NVIDIA GeForce GT 635M vs NVIDIA GeForce GTX 460M Comparison

Where Each One Wins

The recorded benchmark data splits cleanly between these two mobile NVIDIA parts, though the split is heavily lopsided. The NVIDIA GeForce GTX 460M wins the only head-to-head benchmark category present in the database, the Geekbench OpenCL test, with a score of 4282 against the GT 635M's 3740. That is a 14.5% advantage for the older, higher-tier part. In terms of raw compute output, the GTX 460M delivers 518.4 GFLOPS of FP32 performance, while the GT 635M manages only 182.4 GFLOPS. This is not a close contest on paper, and the benchmark results confirm it.

The GT 635M does have one area where it claims a nominal edge: memory capacity. It carries 2 GB of DDR3 memory, whereas the GTX 460M has 1536 MB (1.5 GB) of GDDR5. However, having more memory does not translate into a win in the recorded workload. The GTX 460M's memory subsystem is fundamentally faster: its 192-bit bus and GDDR5 type produce 60.00 GB/s of bandwidth, versus the GT 635M's 128-bit bus and DDR3 type at 28.80 GB/s. The GTX 460M more than doubles the memory bandwidth of the GT 635M, which likely explains why it dominates in the OpenCL compute test despite carrying less VRAM.

The percentile rankings reinforce this hierarchy. The GTX 460M sits at the 25th percentile of all GPUs in the database, while the GT 635M sits at the 22nd percentile. That is a narrow gap in overall position, but the performance delta within the tested workload is substantial. The GTX 460M wins 1 out of 1 head-to-head benchmarks; the GT 635M wins none. If a user's priority is compute throughput, the GTX 460M is the clear choice. If a user prioritizes VRAM capacity for holding larger textures, the GT 635M at least offers more storage, though the slower memory type may bottleneck that advantage.

Architecture Differences

Both GPUs share the same fundamental Fermi architecture from NVIDIA, and both are manufactured by TSMC on the same 40 nm process node. The similarities end there. The GTX 460M uses the GF106 chip, while the GT 635M uses the GF108 chip. These are different silicon dies with different transistor counts and physical sizes. The GF106 packs 1,170 million transistors onto a 238 mm² die, giving a transistor density of 4.9 million transistors per square millimeter. The GF108 is a smaller, cut-down chip with 585 million transistors on a 116 mm² die, which works out to 5.0 million transistors per square millimeter. The GTX 460M's die is more than twice the size of the GT 635M's, and it has exactly twice the transistor count.

The internal execution resources differ by the same 2:1 ratio. The GTX 460M has 192 shading units, 32 texture mapping units (TMUs), and 24 raster operation units (ROPs). The GT 635M has 96 shading units, 16 TMUs, and only 4 ROPs. That ROP count is particularly telling. The GTX 460M has six times the pixel output capability of the GT 635M, which shows up in the pixel rate figures: 5.400 GPixel/s versus 1.900 GPixel/s. Similarly, the texture rate is 21.60 GTexel/s for the GTX 460M versus 7.600 GTexel/s for the GT 635M. The GTX 460M has nearly three times the texture fill rate.

Memory architecture is another major divide. The GTX 460M uses GDDR5 memory with a 192-bit bus, while the GT 635M uses DDR3 with a 128-bit bus. The memory clock is recorded at 625 MHz (2.5 Gbps effective) for the GTX 460M and 900 MHz (1800 Mbps effective) for the GT 635M. Despite the lower clock, the GTX 460M's wider bus and faster memory type deliver 60.00 GB/s versus 28.80 GB/s. The GT 635M's memory is clocked higher, but the DDR3 type and narrower bus limit its throughput.

Power draw and physical form factor also differ. The GTX 460M has a TDP of 50 W and uses an MXM Module slot width, meaning it is a replaceable mobile graphics card. The GT 635M has a TDP of 35 W and is listed as an IGP (integrated graphics processor), though it still uses a PCIe 2.0 x16 bus interface like the GTX 460M. Neither requires external power connectors, and both have portable-device-dependent display outputs.

The generational placement is curious. The GTX 460M belongs to the GeForce 400M series, released in September 2010, while the GT 635M belongs to the GeForce 600M series, released in March 2012. The newer part is slower in compute, which suggests it was positioned as a mainstream or entry-level option, while the older part was a higher-tier mobile GPU. Both are now end-of-life products, with the GTX 460M succeeding the GeForce 300M and preceding the GeForce 500M, while the GT 635M succeeds the GeForce 500M and precedes the GeForce 700M.

Head-to-Head Benchmarks

The database contains one head-to-head benchmark between these two GPUs: the Geekbench OpenCL test. The GTX 460M scores 4282, and the GT 635M scores 3740. The GTX 460M wins by a delta of 14.5%. To put that in perspective, the GTX 460M's nearest rivals in the database include the AMD FirePro W2100 at 4295 (only 0.3% ahead), the AMD Radeon Vega 3 at 4268 (0.3% behind), the NVIDIA Quadro K3000M at 4241 (1% behind), and the NVIDIA GeForce RTX 4070 GDDR6 at 4335 (1.2% ahead). The GTX 460M is essentially trading blows with modern entry-level integrated graphics and workstation cards, which is remarkable for a 2010 mobile part.

The GT 635M's nearest rivals are a different class of hardware. The NVIDIA Quadro 3000M scores 3718 (0.6% behind), the NVIDIA GeForce GT 740M scores 3717 (0.6% behind), the NVIDIA GeForce 825M scores 3694 (1.2% behind), and the Intel UHD Graphics 710 scores 3792 (1.4% ahead). The GT 635M sits in a cluster of low-end mobile and integrated GPUs, all within a couple of percentage points of each other. Its 3740 score is respectable for its class, but it is nowhere near the GTX 460M's level.

