Definition
The GeForce 600 series is a family of graphics processing units (GPUs) marketed by Nvidia Corporation for the consumer desktop and notebook markets. The series implements Nvidia’s Kepler microarchitecture and was introduced in 2012 as the successor to the GeForce 500 series.
Overview
The GeForce 600 series comprises a range of models targeting various performance and price segments, including entry‑level (e.g., GT 610, GT 620), mid‑range (GT 630, GT 640, GT 650, GT 660, GTX 660, GTX 660 Ti), high‑end (GTX 670, GTX 680) and enthusiast (GTX 690) products. All GPUs in the series are fabricated on a 28 nm silicon‑on‑insulator (SOI) process and support DirectX 11, OpenGL 4.2, and CUDA Compute Capability 3.0. Key technological features introduced with the Kepler architecture, such as GPU Boost (dynamic clock‑frequency scaling), Dynamic Parallelism, and Hyper-Q, are present across the lineup.
The series was positioned to improve performance‑per‑watt relative to the preceding generation, emphasizing energy efficiency for both desktop PCs and laptops. Production of 600‑series cards continued into the early years of the subsequent GeForce 700 series, and many models received driver updates supporting later APIs such as DirectX 12 and Vulkan.
Etymology/Origin
- GeForce: Nvidia’s long‑standing brand name for its consumer‑oriented GPU products, originally introduced in 1999. The term combines “geometry” (referring to the processing of geometric data in 3D graphics) and “force” to imply processing power.
- 600 series: Nvidia’s internal naming convention uses three‑digit series numbers to denote generational progression. The “600” designation follows the “500” series and precedes the “700” series, indicating its place in the chronological product line.
Characteristics
| Attribute | General Attributes (across the series) |
|---|---|
| Microarchitecture | Kepler |
| Process node | 28 nm (TSMC) |
| CUDA cores | Varies by model (e.g., 192 in GT 620, 960 in GTX 660, 1536 in GTX 680, 3072 in GTX 690) |
| Memory interface | 64‑bit to 384‑bit, depending on model |
| Maximum VRAM | Up to 4 GB GDDR5 (GTX 680/690) |
| API support | DirectX 11 (native), OpenGL 4.2, CUDA 3.0 |
| Key technologies | GPU Boost, Dynamic Parallelism, Hyper‑Q, Adaptive VSync, PureVideo HD (video decoding) |
| Power consumption | Ranges from ~30 W (GT 610) to ~300 W (GTX 690) TDP |
| Form factor | Available in PCI‑Express 3.0 x16 desktop cards and MXM/PCIe mobile variants |
Related Topics
- Nvidia Corporation – The developer and manufacturer of the GeForce 600 series.
- Kepler microarchitecture – The GPU architecture underlying the series.
- GeForce 500 series – Predecessor generation based on the Fermi architecture.
- GeForce 700 series – Successor generation that extended Kepler and introduced the Maxwell architecture.
- CUDA – Nvidia’s parallel computing platform, supported by all 600‑series GPUs.
- AMD Radeon HD 7000/8000 series – Contemporary competing GPU product lines from AMD.
- DirectX 11 – Graphics API standard that the 600 series was designed to fully support.