In this new all-in-one review we will
compare the new DX11 HD5xxx series cards from ATI, the predecessor DX10
the HD 4xxx series and the GTX275 from Nvidia
in addition to that you can find out a wide range of the most popular
graphic cards benchmarked, collected via popular websites reviewers , So
lets begin with the specifications of each card:
ATI Radeon HD5770
The ATI Radeon HD 5770 is expected to be
released on October 13th for $159. this GPU features 800 stream
processors, a 850MHz core and use 1GB GDDR5 memory at 4.8GHz, 1.36
teraFLOPS computing power, 128-bit memory bus, 76.8GB/s memory
bandwidth, 18W idle power consumption and 108W load power consumption.
GPU Specifications:
* 1.04 billion 40nm transistors
* TeraScale 2 Unified Processing Architecture
o 800 Stream Processing Units
o 40 Texture Units
o 64 Z/Stencil ROP Units
o 16 Color ROP Units
* GDDR5 memory interface
* PCI Express 2.1 x16 bus interface
* DirectX® 11 support
o Shader Model 5.0
o DirectCompute 11
o Programmable hardware tessellation unit
o Accelerated multi-threading
o HDR texture compression
o Order-independent transparency
* OpenGL 3.2 support1
* Image quality enhancement technology
o Up to 24x multi-sample and super-sample anti-aliasing modes
o Adaptive anti-aliasing
o 16x angle independent anisotropic texture filtering
o 128-bit floating point HDR rendering
* ATI Eyefinity multi-display technology2,3
o Three independent display controllers
+ Drive three displays
simultaneously with independent resolutions, refresh rates, color
controls, and video overlays
o Display grouping
+ Combine multiple displays to behave like a single large display
* ATI Stream acceleration technology
o OpenCL 1.0 compliant
o DirectCompute 11
o Accelerated video encoding, transcoding, and upscaling4,5
+ Native support for common video encoding instructions
* ATI CrossFireX™ multi-GPU technology6
o Dual GPU scaling
ATI Radeon HD5750
The ATI Radeon HD 5750 will be launched
on the same day, this model will be available in with 512MB GDDR5 for
$109 and with 1GB GDDR5 for $129. It has 720 stream processors, 700MHz
core, 4.6GHz memory, 1.008 teraFLOPS computing power, 128-bit memory
bus, 73.6GB/s memory bandwidth, 16W idle power consumption and 86W load
power consumption.
GPU Specifications:
* 1.04 billion 40nm transistors
* TeraScale 2 Unified Processing Architecture
o 720 Stream Processing Units
o 36 Texture Units
o 64 Z/Stencil ROP Units
o 16 Color ROP Units
* GDDR5 memory interface
* PCI Express 2.1 x16 bus interface
* DirectX® 11 support
o Shader Model 5.0
o DirectCompute 11
o Programmable hardware tessellation unit
o Accelerated multi-threading
o HDR texture compression
o Order-independent transparency
* OpenGL 3.2 support1
* Image quality enhancement technology
o Up to 24x multi-sample and super-sample anti-aliasing modes
o Adaptive anti-aliasing
o 16x angle independent anisotropic texture filtering
o 128-bit floating point HDR rendering
o Three independent display controllers
+ Drive three displays
simultaneously with independent resolutions, refresh rates, color
controls, and video overlays
o Display grouping
+ Combine multiple displays to behave like a single large display
* ATI Stream acceleration technology
o OpenCL 1.0 compliant
o DirectCompute 11
o Accelerated video encoding, transcoding, and upscaling4,5
* ATI CrossFireX™ multi-GPU technology6
o Dual GPU scaling
Radeon HD 4850
The ATI Radeon HD 4850 is a DirextX 10.1 based on the RV770PRO GPU which first announced at June 16 2008 and official release on June 25 2008, packing 800 stream processors interconnected with a GDDR3 video memory using a 256-bit bus, the HD 4850 use the new TeraScale graphics engine features up to 1 teraFLOPS with about one billion transistors inside
Radeon HD 4850 specifications:
* 956 million transistors on 55nm fabrication process
* PCI Express 2.0 x16 bus interface
* 256-bit GDDR3/4/5 memory interface
* Microsoft® DirectX® 10.1 support
o Shader Model 4.1
o 32-bit floating point texture filtering
o Indexed cube map arrays
o Independent blend modes per render target
o Pixel coverage sample masking
