NVIDIA GeForce 830M vs NVIDIA GeForce GTX 650 Comparison

The NVIDIA GeForce 830M and the NVIDIA GeForce GTX 650 are both end-of-life graphics processors from NVIDIA, but they target entirely different segments of the market. The 830M is a mobile IGP (integrated graphics processor) designed for portability and low power draw, while the GTX 650 is a desktop single-slot card built for gaming performance. The benchmark data shows a clear split: the GTX 650 dominates in raw compute and graphics workloads, while the 830M offers a level of performance in a power envelope that a desktop card cannot match. This analysis will break down where each chip wins based strictly on the provided data.

Where Each One Wins

The GTX 650 is the clear performance winner in every benchmark where both were tested. It wins the Geekbench OpenCL test with a score of 4545 against the 830M's 4324, a margin of 4.9%. The gap widens significantly in the Geekbench Vulkan test, where the GTX 650 scores 4524 against the 830M's 3590, a substantial 20.6% advantage. If the workload is compute-heavy or requires modern graphics APIs, the GTX 650 is the only choice between the two.

The 830M's primary victory is not in benchmark scores but in its physical and power profile. It is an IGP with a 33 W TDP, meaning it requires no power connectors and is designed to be integrated into a portable device. The GTX 650, by contrast, has a 65 W TDP and requires a 1x 6-pin power connector. The 830M also supports a more recent Vulkan version (1.4 vs 1.2.175), which could matter for compatibility with newer software. For a user building a low-power, compact system where a discrete GPU is not an option, the 830M is the only viable path, but it is a path to lower performance.

Architecture Differences

The two GPUs come from different architectural generations. The 830M uses the Maxwell architecture on the GM108 chip, while the GTX 650 uses the older Kepler architecture on the GK106 chip. Both are built on the same 28 nm process at TSMC, but the chip designs diverge significantly. The GM108 is a small chip with 1,020 million transistors on a 77 mm² die, resulting in a transistor density of 13.2 million transistors per mm². The GK106 is a much larger chip, packing 2,540 million transistors on a 221 mm² die, with a lower density of 11.5 million transistors per mm².

The GTX 650's larger chip translates to more execution resources. It has 384 shading units, 32 texture mapping units (TMUs), and 16 ROPs. The 830M is cut down to 256 shading units, 16 TMUs, and 8 ROPs. This gives the GTX 650 a theoretical texture rate of 33.86 GTexel/s and a pixel rate of 8.464 GPixel/s, compared to the 830M's 18.40 GTexel/s and 9.200 GPixel/s. Notably, the 830M has a higher pixel rate, but the GTX 650's much higher texture rate and shading unit count gives it the overall edge in fill-rate-bound workloads.

Memory architecture is another major divergence. The 830M uses 2 GB of DDR3 memory on a 64-bit bus, yielding a bandwidth of 14.40 GB/s. The GTX 650 uses 1024 MB of GDDR5 memory on a 128-bit bus, delivering a much higher 80.00 GB/s of bandwidth. The GTX 650 also has a higher memory clock at 1250 MHz (5 Gbps effective) versus the 830M's 900 MHz (1800 Mbps effective). This memory bandwidth difference is a key reason for the GTX 650's superior performance in Vulkan, which often stresses memory throughput.

Head-to-Head Benchmarks

The head-to-head data contains two benchmark results, and the GTX 650 wins both. In the Geekbench OpenCL test, the GTX 650 scores 4545 against the 830M's 4324. The delta is -4.9% from the perspective of the 830M, meaning the 830M is 4.9% slower. This is a modest win, suggesting that the 830M's Maxwell architecture is efficient enough to keep it competitive in general compute tasks despite having fewer cores.

The second test, Geekbench Vulkan, shows a much larger gap. The GTX 650 scores 4524, while the 830M manages only 3590. This is a -20.6% delta for the 830M. The GTX 650 is more than a fifth faster in this API. This is likely due to the GTX 650's 80.00 GB/s of memory bandwidth, which is over five times higher than the 830M's 14.40 GB/s. Vulkan workloads often involve heavy data movement, and the GTX 650's GDDR5 memory gives it a decisive advantage.

It is also worth noting the GTX 650's additional benchmark result in Geekbench Metal, where it scores 2400. The 830M has no Metal score in the data, so a direct comparison is not possible, but this shows the GTX 650 has broader API support for Apple's framework.

