XFX 8800 GTX Extreme

PVT80FSHE9

Graphics Processor

G80

Cores

128

TMUs

ROPs

Memory Size

768 MB

Memory Type

GDDR3

Bus Width

384 bit

GPU

Graphics Processor

GPU Name: G80

GPU Variant: G80-300-A2

Architecture: Tesla

Foundry: TSMC

Process Size: 90 nm

Transistors: 681 million

Density: 1.4M / mm²

Die Size: 484 mm²

Chip Package: BGA-1449

Graphics Card

Release Date: ~~Nov 8th, 2006~~
Jan 9th, 2007

Generation: GeForce 8
(8800)

Predecessor: GeForce 7 PCIe

Successor: GeForce 9

Production: End-of-life

Bus Interface: PCIe 1.0 x16

Clock Speeds

GPU Clock: ~~576 MHz~~
600 MHz (+4%)

Shader Clock: 1350 MHz

Memory Clock: ~~900 MHz~~
950 MHz (+6%)
~~1800 Mbps effective~~
1900 Mbps effective

Memory

Memory Size: 768 MB

Memory Type: GDDR3

Memory Bus: 384 bit

Bandwidth: ~~86.40 GB/s~~
91.20 GB/s

Render Config

Shading Units: 128

TMUs: 32

ROPs: 24

SM Count: 16

L2 Cache: 96 KB

Theoretical Performance

Pixel Rate: ~~13.82 GPixel/s~~
14.40 GPixel/s

Texture Rate: ~~36.86 GTexel/s~~
38.40 GTexel/s

FP32 (float): 345.6 GFLOPS

Board Design

Slot Width: Dual-slot

TDP: 155 W

Suggested PSU: 450 W

Outputs: 2x DVI
1x S-Video

Power Connectors: 2x 6-pin

Board Number: P355 SKU 0

Graphics Features

DirectX: 11.1 (10_0)

OpenGL: 3.3

OpenCL: 1.1 (1.0)

Vulkan: N/A

CUDA: 1.0

Shader Model: 4.0

Card Notes

518.4 GFLOPs when Multiply Operation is available.

G80 GPU Notes

Each Streaming Multiprocessor(SM) in the chip architecture contains 8 SPs and 2 SFUs.

Each SP can fulfill up to two single precision operations MAD per clock.

Each SFU can fulfill up to four operations SF per clock (these units can also handle single-precision floating-point multiplications per clock).

The approximate ratio of operations MAD to operations SF is equal 2:1.

The theoretical SP + SFU performance in single-precision floating point operations
[FLOPSsp + sfu, GFLOPS] of the graphics card with shader count [n] and shader frequency [f, GHz], is estimated by the following formula: FLOPSsp+sfu; f × n × 3.

Alternative formula: FLOPSsp+sfu; f × m × ( 8 SPs * 2 (MAD) + 4 * 2 SFUs ). [m] - SM count.

SP - Shader Processor (Unified Shader, CUDA Core)
SFU - Special Function Unit
SM - Streaming Multiprocessor
MAD - ADD+MUL

Latest Drivers:
Windows 2000:
Quadro Release 261.19

Windows XP / Server 2003 x64:
GeForce Release 340.52
Quadro Release R319 U2 (321.01)
Tesla Release 258.96

Windows Vista / 7 / 8 / 8.1 / 10 32-bit:
GeForce Release 342.01
Quadro Release R340 U15 (342.01)
Tesla Release 258.96

Other retail boards based on this design (15)

Name	GPU Clock	Memory Clock	Other Changes
ASUS 8800 GTX AquaTank	630 MHz	1030 MHz	Triple-slot
EVGA 8800 GTX	576 MHz	900 MHz
EVGA 8800 GTX ACS³	576 MHz	900 MHz
EVGA 8800 GTX KO ACS³	626 MHz	1000 MHz
EVGA 8800 GTX Superclocked	621 MHz	1000 MHz
EVGA 8800 GTX Superclocked v2	621 MHz	1000 MHz
EVGA 8800 GTX v2	576 MHz	900 MHz
Foxconn 8800 GTX OC	650 MHz	1000 MHz
OCZ 8800 GTX	576 MHz	900 MHz
SPARKLE Calibre 8800 GTX	630 MHz	980 MHz
XFX 8800 GTX	576 MHz	900 MHz
XFX 8800 GTX Extreme	600 MHz	950 MHz
XFX 8800 GTX Premium	630 MHz	1030 MHz
XFX 8800 GTX XXX	630 MHz	1000 MHz
ZOTAC 8800 GTX AMP! Edition	630 MHz	1000 MHz