Overclocking GeForce 9000 Series
Modified previous series with unified shader units, increased pixel fill rate. SLI mode supports up to four identical devices. Switching to a new technical process saved power consumption by up to 15%, lowered the temperature, and increased overclocking potential. HybridPower technology automatically switches the built-in and external video cores depending on the load, PureVideo HD will speed up dual-stream video decoding.
| Parameter Map Model | 9300 GS | 9400 GT | 9500 GT | 9600 GT | 9800 GT |
| Core Clock, MHz | 700 | 600 | 700 | 850 | 700 |
| Memory Clock, MHz | 516 | 400 | 530 | 500 | 550 |
Overclocking GeForce 100 Series
A line of budget graphics accelerators for the OEM segment. In fact, these are renamed GeForce 9 . There is no point in overclocking, raising the memory and core frequencies by 3-5% hardly affects performance.
Overclocking GeForce 200 Series
The line was marked by the appearance of CUDA v2 and PhysX technologies, a qualitative leap in the development of PureVideo. The graphics chip has increased in area, which has increased the efficiency of its cooling. The series demonstrated video chips adapted for general computing – Tesla. With relatively the same power consumption as the devices of the previous line, the 200 series video cards overtake them in tests and in real calculations. Made on two-layer boards for mounting two-chip video cards. The fan removes up to 50% more air than before. Most devices require 6 and 8 pin power.
| Parameter Map Model | GTX 295 | GTX 285 | GTX 280 | GTX 275 | GTX 260 | GTX 260 Core 216 | GTS 250 |
| Core Clock, MHz | 640 | 698 | 633 | 701 | 700 | 720 | 700 |
| Memory Clock, MHz | 1380 | 1300 | 1260 | 1352 | 1200 | 1120 | 1200 |
Overclocking GeForce 400 Series
40mm chips with DirectX 11 support on the Fermi architecture.
Features include:
- using MIMD for parallel processing of information;
- support for ECC – automatic memory error correction;
- transition to 64-bit architecture (graphic memory registers);
- compatibility with OpenGL, DirectCompute.
The difference in performance between the cheapest and most expensive devices reaches 20-25%. The lag is partially offset by good acceleration.
| Parameter Map Model | GTX 480 | GTX 470 | GTX 465 | GTX 460 | GTX 460 SE | GTS 450 |
| Core Clock, MHz | 825 | 650 | 750 | 750 | 851 | 830 |
| Memory Clock, MHz | 1650 | 1801 | 1800 | 1900 | 1990 | 1950 |
Overclocking GeForce 500 Series
Chips based on modified Fermi architecture with Quad SLI technology. Thanks to the optimization of power consumption at the transistor level, it has an attractive overclocking potential. In addition to transistors with low leakage current and fast switching with high leakage, a third type has appeared that combines the advantages of its predecessors. The solution increased energy efficiency, the temperature of the chips. Increased texture processing speed gives an increase of up to 15% in games.
Along with other improvements – up to 30%, compared to the 400th series. The acoustic comfort of the fan has been increased.
| Parameter Map Model | GTX 590 | GTX 580 | GTX 570 | GTX 560 | GTX 550 TI |
| Core Clock, MHz | 662 | 950 | 880 | 960 | 1025 |
| Memory Clock, MHz | 900 | 1100 | 1050 | 1165 | 1040 |
| Core Voltage, В | – | 1.037 | 1.1 | 1.15 |
Overclocking GeForce 600 Series
The line presented a 28-nm graphics core with GPU Boost dynamic frequency control technology – changing characteristics in accordance with the load without exceeding TDP. TXAA image enhancement is implemented in the chips – it cleans the scenes from flickering, smoothes out movements. Now you can connect four displays to one graphics accelerator – a clear focus on the gaming audience, and the NVENC block – hardware video encoding.
Drivers implemented dynamic ultra-high resolution to adapt content resolution to screens (usually lowering the resolution of a high-quality picture for screens with low resolution).
Important. Mobile chips are produced on the previous Fermi architecture.
| Parameter Map Model | GTX 670 | GTX 680 | GTX 690 | GTX 650 | GTX 650 Ti | GTX 660 | GTX 650 Ti Boost | GTX 660 Ti |
| Core Voltage, В | 1.175 | 1.075 | 1.125 | 1.010 | 1.037 | 1.1750 | 1.070 | 1.162 |
| Power Limit | 106% | 125% | 135% | – | – | 107.5% | – | 123% |
| Core Clock, MHz | 1061 | 1280 | 1172 | 1240 | 1176 | 1180 | + 400 | 1071 |
| Memory Clock, MHz | 1615 | 1800 | 1700 | 1552 | 1575 | 1652 | – | 1765 |
Overclocking GeForce 700 Series
Chip from Nvidia, the main area of which is occupied by SMX clusters and caches. Video cards have a markedly increased speed of double precision calculations and a slight increase in performance in single precision operations. For overclockers, the family is interesting for the Power Balancing function – synchronization of various power supplies (with high consumption, a hefty load on one line can become a bottleneck during overclocking). Elevated to 95 ° C, the operating temperature of the crystal plays an important role in overclocking.
| Parameter Map Model | GTX 770 | GTX 780 | GTX 780 Ti | GTX Titan | GTX Titan Black | GTX Titan Z | GTX 750 | GTX 750 Ti | GTX 760 | GT 710 | GT 720 | GT 730 | GT 740 |
| Core Voltage | – | 1.200 | 1.175 | +0.25 | 1.1 | 1.025 | – | – | 1.2 | 1.125 | + 0.075 | – | |
| Power Limit | – | 110% | 106% | 106% | – | 120% | – | – | 145% | – | – | – | |
| Core Clock, MHz | 1300 | 1070 | 1130 | 947 | 1225 | 1071 | 1295 | 1284 | 1130 | 1366 | 1199 | +1006 | 1110 |
| Memory Clock, MHz | 1850 | 1777 | 2025 | 1710 | 1850 | 2000 | 1424 | 1634 | 1800 | 1189 | 1371 | 1253 | 1651 |
