AMD Radeon R5 Graphics vs NVIDIA GeForce GT 745M Comparison

# NVIDIA GeForce GT 745M vs AMD Radeon R5 Graphics

The NVIDIA GeForce GT 745M and AMD Radeon R5 Graphics represent two different approaches to mobile graphics from the same era: a discrete GPU from NVIDIA's Kepler family versus an integrated GPU from AMD's GCN 2.0 lineup. Both cards sit at the 23rd percentile of all GPUs, meaning they occupy nearly identical performance tiers, yet their underlying designs and benchmark results reveal clear strengths in different workloads. The GT 745M averages 3953 points across all benchmark tests, while the Radeon R5 Graphics averages 3883 points, a gap of roughly 1.8% that stays within the margin of error for most synthetic workloads.

FAQ

Q: Which GPU has the higher average benchmark score?

A: The NVIDIA GeForce GT 745M leads with an average benchmark score of 3953, compared to 3883 for the AMD Radeon R5 Graphics. This puts the GT 745M slightly ahead, though both occupy the same 23rd percentile of all GPUs.

Q: How do the two GPUs compare in OpenCL performance?

A: The AMD Radeon R5 Graphics is the clear winner in Geekbench OpenCL, scoring 5183 versus 3580 for the GT 745M. This represents a 30.9% advantage for the AMD part, a substantial margin that favors AMD in compute-oriented workloads.

Q: Which GPU performs better in Vulkan benchmarks?

A: The NVIDIA GeForce GT 745M dominates Vulkan testing, scoring 5502 against 2582 for the Radeon R5 Graphics. The delta is 113.1% in NVIDIA's favor, making this the largest performance gap between the two cards in any benchmark.

Q: What are the thermal design power ratings for each GPU?

A: The GT 745M has a TDP of 45 W, while the Radeon R5 Graphics is rated at 15 W. The AMD part consumes significantly less power, reflecting its integrated design within a Kaveri APU.

Q: Do both GPUs support DirectX 12?

A: Yes, but at different feature levels. The GT 745M supports DirectX 12 (11_0), meaning it can run DX12 titles but at the 11_0 feature set. The Radeon R5 Graphics supports DirectX 12 (12_0), offering the full feature level.

Q: How does transistor count compare between the two chips?

A: The AMD Radeon R5 Graphics, built on the Spectre SL chip, contains 2,410 million transistors on a 245 mm² die. The NVIDIA GT 745M uses the GK107 chip with 1,270 million transistors on a 118 mm² die.

Architecture Differences

The architectural split between these two GPUs is fundamental. NVIDIA's GT 745M is built on the Kepler architecture using the GK107 chip, manufactured by TSMC on a 28 nm process. Kepler was designed for efficiency and scaling, with a transistor density of 10.8 million transistors per square millimeter. AMD's Radeon R5 Graphics uses the GCN 2.0 architecture with the Spectre SL chip, fabricated by GlobalFoundries on the same 28 nm node, but with a lower transistor density of 9.8M per mm². The AMD chip is substantially larger at 245 mm² versus 118 mm² for NVIDIA, reflecting the integrated nature of the Kaveri APU design that includes CPU cores alongside the GPU.

The memory subsystems diverge sharply. The GT 745M has 2 GB of dedicated GDDR5 memory on a 128-bit bus, delivering 64.00 GB/s of bandwidth. The Radeon R5 Graphics uses System Shared memory, with bandwidth described as System Dependent — meaning performance scales with the host system's DDR3 memory configuration. This is a critical architectural difference: NVIDIA's discrete memory provides predictable bandwidth, while AMD's shared memory approach trades latency and bandwidth for cost and power savings.

Shader resources also differ. The GT 745M packs 384 shading units, 32 texture mapping units, and 16 ROPs. The Radeon R5 Graphics has 256 shading units, 16 TMUs, and only 4 ROPs. The ROP deficit is particularly striking — the NVIDIA part has four times the rasterization units. Pixel rate reflects this: 4.392 GPixel/s for the GT 745M versus 3.032 GPixel/s for the AMD. Texture rate follows the same pattern at 17.57 GTexel/s versus 12.13 GTexel/s. FP32 compute is closer, with 421.6 GFLOPS for NVIDIA and 388.1 GFLOPS for AMD, a difference of about 8.6%.

API support shows one notable divergence. Both support OpenGL 4.6 and Vulkan (1.2.175 for NVIDIA, 1.2.170 for AMD). However, DirectX support differs: the GT 745M is rated for DirectX 12 (11_0), while the Radeon R5 Graphics supports DirectX 12 (12_0). For modern games that use DX12 features, the AMD part has an architectural advantage, though the practical impact depends on the workload.

Head-to-Head Benchmarks

The benchmark data reveals two completely opposite performance profiles. In Geekbench OpenCL, the AMD Radeon R5 Graphics delivers a decisive victory with a score of 5183 versus 3580 for the GT 745M. The delta of -30.9% from NVIDIA's perspective means the AMD card is roughly 44.8% faster in raw OpenCL compute. This aligns with the GCN architecture's compute-oriented design, which was heavily optimized for asynchronous compute and general-purpose GPU workloads.

