop-class performance ambitions.

FAQ

Q: Which GPU is faster in the only shared benchmark?

A: The NVIDIA GeForce RTX 2080 is significantly faster, scoring 91,313 in Geekbench OpenCL compared to the AMD Radeon HD 8970M’s 21,237. This represents a 330% performance advantage for the RTX 2080.

Q: Does the AMD Radeon HD 8970M support ray tracing or AI acceleration?

A: No. The HD 8970M is based on GCN 1.0 architecture and has no ray tracing cores or tensor cores. The RTX 2080 includes 46 ray tracing cores and 368 tensor cores for those workloads.

Q: How do their memory bandwidths compare?

A: The RTX 2080 has a memory bandwidth of 448.0 GB/s using 8 GB of GDDR6 on a 256-bit bus. The HD 8970M has 153.6 GB/s using 4 GB of GDDR5 on the same 256-bit bus, making the RTX 2080 roughly 2.9 times faster in memory bandwidth.

Q: What is the transistor density difference?

A: The RTX 2080 has a transistor density of 25.0M per mm² on a 12 nm process, while the HD 8970M has a density of 13.2M per mm² on a 28 nm process. The RTX 2080 also has a much larger die at 545 mm² versus 212 mm².

Q: Which GPU has a higher average benchmark score?

A: The RTX 2080 has an average benchmark score of 22,895 across all tests, while the HD 8970M has an average of 21,237. The RTX 2080’s score is higher, but the HD 8970M is only tested in one benchmark in the database.

Q: Are both GPUs still in production?

A: No. Both are marked as end-of-life products. The RTX 2080 was released on 2018-09-19, while the HD 8970M was released on 2013-05-13.

Specification Differences

The specification tables for these two GPUs diverge on nearly every key metric. The process node differs: the RTX 2080 uses 12 nm, while the HD 8970M uses 28 nm. Transistor count is 13,600 million for the RTX 2080 versus 2,800 million for the HD 8970M. Die size is 545 mm² versus 212 mm². The RTX 2080 has a base clock of 1515 MHz and a boost of 1710 MHz, while the HD 8970M has a base of 850 MHz and a boost of 900 MHz. Memory size is 8 GB GDDR6 for the RTX 2080, versus 4 GB GDDR5 for the HD 8970M. Memory bandwidth is 448.0 GB/s versus 153.6 GB/s. Shading units number 2,944 versus 1,280; TMUs number 184 versus 80; ROPs number 64 versus 32. The RTX 2080 has 46 RT cores and 368 tensor cores; the HD 8970M has none. Pixel rate is 109.4 GPixel/s versus 28.80 GPixel/s. Texture rate is 314.6 GTexel/s versus 72.00 GTexel/s. FP32 performance is 10.07 TFLOPS versus 2.304 TFLOPS. The RTX 2080 has FP16 performance of 20.14 TFLOPS (2:1), while the HD 8970M has no listed FP16 capability. TDP is 215 W versus 100 W. The RTX 2080 is a dual-slot card with 1x 6-pin and 1x 8-pin connectors, requiring a 550 W PSU, while the HD 8970M is an MXM module with no listed power connectors or PSU requirement. The RTX 2080 has display outputs of 1x HDMI 2.0, 3x DisplayPort 1.4a, and 1x USB Type-C, while the HD 8970M’s outputs are portable-device dependent. The RTX 2080 supports DirectX 12 Ultimate (12_2) and Vulkan 1.4, versus DirectX 12 (11_1) and Vulkan 1.2.170 for the HD 8970M. The RTX 2080 is 267 mm long, 116 mm high, and 35 mm wide; the HD 8970M has no listed dimensions. The RTX 2080’s predecessor is GeForce 10 and successor is GeForce 30; the HD 8970M’s predecessor is London and successor is Gem System.

The Verdict

The data unequivocally favors the NVIDIA GeForce RTX 2080 for any user requiring maximum compute performance. Its 330% lead in the only shared benchmark, combined with a 4.4x advantage in FP32 throughput (10.07 TFLOPS versus 2.304 TFLOPS) and a 2.9x advantage in memory bandwidth, makes it the obvious choice for gaming, content creation, or any GPU-accelerated workload. The RTX 2080 also brings modern features like ray tracing and tensor cores, which the HD 8970M simply cannot offer. The RTX 2080’s percentile ranking of 68th places it among contemporary desktop GPUs like the RTX 3080, while the HD 8970M’s 66th percentile is misleading—it competes with low-end mobile parts like the RTX 5050 and RX Vega M GL. For a user with a desktop system that can accommodate a dual-slot, 215 W card, the RTX 2080 is the superior product in every measurable way. The HD 8970M, with its 100 W TDP and MXM form factor, is only relevant for legacy laptops where upgradability is constrained by the chassis. In such a scenario, the HD 8970M provides a baseline level of performance that is now several generations old, but it remains the only option for that specific hardware class. For all other use cases, the RTX 2080’s raw power, modern feature set, and higher average benchmark score make it the definitive choice. There is no benchmark scenario in the data where the HD 8970M surpasses the RTX 2080.