欢迎访问ic37.com |
会员登录 免费注册
发布采购

319537-003US 参数 Datasheet PDF下载

319537-003US图片预览
型号: 319537-003US
PDF下载: 下载PDF文件 查看货源
内容描述: 英特尔系统控制器中心 [Intel System Controller Hub]
分类和应用: 控制器
文件页数/大小: 450 页 / 2593 K
品牌: INTEL [ INTEL ]
 浏览型号319537-003US的Datasheet PDF文件第91页浏览型号319537-003US的Datasheet PDF文件第92页浏览型号319537-003US的Datasheet PDF文件第93页浏览型号319537-003US的Datasheet PDF文件第94页浏览型号319537-003US的Datasheet PDF文件第96页浏览型号319537-003US的Datasheet PDF文件第97页浏览型号319537-003US的Datasheet PDF文件第98页浏览型号319537-003US的Datasheet PDF文件第99页  
Graphics, Video, and Display (D2:F0)  
9 Graphics, Video, and Display  
(D2:F0)  
9.1  
Graphics Overview  
9.1.1  
3-D Core Key Features  
Two pipe scaleable unified shader implementation.  
— 3-D Peak Performance  
— Fill Rate: 2 Pixels per clock  
— Vertex Rate: One Triangle 15 clocks (Transform Only)  
— Vertex/Triangle Ratio average = 1 vtx/tri, peak 0.5 vtx/tri  
Texture max size = 2048 x 2048  
• Programmable 4x multi-sampling anti-aliasing (MSAA)  
— Rotated grid  
— ISP performance related to AA mode, TSP performance unaffected by AA mode  
• Optimized memory efficiency using multi-level cache architecture  
9.1.2  
Shading Engine Key Features  
The unified pixel/vertex shader engine supports a broad range of instructions.  
• Unified programming model  
— Multi-threaded with four concurrently running threads  
— Zero-cost swapping in/out of threads  
— Cached program execution model – unlimited program size  
— Dedicated pixel processing instructions  
— Dedicated vertex processing instructions  
— 2048 32-bit registers  
• SIMD pipeline supporting operations in:  
— 32-Bit IEEE Float  
— 2-way, 16-bit fixed point  
— 4-way, 8-bit integer  
— 32-bit, bit-wise (logical only)  
• Static and Dynamic flow control  
— Subroutine calls  
— Loops  
— Conditional branches  
— Zero-cost instruction predication  
• Procedural Geometry  
— Allows generation of more primitives on output compared with input data  
— Effective geometry compression  
— High order surface support  
• External data access  
— Permits reads from main memory by cache (can be bypassed)  
— Permits writes to main memory  
— Data fence facility provided  
— Dependent texture reads  
Datasheet  
95