GeekRadar / The comparison desk
COMPARE.
GPUNVIDIA Riva TNT2 Ultra
Graphics processor Released
- Video memory
- 32 MB
- Memory bus
- 128 bit
- Rated memory bandwidth
- 2.93 GB/s
Benchmark results
- Reported score · version unreported · passmark_g3d_mark3*Platform unreported Compare results
Processing & platform
- Architecture
- Fahrenheit
- Manufacturing process
- TSMC 250 nm
- Process node
- 250 nm
- GPU base clock
- 150 MHz*
- Graphics interface
- AGP 4x*
- Video memory
- 32 MB*
- Reported chip die area
- 63 mm²*
- Transistors
- 15 million*
- GPU boost clock
- 150 MHz*
- Graphics processor
- NV5*
- Texture mapping units
- 2*
- Raster output units
- 2*
- Chip foundry
- TSMC*
- Card or module format
- Single-slot*
- Suggested system PSU
- 200 W*
- External power connectors
- None*
- DirectX / feature level
- 6.0*
- OpenGL
- 1.2*
GPU configurations
- 0.03125 GB / SDR / 128-bit / NV5
- Capacity
- 32 MB
- Rated memory bandwidth
- 2.93 GB/s
- Chip variant
- NV5
- Reference clock
- 150 MHz
- Memory bus
- 128 bit
- Technology
- SDR
Memory & storage
- Memory options
- 32*
- Video memory
- 32 MB
- Video memory type
- SDR
- Memory bus
- 128 bit
- Rated memory bandwidth
- 2.93 GB/s
Power specifications
- Thermal design power
- 15 W*
System configuration
Detailed specifications
Specifications 1 · 18 fields
- proc transistors
- 15 million*
- proc die size
- 63* mm²
- memory clock
- 183*
- ram
- 32*
- name
- NVIDIA Riva TNT2 Ultra*
- ram type
- SDR*
- bus width
- 128*
- bus interface
- AGP 4x*
- ram bandwidth
- 2.928* GB/s
- suggested psu
- 200* W
- slot width
- Single-slot*
- is retail board
- No*
- vendor
- nvd*
- manufacturer
- NVIDIA*
- arch name
- Fahrenheit*
- card release date
- 1999-03-15*
- proc foundry
- TSMC*
- proc process size
- 250* nm
Specifications 2 · 24 fields
- DirectX / feature level
- 6.0*
- OpenGL
- 1.2*
- Architecture
- Fahrenheit*
- Graphics interface
- AGP 4x*
- Memory bus
- 128* bit
- Reported display outputs
- 1x VGA, 2x S-Video*
- Theoretical pixel fill rate
- 300.0 MPixel/s*
- External power connectors
- None*
- Reported chip die area
- 63 mm²*
- Chip foundry
- TSMC*
- Process node declaration
- 250 nm*
- Transistors
- 15 million*
- Graphics processor
- NV5*
- Reported video memory
- 32* MB
- Rated memory bandwidth
- 2.928 GB/s*
- Video memory type
- SDR*
- Raster output units
- 2*
- Card or module format
- Single-slot*
- Suggested system PSU
- 200 W*
- 2:4 tensor sparsity
- No*
- Theoretical texture fill rate
- 300.0 MTexel/s*
- Texture mapping units
- 2*
- Video memory
- 32* MB
- Video memory type
- SDR*
Specifications 3 · 9 fields
- model
- Riva TNT2 Ultra*
Scope and conditions
- source_native_row_context
- model
- Riva TNT2 Ultra*
- core config
- 2:2:2*
Scope and conditions
- source_native_row_context
- model
- Riva TNT2 Ultra*
- manufacturer
- Nvidia*
Scope and conditions
- source_native_row_context
- model
- Riva TNT2 Ultra*
- launch date
- 1999-03-15 00:00:00.000000000*
Scope and conditions
- source_native_row_context
- model
- Riva TNT2 Ultra*
- code name
- NV5*
Scope and conditions
- source_native_row_context
- model
- Riva TNT2 Ultra*
- fab nm
- TSMC 250 nm* nm
Scope and conditions
- source_native_row_context
- model
- Riva TNT2 Ultra*
- die size mm2
- 63.0* mm²
Scope and conditions
- source_native_row_context
- model
- Riva TNT2 Ultra*
- bus interface
- AGP 4x*
Scope and conditions
- source_native_row_context
- model
- Riva TNT2 Ultra*
- core clock mhz
- 150.0* MHz
Scope and conditions
- source_native_row_context
- model
- Riva TNT2 Ultra*
Specifications 4 · 28 fields
- gpu name
- Nvidia RIVA TNT2 Ultra*
- sm count
- 1*
- tensor cores count
- 0*
- rt cores count
- 0*
- base clock mhz
- 150* MHz
- boost clock mhz
- 150* MHz
- memory size gb
- 0.032* GB
- memory type
- SDR*
- bus width bit
- 128*
- bandwidth gbs
- 2.9* GB/s
- series
- RIVA*
- fp32 tflops
- 0.0006* TFLOPS
- tensor tflops
- 0* TFLOPS
- tdp watts
- 15* W
- passmark g3d mark
- 3*
- time spy graphics
- 2*
- launch price usd
- 299*
- transistor density m per mm2
- 0.17* mm²
- fp32 gflops per dollar
- 0* GFLOPS
- fp32 gflops per watt
- 0.04* GFLOPS
- tensor tflops per watt
- 0* TFLOPS
- architecture
- Fahrenheit*
- chip code
- NV5*
- launch year
- 1999*
- process node nm
- 250* nm
- die size mm2
- 90* mm²
- transistor count millions
- 15*
- cuda cores
- 2*
0 selected Compare 0
* Provisional data