AMD FirePro W600 vs NVIDIA GeForce GTX 670M Comparison

# NVIDIA GeForce GTX 670M vs AMD FirePro W600

The GeForce GTX 670M and FirePro W600 represent two distinct 2012-era approaches to GPU design: one aimed at mobile performance, the other at professional workstation output. In the single available OpenCL benchmark, the GTX 670M scores 6513 against the FirePro W600's 6223, a 4.7% margin that places the NVIDIA part at the 38th percentile of all GPUs versus the AMD part's 36th percentile. While the raw score gap is modest, the architectural philosophies behind each card could not be more different, and those differences matter far more than the benchmark delta alone.

Head-to-Head Benchmarks

The only direct comparison available is the Geekbench OpenCL test, where the GTX 670M records 6513 points against the FirePro W600's 6223. That 290-point spread translates to a 4.7% advantage for the NVIDIA part. This is a narrow win—not the kind of margin that suggests a generational leap, but enough to put the GTX 670M ahead in a compute workload that exercises raw throughput.

Context from the nearest rival lists sharpens the picture. The GTX 670M sits just 0.3% above the GeForce GT 555M (6493) and 0.5% above the Quadro M5000M (6481), while trailing the Intel UHD Graphics P750 (6554) by 0.6%. The FirePro W600, meanwhile, beats the Radeon HD 6950 (6210) by 0.2% but falls 0.7% short of the RTX 4070 Ti SUPER AD102 and RTX 5070 Ti SUPER (both 6270), and 0.9% behind the Quadro K620 (6282). In other words, both cards cluster in a tight performance band where a few percent separates them from entirely different generations and market segments.

The 4.7% delta in this single test is the entire head-to-head story. There are no multi-test splits, no workloads where the FirePro W600 pulls ahead, no memory-bound or geometry-heavy scenarios to weigh. The data shows one winner with a modest but real edge in OpenCL compute. What is more revealing is how each card achieves its score—and that requires a closer look at the silicon.

Architecture Differences

The GTX 670M is built on NVIDIA's Fermi 2.0 architecture, using the GF114 chip manufactured on a 40 nm process at TSMC. That chip packs 1,950 million transistors into a 332 mm² die, yielding a transistor density of 5.9 million per square millimeter. The FirePro W600 counters with AMD's GCN 1.0 architecture, using the Cape Verde chip on a 28 nm process—also TSMC. Cape Verde fits 1,500 million transistors into just 123 mm², achieving a density of 12.2 million per square millimeter. The node advantage is clear: AMD crams nearly twice the transistors per area, which is why the smaller die actually carries more modern compute capabilities.

Shader configuration tells a different story. The GTX 670M fields 336 shading units, 56 texture mapping units, and 24 ROPs. The FirePro W600 has 512 shading units but only 32 TMUs and 16 ROPs. That asymmetry explains their respective rate figures: the GTX 670M delivers 33.49 GTexel/s and 8.372 GPixel/s, while the FirePro W600 manages 24.00 GTexel/s and 12.00 GPixel/s. NVIDIA's part textures faster; AMD's part rasterizes faster. In raw FP32 compute, however, the GTX 670M edges ahead at 803.7 GFLOPS versus 768.0 GFLOPS—a 4.7% margin that mirrors the OpenCL benchmark result almost exactly.

Memory subsystems diverge as well. The GTX 670M uses 1536 MB of GDDR5 on a 192-bit bus, producing 72.00 GB/s of bandwidth. The FirePro W600 has 2 GB of GDDR5 on a narrower 128-bit bus, yielding 64.00 GB/s. Memory clocks are 750 MHz (3 Gbps effective) for the NVIDIA part and 1000 MHz (4 Gbps effective) for the AMD part, but the wider bus gives the GTX 670M the throughput advantage. Both cards draw a 75 W TDP and require no power connectors, but the FirePro W600 suggests a 250 W PSU while the GTX 670M lists none.

Interface and form factor differences are stark. The GTX 670M is an MXM Module with an MXM-B (3.0) bus interface, designed for laptops with display outputs that are portable-device-dependent. The FirePro W600 is a single-slot, 168 mm long PCIe 3.0 x16 card with six mini-DisplayPort 1.2 outputs—a workstation-oriented configuration for multi-monitor professional setups. API support also favors the AMD card: it lists DirectX 12 (11_1), OpenGL 4.6, and Vulkan 1.2.170, while the GTX 670M only lists DirectX 12 (11_0) and OpenGL 4.6 with no Vulkan support.

