AMD Radeon 680M vs AMD Radeon HD 7970M Comparison

Head-to-Head Benchmarks

The only direct benchmark comparison recorded in the database is the Geekbench OpenCL test. In this test, the AMD Radeon 680M scores 23,468 points, while the AMD Radeon HD 7970M scores 17,019 points. The delta of -27.5% indicates the HD 7970M trails the 680M by more than a quarter in raw compute throughput. This is a decisive victory for the newer integrated part.

Looking at how each card sits relative to its own nearest rivals, the HD 7970M's 17,019 OpenCL score places it between the NVIDIA Tesla M4 at 16,932 (+0.5% for the HD 7970M) and the NVIDIA GeForce GTX 690 at 17,037 (-0.1%). The gap to the AMD Radeon RX 7600 XT at 17,083 is only -0.4%, and the NVIDIA GeForce RTX 3070 at 17,208 shows a -1.1% deficit for the HD 7970M. The data indicates the HD 7970M is effectively locked in a tight cluster with these four rivals, with all deltas within roughly one percentage point. Its percentile ranking of 60 among all GPUs reflects this mid-pack positioning.

The Radeon 680M, by contrast, produces an OpenCL score of 23,468, which is a substantial step up. Its nearest rival, the NVIDIA GeForce GTX 580, averages 15,283, putting the 680M ahead by -0.1% in the delta column (the negative sign indicates the rival's score is lower). The NVIDIA GeForce RTX 2060 at 15,290 is similarly close, while the NVIDIA GeForce RTX 3050 OEM at 15,199 and the AMD Radeon RX 7600 at 15,171 are both within 0.7% of the 680M's average. Despite this, the 680M's percentile ranking is 57, slightly lower than the HD 7970M's 60, because the 680M's average benchmark score of 15,270 is dragged down by its other recorded tests, including a 3DMark Steel Nomad DX12 score of 378 and a Geekbench Vulkan score of 21,965.

The OpenCL result is the only head-to-head data point, and it shows a clean win for the 680M. However, the HD 7970M still holds a higher percentile position overall, which suggests that the OpenCL comparison alone does not fully capture the two parts' relative standing across all workloads. The 680M's Vulkan score of 21,965 is also notably higher than the HD 7970M's OpenCL figure, though no direct Vulkan measurement exists for the older card in this database.

Architecture Differences

The two GPUs come from completely different architectural generations. The HD 7970M is built on GCN 1.0, with the chip codenamed Wimbledon, and belongs to the London (HD 7900M) generation. It uses a 28 nm process at TSMC, with 2,800 million transistors packed into a 212 mm² die, yielding a transistor density of 13.2 million transistors per square millimeter. The 680M uses RDNA 2.0 architecture on the Rembrandt+ chip, part of the Navi II IGP (Rembrandt Mobile) generation. It is manufactured on TSMC's 6 nm node, housing 13,100 million transistors across a 208 mm² die, for a density of 63.0 million per square millimeter, nearly five times denser than the HD 7970M.

The compute layout also diverges sharply. The HD 7970M fields 1,280 shading units, 80 texture mapping units (TMUs) and 32 render output units (ROP units). Its FP32 throughput is rated at 2.176 TFLOPS, with pixel rate at 27.20 GPixel/s and texture rate at 68.00 GTexel/s. The 680M has 768 shading units, 48 TMUs and 32 ROPs, but outputs 3.379 TFLOPS FP32, 6.758 TFLOPS FP16 (at a 2:1 ratio), 70.40 GPixel/s and 105.6 GTexel/s. The RDNA 680M achieves higher texture and pixel throughput with fewer TMUs, reflecting the efficiency of the newer architecture. The 680M also includes 12 ray tracing cores, a feature entirely absent from the HD 7970M, which has none.

Memory subsystems could not be more different. The HD 7970M uses 2 GB of GDDR5 VRAM on a 256-bit bus, with memory clocks set at 1,200 MHz (4.8 Gbps effective, yielding 153.6 GB/s bandwidth. The 680M relies on system shared memory, with no dedicated VRAM capacity, bus width or bandwidth figure; its memory is marked system dependent and shared with the host. The HD 7970M's TDP is rated at 100 W, while the 680M draws 50 W, half the power budget. The HD 7970M interfaces via MXM-B (3.0), the 680M via PCIe 4.0 x8. The HD 7970M's production status is end-of-life, released on April 23, 2012, while the 680M is active, released on January 2, 2023.

API support differs as well. The HD 7970M supports DirectX 12 (11_1), OpenGL 4.6 and Vulkan 1.2.170. The 680M supports DirectX 12 Ultimate (12_2), OpenGL 2.2, OpenGL 4.6 and Vulkan 1.2.170. The 680M lists DirectX 12 Ultimate (12_2, Vulkan 1.4, OpenGL 4.6, Vulkan 1.2.170 and Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6 and Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, Vulkan 1.4, OpenGL 4.6 and Vulkan 1.2.2. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6 and Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6 and Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6 and Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6 and Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6 and Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6, Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6 Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6, Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6 Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6 Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 1.2.170, Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6 Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6 Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6, Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6, Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6, Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6, Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6, Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6, Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6, Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6, Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6, Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6, Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6, Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6, Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6, Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6, Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6, Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6, Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6, Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6, Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6, Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6, Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6, Vulkan 1.4. The 680M lists DirectX 12 Ultimate (12_2, OpenGL 4.6, Vulkan 1.4.

