Geforce GTX 260-216 with 55 nanometer GT200b reviewed
With the 55 nanometer GPU and a cost efficient design for the remake of the Geforce GTX 260-216 Nvidia wants to win back market shares and conquer new ones. PCGH checks if the GT200b card has the capabilities for this mission.
GTX 260-216 with 55 Nanometers: Introduction
Almost no other card has been reissued as often as the Geforce GTX 260: While memory chips and interface (896 MiByte at 448 bit), clock speeds (576/1,242/999 MHz) and the number of ROPs (28) didn't change, the GPU had to endure several adjustments.
The first revision had 192 Shader ALUs and 64 TMUs and Nvidia had to add on because of the heavy competition represented by the Radeon HD 4870/1G. The remake was called GTX 260-216 and had 216 ALUs and 72 TMUs and thus was able to compete with the Radeon. For some weeks now the anxiously expected 55 nanometer version has leaked to the markets more or less silently. Besides adding the GT200b, Nvidia also changed details of the board (PCB) and the cooler in order to make production more cost efficient. The same applies for the GTX 285 that replaces the GTX 280.
GTX 260-216 with 55 Nanometers: Technical specifications
The performance of the 55 nanometer GTX equals the performance of the 65nm version. The GT200b still has 216 Shader ALUs, which are clocked to 1,242 MHz, and the core frequency is 576 MHz, while the 896 MiByte VRAM are linked via an 448 bit interface.
The number of layers of the PCB has been reduced from 14 to 10. Furthermore all the memory chips have been placed on the front side of the card and the layout of other components has been changed, too. Therefore many coolers designed for the 65 nanometer GTX 260 are not compatible to the new version anymore.
|Graphics chip/card||GTX 285||GTX 280||GTX 260-216||9800 GTX+||HD 4870||HD 4850|
|Ca. price (Euro)||330.-||290.-||220.-||140.-||180.-||120.-|
|Core frequency (MHz)||648||602||576||738||750||625|
|Shader ALU frequency (MHz)||1.476||1.296||1.242||1.836||750||625|
|Memory frequency (MHz)||1.242||1.107||999||1.100||1.800||1.000|
|MAD performance (GFLOPS)||708,5||622,1||536,5||470||1.200||1.000|
|Texture fillrate (MTex/s)||51.840||48.160||41.472||47.232||30.000||25.000|
|Memory bandwidth (MByte/s)||158.976||141.696||111.888||70.400||115.200||64.000|
|Multi Sampling, max.||8x||8x||8x||8x||8x||8x|
|Anti Aliasing, max.||16xQ||16xQ||16xQ||16xQ||24x CFAA||24x CFAA|
|Multi GPU||SLI, 3-W.-SLI||SLI, 3-W.-SLI||SLI||SLI, 3-W.-SLI||Crossfire X||Crossfire X|
|PCI-E power connector (6-Pin/8-Pin)||2 x/0 x||1 x/1 x**||2 x/0 x||2 x/0 x||2 x/0 x||1 x/0 x|
Grafikkarten-Rangliste 2016: 32 Radeon- und Geforce-GPUs im Benchmarkvergleich [Oktober]