The Vulkan benchmark flips the script entirely. The GT 745M scores 5502, while the Radeon R5 Graphics manages only 2582. The 113.1% delta in NVIDIA's favor is extraordinary — the GT 745M more than doubles the AMD part's Vulkan performance. This suggests that Kepler's driver maturity and Vulkan implementation are far stronger than GCN 2.0's early Vulkan support, which was known to be rough in its initial iterations. The 2.13x advantage in Vulkan is the single largest performance differential between these two GPUs in any test.

Neither GPU has a Geekbench Metal score for the AMD side, so no comparison is possible there. The GT 745M records a Metal score of 2777, but without an equivalent AMD result, this data point stands alone. Across the two comparable benchmarks, each GPU wins one test — the AMD takes OpenCL, NVIDIA takes Vulkan — resulting in a 1-1 split in head-to-head wins.

Specification Differences

| Specification | NVIDIA GeForce GT 745M | AMD Radeon R5 Graphics |

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

| Chip | GK107 | Spectre SL |

| Architecture | Kepler | GCN 2.0 |

| Manufacturer | NVIDIA | AMD |

| Foundry | TSMC | GlobalFoundries |

| Process Node | 28 nm | 28 nm |

| Transistors | 1,270 million | 2,410 million |

| Die Size | 118 mm² | 245 mm² |

| Transistor Density | 10.8M / mm² | 9.8M / mm² |

| Memory Type | GDDR5 | System Shared |

| Memory Size | 2 GB | System Shared |

| Memory Bus Width | 128 bit | System Shared |

| Memory Bandwidth | 64.00 GB/s | System Dependent |

| Shading Units | 384 | 256 |

| TMUs | 32 | 16 |

| ROPs | 16 | 4 |

| Pixel Rate | 4.392 GPixel/s | 3.032 GPixel/s |

| Texture Rate | 17.57 GTexel/s | 12.13 GTexel/s |

| FP32 Performance | 421.6 GFLOPS | 388.1 GFLOPS |

| TDP | 45 W | 15 W |

| Bus Interface | PCIe 3.0 x16 | IGP |

| DirectX Support | 12 (11_0) | 12 (12_0) |

| Vulkan Version | 1.2.175 | 1.2.170 |

| Release Date | 2013-03-31 | 2014-09-16 |

The most consequential differences are memory configuration, ROP count, and TDP. The GT 745M's dedicated GDDR5 with 64 GB/s bandwidth provides consistent performance, while the Radeon's shared memory is entirely dependent on the host system. The 16 vs 4 ROP difference means NVIDIA can handle fill-rate-bound scenarios far better. The 45 W versus 15 W TDP gap reflects the discrete versus integrated design philosophy — the AMD part sips power inside an APU, while the NVIDIA card demands a dedicated power budget.

Where Each One Wins

NVIDIA GeForce GT 745M — Vulkan Gaming and Rasterization. The 113.1% Vulkan advantage is the headline story. For any application that leverages Vulkan — which includes many modern game engines and emulators — the GT 745M is decisively superior. The 4x ROP advantage and 64 GB/s dedicated bandwidth also make it the better choice for traditional rasterization workloads, pixel-heavy effects, and high-resolution textures. The 45 W TDP indicates this is a card designed for performance-first laptops where battery life is secondary. Its 16 TMUs versus AMD's 8 also give it a clear edge in texture-heavy scenes. The GT 745M's nearest rivals all sit within 0.6% of its average score, confirming it is well-positioned in its performance class.

AMD Radeon R5 Graphics — OpenCL Compute and Efficiency. The Radeon R5 Graphics wins the OpenCL benchmark by 30.9%, making it the stronger choice for compute workloads like video encoding, physics simulations, and machine learning inference that use OpenCL. Its 15 W TDP is one-third of the NVIDIA part's 45 W, making it dramatically more power-efficient per watt. For ultraportable laptops and 2-in-1 devices where battery life is paramount, the Radeon R5 Graphics is the logical choice. The full DirectX 12 (12_0) feature support also gives it an edge in DX12 titles that utilize the complete feature set. Its shared memory architecture means performance can scale with faster system memory, giving it flexibility in well-configured systems. The nearest rival to the Radeon is the GeForce MX110 at 1.3% higher score, indicating the AMD part remains competitive in its segment despite its integrated nature.

The Verdict. Neither GPU is a universal winner. The GT 745M excels in Vulkan and rasterization, while the Radeon R5 Graphics dominates OpenCL and power efficiency. The 23rd percentile placement for both confirms they are entry-level parts suited for casual gaming and productivity, not high-end workloads. Choose based on the applications you run: Vulkan-heavy gaming favors NVIDIA, while OpenCL compute and battery-conscious designs favor AMD.