FAQ

Q: Which GPU has the higher Geekbench OpenCL score?

A: The AMD Radeon 680M scores 23,468 points, compared to the AMD Radeon HD 7970M's 17,019 points, a delta of -27.5% against the HD 7970M.

Q: How does the HD 7970M compare to its nearest rivals?

A: The HD 7970M's 17,019 score sits within 1.1% of the NVIDIA GeForce GTX 690 (17,037), the AMD Radeon RX 7600 XT (17,083), the NVIDIA Tesla M4 (16,932) and the NVIDIA GeForce RTX 3070 (17,208).

Q: What is the transistor count and process node for each?

A: The HD 7970M has 2,800 million transistors on a 28 nm process, while the 680M has 13,100 million transistors on a 6 nm process, both fabricated by TSMC.

Q: Does the Radeon 680M support ray tracing?

A: Yes, the 680M includes 12 ray tracing cores. The HD 7970M has no ray tracing cores listed.

Q: What are the FP32 compute ratings?

A: The HD 7970M is rated at 2.176 TFLOPS FP32, while the 680M is rated at 3.379 TFLOPS FP32, a 55% advantage for the 680M in this metric.

Q: What memory configurations do the two GPUs use?

A: The HD 7970M uses 2 GB of GDDR5 VRAM on a 256-bit bus with 153.6 GB/s bandwidth. The 680M uses system shared memory with bandwidth listed as system dependent.

Specification Differences

The HD 7970M and 680M differ across nearly every major specification field. The HD 7970M uses 28 nm process, the 680M uses 6 nm. Transistor counts are 2,800 million versus 13,100 million. Die size is 212 mm² for the HD 7970M and 208 mm² for the 680M, with densities of 13.2 million per mm² versus 63.0 million per mm². The HD 7970M's shading units total 1,280, TMUs 80, ROPs 32, with no RT cores. The 680M has 768 shading units, 48 TMUs, 32 ROPs and 12 RT cores.

Clock speeds differ: the HD 7970M lists no base or boost clock in the database, only memory at 1,200 MHz (4.8 Gbps effective). The 680M lists base at 2,000 MHz, boost at 2,200 MHz, memory as system shared. Pixel rate is 27.20 GPixel/s for the HD 7970M versus 70.40 GPixel/s for the 680M. Texture rates are 68.00 GTexel/s versus 105.6 GTexel/s. FP32 is 2.176 TFLOPS versus 2.176 TFLOPS 2.176 TFLOPS. 2.176 TFLOPS 2.176 TFLOPS. FP32 2.176 TFLOPS versus 3.379 TFLOPS. FP16 is null for the HD 7970M, 6.758 TFLOPS (2:1) for the 680M. TDP is 100 W versus 50 W. The bus interface is MXM-B (3.0) versus PCIe 4.0 x8. The HD 7970M is end-of-life, released 2012-04-23, while the 680M is active, released 2023-01-02. API support: HD 7970M has DirectX 12 (11_1), Vulkan 1.2.170; 680M has DirectX 12 Ultimate (12_2), Vulkan 1.4. The 680M has no fixed memory size, bus width or bandwidth, all system dependent. Both have no power connectors, no suggested PSU, and display outputs marked portable device dependent.

Where Each One Wins

The AMD Radeon 680M is the clear winner in raw OpenCL compute, scoring 23,468 versus 17,019, a 27.5% advantage. It also wins on architecture features: ray tracing cores, higher FP32 throughput (3.379 TFLOPS), higher FP16 throughput (6.758 TFLOPS), and a much lower TDP of 50 W versus 100 W. Its higher clock speeds (2,000 MHz base, 2,200 MHz boost) and higher pixel/texture rates (70.40 GPixel/s, 105.6 GTexel/s) make it the better choice for compute-heavy workloads that leverage modern APIs like DirectX 12 Ultimate and Vulkan 1.4. The 680M's nearest rivals are all within 0.7% of its average score, indicating it sits in a competitive but tight performance band.

The AMD Radeon HD 7970M wins in the percentile comparison, holding a 60th percentile ranking versus the 680M's 57th. Its dedicated 2 GB GDDR5 memory on a 256-bit bus provides 153.6 GB/s of bandwidth, a fixed resource that does not depend on system memory. In the OpenCL test it is only 0.1% behind the GTX 690 and 0.4% behind the RX 7600 XT, showing that despite its age, it remains within a percentage point of those parts. It also has a higher shading unit count (1,280) and TMU count (80) than the 680M, though this does not translate to a higher benchmark score. For scenarios where dedicated VRAM is essential and system memory bandwidth is a limiting factor, the HD 7970M's fixed memory configuration could be advantageous.

The recorded benchmark data shows the 680M winning the only direct head-to-head test, while the HD 7970M retains a higher overall percentile. This suggests the HD 7970M performs relatively better across the broader set of all recorded GPUs, but the 680M is the stronger part in the specific OpenCL workload tested. The 680M's Vulkan score of 21,965 further reinforces its modern API strength, and its active production status versus the HD 7970M's end-of-life status indicates continued support. The HD 7970M remains competitive in its narrow band, but the architecture and feature advantages of the 680M make it the more capable GPU in current workloads.