o Read/write multi-sample surfaces with shaders
o Gather4 texture fetching
* Unified Superscalar Shader Architecture
o 800 stream processing units
+ Dynamic load balancing and resource allocation for vertex, geometry, and pixel shaders
+ Common instruction set and texture unit access supported for all types of shaders
+ Dedicated branch execution units and texture address processors
o 128-bit floating point precision for all operations
o Command processor for reduced CPU overhead
o Shader instruction and constant caches
o Up to 160 texture fetches per clock cycle
o Up to 128 textures per pixel
o Fully associative multi-level texture cache design
o DXTC and 3Dc+ texture compression
o High resolution texture support (up to 8192 x 8192)
o Fully associative texture Z/stencil cache designs
o Double-sided hierarchical Z/stencil buffer
o Re-Z, and Z Range optimization
o Lossless Z & stencil compression (up to 128:1)
o Lossless color compression (up to 8:1)
o Up to 8 render targets (MRTs) with anti-aliasing
o Accelerated physics processing
* Dynamic Geometry Acceleration
o High performance vertex cache
o Programmable tessellation unit
o Accelerated geometry shader path for geometry amplification
o Memory read/write cache for improved stream output performance
* Anti-aliasing features
o Multi-sample anti-aliasing (2, 4, or 8 samples per pixel)
o Up to 24x Custom Filter Anti-Aliasing (CFAA) for improved quality
o Adaptive super-sampling and multi-sampling
o Gamma correct
o Super AA (ATI CrossFireX™ configurations only)
o All anti-aliasing features compatible with HDR renderin
* Texture filtering features
o 2x/4x/8x/16x high quality adaptive anisotropic filtering modes (up to 128 taps per pixel)
o 128-bit floating point HDR texture filtering
o sRGB filtering (gamma/degamma)
o Percentage Closer Filtering (PCF)
o Depth & stencil texture (DST) format support
o Shared exponent HDR (RGBE 9:9:9:5) texture format support
* OpenGL 2.1 support
* ATI Avivo™ HD Video and Display Platform6
GeForce GTX 275
The Nvidia GeForce GTX 275 belong to the the latest 10th generation of graphic cards from the green monster, and they here for one reason, attack the new ATI HD 4000 team, the GTX 275 use the GT200b GPU and support the DirectX 10.0 (Shader Model 4.0), first appearance on April 2, 2009 with a price tag set to $210.
GeForce GTX 275 specifications:
GPU Engine Specs:
CUDA Cores | 240 |
Graphics Clock (MHz) | 633 MHz |
Processor Clock (MHz) | 1404 MHz |
Texture Fill Rate (billion/sec) | 50. |
Memory Specs:
Memory Clock (MHz) | 1134 MHz |
Standard Memory Config | 896 MB GDDR3 |
Memory Interface Width | 448-bit |
Memory Bandwidth (GB/sec) | 127.0 |
Feature Support:
NVIDIA SLI®-ready* | 2-way/3-Way |
GeForce 3D Vision Ready | yes |
NVIDIA PureVideo® Technology* | HD |
NVIDIA PhysX™-ready | yes * |
NVIDIA CUDA™ Technology | yes |
Microsoft DirectX | 10 |
OpenGL | 3.0 |
Bus Support | PCI-E 2.0 |
Certified for Windows Vista | yes |
Benchmark results
All-in-one
All-in-one
Overclocking performance on this card
is impressive! some rumours over the net report that it can touch the
toes of the HD5870. 30% OC on stock voltage.
Our conclusion:
Based on all these reviews we've selected, the Radeon HD 5770 seems to be a great performing video card with its powerful GPU running @ 800Mhz and 800 unified shaders units seems to get a good performance/price result, and if you are seeking a powerful one go ahead to the Nvidia GTX 275, for those who already have a HD 4850 and want to upgrade to the HD 5770 we advise to move up to the next generation and you will be satisfied with this last one.
Our conclusion:
Based on all these reviews we've selected, the Radeon HD 5770 seems to be a great performing video card with its powerful GPU running @ 800Mhz and 800 unified shaders units seems to get a good performance/price result, and if you are seeking a powerful one go ahead to the Nvidia GTX 275, for those who already have a HD 4850 and want to upgrade to the HD 5770 we advise to move up to the next generation and you will be satisfied with this last one.
0 Comments:
Post a Comment