NVIDIA GeForce GTX 680
The NVIDIA GeForce GTX 680 is a graphics card released on March 22, 2012, built on the Kepler architecture using a 28 nm process. It has 1,536 shading units, 4 GB of GDDR5 memory on a 256-bit bus, 192.3 GB/s of memory bandwidth, a boost clock of 1.06 GHz and a 195 W board power rating. It scores 5,609 in the PassMark G3D benchmark. It supports DirectX 11 and OpenGL 4.5.
Depth: 42 recorded fields, 5 PassMark workloads
Specification data last verified October 1, 2026 against PassMark Videocard Benchmark and Wikipedia GPU list tables (wikigrid-2026-09). Coverage: 10 of 10 tracked specifications recorded.
Performance
The NVIDIA GeForce GTX 680 scores 5,609 in the PassMark G3D benchmark, higher than 82.2% of the 2,658 GPUs benchmarked (PassMark rank #488), measured across 7,482 submitted results. Its PassMark G2D score is 531.
PassMark workload tests
In the PassMark workload tests the NVIDIA GeForce GTX 680 averages DirectX 9 89 fps, DirectX 10 23 fps, DirectX 11 49 fps, DirectX 12 21 fps and GPU Compute 2,638 ops/s, across 7,482 submitted results. Frames per second are PassMark test-scene rates, not game frame rates.
As stated by PassMark
PassMark's own listing for the NVIDIA GeForce GTX 680 states the PCIe 3.0 x16 interface, up to 4 GB of memory, a core clock of 1006 MHz, DirectX 11 and OpenGL 4.5, first benchmarked on 2012-08-17. These are PassMark's figures, shown separately from the specification tables above.
Not to be confused with
This page covers the NVIDIA GeForce GTX 680 specifically. It is a different product from NVIDIA GeForce GTX 680M, NVIDIA GeForce GTX 680M KY_Bullet Edition, NVIDIA GeForce GTX 680MX, NVIDIA GeForce GTX 680 Mac Edition, which have their own specifications. Check the exact model name before comparing figures.
Graphics processor
The NVIDIA GeForce GTX 680 is based on the Kepler architecture, uses the GK104-400-A2 graphics processor, is manufactured on a 28 nm process, contains 3,540 million transistors and has a die size of 294 mm².
Shader configuration
The NVIDIA GeForce GTX 680 has 1,536 shading units, 128 texture mapping units and 32 render output units.
The NVIDIA GeForce GTX 680 does not use XMX engines; those are specific to Intel Arc graphics. Infinity Cache is an AMD architectural feature and does not apply to the NVIDIA GeForce GTX 680. Feex has no recorded ray tracing core count for the NVIDIA GeForce GTX 680.
Clock speeds
The NVIDIA GeForce GTX 680 runs at a base clock of 1.01 GHz and boosts to 1.06 GHz.
Memory
The NVIDIA GeForce GTX 680 is equipped with 4 GB of GDDR5 memory, uses a 256-bit memory bus and delivers 192.3 GB/s of memory bandwidth.
Board and power
The NVIDIA GeForce GTX 680 has a board power rating of 195 W and it connects over PCIe 3.0 x16.
Market
The NVIDIA GeForce GTX 680 was released on March 22, 2012 and launched at a price of $499.
Benchmarks
Graphics and compute APIs
The NVIDIA GeForce GTX 680 supports DirectX 11 and OpenGL 4.5.
Feex has no recorded CUDA compute capability for the NVIDIA GeForce GTX 680. Feex has no recorded Vulkan version for the NVIDIA GeForce GTX 680. Feex has no recorded OpenCL version for the NVIDIA GeForce GTX 680.
Frequently asked questions
How much VRAM does the NVIDIA GeForce GTX 680 have?
The NVIDIA GeForce GTX 680 has 4 GB of GDDR5 memory on a 256-bit bus. Its memory bandwidth is 192.3 GB/s.
How much power does the NVIDIA GeForce GTX 680 use and what power supply does it need?
The NVIDIA GeForce GTX 680 has a board power rating of 195 W.
How fast is the NVIDIA GeForce GTX 680?
The NVIDIA GeForce GTX 680 scores 5,609 in the PassMark G3D benchmark, higher than 82.2% of the 2,658 GPUs benchmarked. Its PassMark rank is #488. In the PassMark DirectX 12 test it averages 21 fps.
What bus interface does the NVIDIA GeForce GTX 680 use?
The NVIDIA GeForce GTX 680 connects over PCIe 3.0 x16.
When was the NVIDIA GeForce GTX 680 released and how much did it cost at launch?
The NVIDIA GeForce GTX 680 was released on March 22, 2012 and launched at a price of $499.
Corrections to this record
After an audit, 1 empty field was filled from sources. Each change is listed with its old value, new value, rule and sources.
1 filled value
- Architecture: Kepler (was Not recorded). Rule: Architecture from chip codename. chip GK104 -> Kepler. Sources: Wikipedia: List of Nvidia graphics processing units (revision 1377107724); Wikipedia: List of Nvidia desktop graphics processing units (revision 1375708460).
The audit log is public: all changes for this GPU, the corrections overview and the method.
Similar performance
Compare the NVIDIA GeForce GTX 680
Side-by-side specification comparisons against cards in the same performance band.