Where Each One Wins

The GTX 670M wins in compute throughput and memory bandwidth. Its 803.7 GFLOPS FP32 figure tops the FirePro W600's 768.0 GFLOPS, and its 72.00 GB/s bandwidth exceeds the AMD card's 64.00 GB/s. The 33.49 GTexel/s texture rate is also 39.5% higher than the FirePro W600's 24.00 GTexel/s, which matters for texture-heavy workloads. The benchmark result confirms this: 6513 versus 6223 in OpenCL, a 4.7% win.

The FirePro W600 wins in pixel throughput and configuration flexibility. Its 12.00 GPixel/s pixel rate is 43.4% higher than the GTX 670M's 8.372 GPixel/s, making it the better choice for fill-rate-bound rendering tasks. It also has more memory (2 GB versus 1536 MB), a modern 28 nm process, and six display outputs versus the GTX 670M's portable-device-dependent outputs. The Vulkan 1.2.170 support is a feature the NVIDIA part completely lacks.

Release timing also differs. The GTX 670M launched on 2012-03-21, while the FirePro W600 followed on 2012-06-12. Both are end-of-life, with the GTX 670M succeeding the GeForce 500M and preceding the GeForce 700M, while the FirePro W600 follows FirePro Terascale and precedes Radeon Pro Polaris.

FAQ

Q: Which GPU has higher FP32 compute performance?

A: The NVIDIA GeForce GTX 670M delivers 803.7 GFLOPS, which is 4.7% higher than the AMD FirePro W600's 768.0 GFLOPS. This aligns with the 4.7% OpenCL benchmark margin.

Q: What are the memory configuration differences?

A: The GTX 670M uses 1536 MB of GDDR5 on a 192-bit bus with 72.00 GB/s bandwidth. The FirePro W600 uses 2 GB of GDDR5 on a 128-bit bus with 64.00 GB/s bandwidth. The NVIDIA card has higher bandwidth despite less capacity.

Q: Does the FirePro W600 support Vulkan?

A: Yes, the FirePro W600 lists Vulkan 1.2.170 support. The GTX 670M lists no Vulkan support in its API specifications, offering only DirectX 12 (11_0) and OpenGL 4.6.

Q: How do the process nodes compare?

A: The GTX 670M is built on TSMC's 40 nm process with 1,950 million transistors on a 332 mm² die. The FirePro W600 uses TSMC's 28 nm process with 1,500 million transistors on a 123 mm² die, giving it a transistor density of 12.2M per mm² versus 5.9M per mm².

Q: Which card has more shading units?

A: The FirePro W600 has 512 shading units, while the GTX 670M has 336. However, the GTX 670M has more TMUs (56 versus 32) and ROPs (24 versus 16), which contributes to its higher texture rate.

Q: What is the form factor of each card?

A: The GTX 670M is an MXM Module with an MXM-B (3.0) bus interface, intended for portable devices. The FirePro W600 is a single-slot PCIe 3.0 x16 card measuring 168 mm in length, with six mini-DisplayPort 1.2 outputs.

The Verdict

The data points to a clear but narrow winner in raw compute: the GTX 670M takes the OpenCL benchmark 6513 to 6223, a 4.7% edge that is consistent across its FP32 throughput and memory bandwidth advantages. For workloads that are texture-heavy or bandwidth-limited, the NVIDIA part is the better choice. Its 72.00 GB/s bandwidth and 33.49 GTexel/s texture rate are both superior, and it does so within the same 75 W TDP envelope.

However, the FirePro W600 is not a loser—it is a different tool. Its 12.00 GPixel/s pixel rate is 43.4% higher than the GTX 670M's, making it preferable for fill-rate-bound scenarios. The 28 nm process yields a smaller, denser chip. The six mini-DisplayPort outputs and PCIe 3.0 x16 interface make it a workstation card, not a mobile part. The Vulkan 1.2.170 support gives it modern API coverage that the GTX 670M lacks entirely.

Who should pick which? If the task is portable compute with maximum OpenCL throughput and bandwidth, the GTX 670M is the data-backed choice. If the task is professional multi-display output, fill-rate-heavy rendering, or needs the latest API support, the FirePro W600 is the only option that fits. The benchmark says the GTX 670M is 4.7% faster in one test; the specification sheet says the FirePro W600 is built for a different job entirely. The verdict is not about which is "better"—it is about which matches the workload. The data supports the GTX 670M for compute and the FirePro W600 for workstation display tasks.