GeekRadar / The comparison desk
COMPARE.
GPU
EVGA

EVGA 8800 Ultra KO

Graphics processor

Memory bus
384 bit*
Rated memory bandwidth
108 GB/s*

No matched benchmark results are available for this entry yet.

Processing & platform

Architecture
Tesla*
Process node
90 nm*
GPU cores
128*
Graphics interface
PCIe 1.0 x16*
Video memory
768 MB*
Reported chip die area
484 mm²*
Transistors
681 million*
Graphics processor
G80*
Texture mapping units
32*
Raster output units
24*
Streaming multiprocessors
16*
Chip foundry
TSMC*
Card or module format
Dual-slot*
Declared GPU TDP
171 W*
Suggested system PSU
450 W*
External power connectors
2x 6-pin*
DirectX / feature level
11.1 (10_0)*
OpenGL
3.3*
OpenCL
1.1 (1.0)*
CUDA compute capability
1.0*
Shader model
4.0*
GPU L2 cache
96 KB*

Memory & storage

Memory options
768*
Video memory type
GDDR3*
Memory bus
384 bit*
Rated memory bandwidth
108 GB/s*

Power specifications

Thermal design power
171 W*

System configuration

Detailed specifications

Specifications 1 · 24 fields
proc transistors
681 million*
proc die size
484* mm²
cores
128*
sm count
16*
memory clock
1,125*
ram
768*
name
EVGA 8800 Ultra KO*
ram type
GDDR3*
bus width
384*
bus interface
PCIe 1.0 x16*
ram bandwidth
108* GB/s
fp32 performance
0.4147*
tdp
171* W
suggested psu
450* W
slot width
Dual-slot*
power connectors
2x 6-pin*
api cuda
1*
is retail board
Yes*
gpi value
1.88*
manufacturer
EVGA*
arch name
Tesla*
card release date
2007-05-10*
proc foundry
TSMC*
proc process size
90* nm
Specifications 2 · 34 fields
CUDA compute capability
1.0*
DirectX / feature level
11.1 (10_0)*
OpenCL
1.1 (1.0)*
OpenGL
3.3*
Shader model
4.0*
Architecture
Tesla*
Graphics interface
PCIe 1.0 x16*
Memory bus
384* bit
Reported shader processors
128*
Reported theoretical FP32 peak
0.4147* TFLOPS
Scope and conditions
Arithmetic density
Not stated*
Declared GPU L2 cache
96 KB*
Reported display outputs
2x DVI, 1x S-Video*
Theoretical pixel fill rate
15.31 GPixel/s*
External power connectors
2x 6-pin*
Reported chip die area
484 mm²*
Chip foundry
TSMC*
Process node declaration
90 nm*
Transistors
681 million*
Graphics processor
G80*
Reported video memory
768* MB
Rated memory bandwidth
108.0 GB/s*
Video memory type
GDDR3*
Raster output units
24*
Reported shading units
128*
Card or module format
Dual-slot*
Streaming multiprocessors
16*
Suggested system PSU
450 W*
2:4 tensor sparsity
No*
Declared GPU TDP
171 W*
Theoretical texture fill rate
40.83 GTexel/s*
Texture mapping units
32*
GPU cores
128*
Video memory
768* MB
Video memory type
GDDR3*

* Provisional data