The biggest single gap in the head-to-head data is the 14.5% delta between the two NVIDIA parts. That is a substantial margin for a single benchmark, and it aligns with the raw specification differences. The GTX 460M has double the shading units, double the TMUs, six times the ROPs, and more than double the memory bandwidth. The FP32 compute difference is even starker: 518.4 GFLOPS versus 182.4 GFLOPS, which is a 184% advantage for the GTX 460M. The benchmark delta of 14.5% actually understates the theoretical compute gap, likely because OpenCL workloads do not fully stress all execution resources equally, or because the GT 635M's higher memory clock (900 MHz vs 625 MHz) helps it close some distance.

Looking at the rival deltas, the GTX 460M's 14.5% win over the GT 635M is larger than any gap between either card and its nearest rivals. The GTX 460M's closest rival is only 1.2% away, and the GT 635M's closest rival is only 1.4% away. This means the two cards occupy different performance tiers entirely, with no overlap in their competitive clusters. The GTX 460M competes with chips like the Quadro K3000M and Vega 3, while the GT 635M competes with the GT 740M and UHD 710. The head-to-head result simply confirms that these are not peers.

FAQ

Q: Which GPU is faster in the Geekbench OpenCL benchmark?

A: The NVIDIA GeForce GTX 460M is faster, scoring 4282 compared to the GT 635M's 3740, a 14.5% difference.

Q: Do both GPUs use the same architecture?

A: Yes, both use NVIDIA's Fermi architecture, but they use different chips: the GTX 460M uses GF106, and the GT 635M uses GF108.

Q: How much memory does each GPU have, and what type?

A: The GTX 460M has 1536 MB of GDDR5 memory, while the GT 635M has 2 GB of DDR3 memory.

Q: Which GPU has higher memory bandwidth?

A: The GTX 460M has 60.00 GB/s of bandwidth, more than double the GT 635M's 28.80 GB/s.

Q: What is the transistor count for each chip?

A: The GTX 460M's GF106 has 1,170 million transistors, while the GT 635M's GF108 has 585 million transistors.

Q: Are these GPUs still in production?

A: No, both are listed as end-of-life products. The GTX 460M released in September 2010, and the GT 635M released in March 2012.

Specification Differences

| Specification | NVIDIA GeForce GTX 460M | NVIDIA GeForce GT 635M |

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

| Chip | GF106 | GF108 |

| Generation | GeForce 400M | GeForce 600M |

| Transistors | 1,170 million | 585 million |

| Die Size | 238 mm² | 116 mm² |

| Transistor Density | 4.9M / mm² | 5.0M / mm² |

| Memory Size | 1536 MB | 2 GB |

| Memory Type | GDDR5 | DDR3 |

| Memory Bus Width | 192 bit | 128 bit |

| Memory Bandwidth | 60.00 GB/s | 28.80 GB/s |

| Memory Clock | 625 MHz (2.5 Gbps effective) | 900 MHz (1800 Mbps effective) |

| Shading Units | 192 | 96 |

| TMUs | 32 | 16 |

| ROPs | 24 | 4 |

| Pixel Rate | 5.400 GPixel/s | 1.900 GPixel/s |

| Texture Rate | 21.60 GTexel/s | 7.600 GTexel/s |

| FP32 Performance | 518.4 GFLOPS | 182.4 GFLOPS |

| TDP | 50 W | 35 W |

| Slot Width | MXM Module | IGP |

| Release Date | September 2010 | March 2012 |

The two GPUs share the same process node (40 nm), foundry (TSMC), bus interface (PCIe 2.0 x16), API support (DirectX 12 (11_0), OpenGL 4.6), and lack of external power connectors. They differ on every other major specification.

The Verdict

The data presents a straightforward conclusion: the NVIDIA GeForce GTX 460M is the superior performer in compute workloads, despite being the older product. Its 4282 OpenCL score versus the GT 635M's 3740 represents a 14.5% advantage, and its raw specifications are dramatically better across the board. The GTX 460M has double the shading units, double the TMUs, six times the ROPs, and more than double the memory bandwidth. Its FP32 throughput of 518.4 GFLOPS is nearly three times that of the GT 635M's 182.4 GFLOPS.

The GT 635M does have one meaningful advantage: memory capacity. Its 2 GB of VRAM exceeds the GTX 460M's 1536 MB, which could matter for workloads that require holding larger datasets in local memory. However, the DDR3 memory type and 128-bit bus severely limit the speed at which that data can be accessed. The GTX 460M's GDDR5 memory on a 192-bit bus is categorically faster, and the benchmark results reflect that.

The generation gap is worth noting. The GT 635M is from the newer GeForce 600M series and has a lower TDP (35 W versus 50 W), making it more power-efficient. It also uses an IGP form factor, which may be simpler to integrate into thin laptops. But efficiency and integration do not compensate for the compute deficit. The GTX 460M wins the only recorded benchmark, and its nearest rivals in the database (Quadro K3000M, Vega 3, FirePro W2100) are all in a higher performance tier than the GT 635M's rivals (GT 740M, Quadro 3000M, UHD 710).

For a user selecting between these two for a mobile workstation or gaming laptop, the choice depends on the workload. If the task is compute-heavy, such as OpenCL acceleration, the GTX 460M is the only rational pick. Its 14.5% benchmark lead is backed by a 184% advantage in theoretical FP32 throughput. If the task is purely about fitting larger textures or datasets into VRAM, the GT 635M's 2 GB capacity is the sole argument in its favor, but the slow DDR3 memory will likely negate that benefit in practice. The database records the GTX 460M as the winner in 1 out of 1 head-to-head benchmarks, and the GT 635M as the winner in none. That is the verdict.