Specification Differences

The following specifications differ between the two GPUs, based on the FACT PACK:

• Chip: GM108 (830M) vs GK106 (GTX 650)
• Architecture: Maxwell (830M) vs Kepler (GTX 650)
• Transistors: 1,020 million (830M) vs 2,540 million (GTX 650)
• Die Size: 77 mm² (830M) vs 221 mm² (GTX 650)
• Transistor Density: 13.2M / mm² (830M) vs 11.5M / mm² (GTX 650)
• Base Clock: 1082 MHz (830M) vs null (GTX 650)
• Boost Clock: 1150 MHz (830M) vs null (GTX 650)
• Memory Clock: 900 MHz / 1800 Mbps effective (830M) vs 1250 MHz / 5 Gbps effective (GTX 650)
• Memory Size: 2 GB (830M) vs 1024 MB (GTX 650)
• Memory Type: DDR3 (830M) vs GDDR5 (GTX 650)
• Memory Bus Width: 64 bit (830M) vs 128 bit (GTX 650)
• Memory Bandwidth: 14.40 GB/s (830M) vs 80.00 GB/s (GTX 650)
• Shading Units: 256 (830M) vs 384 (GTX 650)
• TMUs: 16 (830M) vs 32 (GTX 650)
• ROPs: 8 (830M) vs 16 (GTX 650)
• Pixel Rate: 9.200 GPixel/s (830M) vs 8.464 GPixel/s (GTX 650)
• Texture Rate: 18.40 GTexel/s (830M) vs 33.86 GTexel/s (GTX 650)
• FP32: 588.8 GFLOPS (830M) vs 812.5 GFLOPS (GTX 650)
• TDP: 33 W (830M) vs 65 W (GTX 650)
• Slot Width: IGP (830M) vs Single-slot (GTX 650)
• Power Connectors: None (830M) vs 1x 6-pin (GTX 650)
• Suggested PSU: null (830M) vs 250 W (GTX 650)
• Bus Interface: PCIe 3.0 x8 (830M) vs PCIe 3.0 x16 (GTX 650)
• Display Outputs: Portable Device Dependent (830M) vs 1x DVI, 2x DisplayPort 1.2 (GTX 650)
• Vulkan Version: 1.4 (830M) vs 1.2.175 (GTX 650)
• Dimensions: null (830M) vs 147 mm / 5.8 inches length (GTX 650)
• Release Date: 2014-03-11 (830M) vs 2013-11-26 (GTX 650)

FAQ

Q: Which GPU is faster in Geekbench OpenCL?

A: The GTX 650 is faster, scoring 4545 versus the 830M's 4324, a 4.9% advantage.

Q: How much faster is the GTX 650 in Geekbench Vulkan?

A: The GTX 650 scores 4524, while the 830M scores 3590. This gives the GTX 650 a 20.6% lead over the 830M.

Q: Does the 830M have a higher pixel rate than the GTX 650?

A: Yes. The 830M has a pixel rate of 9.200 GPixel/s, whereas the GTX 650 has a pixel rate of 8.464 GPixel/s.

Q: What is the memory bandwidth difference?

A: The GTX 650 has 80.00 GB/s of bandwidth, which is significantly higher than the 830M's 14.40 GB/s.

Q: Which GPU has a lower power draw?

A: The 830M has a TDP of 33 W, which is lower than the GTX 650's 65 W. The 830M also requires no power connectors, while the GTX 650 needs a 1x 6-pin connector.

Q: Do both GPUs support DirectX 12?

A: Yes, both support DirectX 12 (11_0) and OpenGL 4.6. They differ in Vulkan support, with the 830M supporting version 1.4 and the GTX 650 supporting version 1.2.175.

The Verdict

The data presents a straightforward choice. The GTX 650 is the superior performer in every head-to-head benchmark. It is 4.9% faster in OpenCL and 20.6% faster in Vulkan. It also has a higher FP32 throughput (812.5 GFLOPS vs 588.8 GFLOPS), more shading units (384 vs 256), and more memory bandwidth (80.00 GB/s vs 14.40 GB/s). The GTX 650's percentile rank of 22% versus the 830M's 24% is slightly lower, but this is based on a different average benchmark score (3823 vs 3957) and should not overshadow the direct comparison where the GTX 650 wins both tests.

The 830M is not without merit. It is a low-power IGP with a 33 W TDP, no power connectors, and a smaller die size (77 mm² vs 221 mm²). It also supports a newer Vulkan version (1.4 vs 1.2.175). Its higher pixel rate of 9.200 GPixel/s is a notable advantage, but this does not translate into a win in the provided benchmarks. The 830M is the right choice for a portable, integrated system where power consumption is critical and a discrete card is not physically possible.

For anyone building a desktop or a system where a discrete card can be installed, the GTX 650 is the clear pick. Its performance lead is decisive, and its single-slot form factor and 250 W suggested PSU make it a practical addition to a budget build. The 830M is a capable low-power processor, but the GTX 650 is a faster GPU, plain and simple. Choose the GTX 650 for performance, or the 830M for integration and efficiency.