New INFO: R600 is based on 65nm Process!
ATI'S R600 GPU – NEW INFO – November 15, 2006
ATI AMD's R600 board is a monster
Part 2 Most complex PCB and the heaviest 3D card ever
- R600 to feature 700 to 800MHz clock for the GPU
- Complex 12-layer PCB monster - 512-bit memory controller - Each memory chip has its own 32-bit wide physical connection to the chip's bi-directional bus of 1024-Bit RingBus memory interface.- Memory bandwidth 115GB/s to 140.1GB/s (GDDR4 at 1.1GHz - 2.2GHz in DDR) (8800GTX only =’s 86GB/s)
- The R600 board is FAT. The PCB will be shorter than 8800GTX's in every variant & can be compared to X1950XT and 7900GTX.
- R600 will be the heaviest board that will hit the market in history.
- Monstrous all copper cooler, longer-than-the-PCB quad-heat pipe.
- The cooler actually enhances the structural integrity of the PCB, so you should be aware that R600 will bring some interesting things to the table.
- Why is it important that a cooler is so big? Well, it needs to dissipate heat from practically every element of the board: GPU chip, memory chips and the power regulation unit.
- Two versions of the board: #1 called Pele w/ 1GB to 2GB of GDDR4 memory, and UFO w/ 1GB to 2GB of GDDR3 memory.
(One version ONLY - The GDDR4 one. No More GDDR3)- Newer digital Rage Theatre 200 to replace the aging original Rage Theatre chip.
- R600 will offer video-in and video-out functions, VIVO, HD in and out connectors (Built in All-in-Wonder???)- R600 = 80nm @ 200 to 220 Watts – should consume less power then Geforce 8800GTX.
- R600 features new dual-bridge connector for Crossfire capability.
- Expect this baby to be in stores during Q1'07, or around January 30th.
- ATI's R600 features 64 Shader 4-way SIMD units. This is a very different and complex approach compared to Nvidia's relatively simple scalar Shader units.
- Since R600 SIMD Shader can calculate the result of four scalar units, it yields with scalar performance of 256 units
- R600 should feature at least five times more eDRAM than Xbox 360 has.
- R600 will do 16 pixels out in every clock, regardless of how complex the pixel might be.
- Twenty-four ROPs at 575MHz equals 13.8 billion pixels per clock, while 16 ROPs at 750MHz will end up at 12.0 billion pixels. At the same time, expect ATI to far better in more complex Shader-intensive applications.
New INFO on ATI's R600. I looked really nice but now it turned into a MONSTER Just look at the BOLD!!!!!!!!!!!
65nm (Advanced Process)
64 Shader pipelines (Vec4+Scalar) = yields with scalar performance of 256 units
720+ million transistor
32+ TMU's
32+ ROPs
128 Shader Operations per Cycle
800 MHz to 1,200 MHz Core
102.4 billion shader ops/sec
512GFLOPs for the shaders
2 Billion triangles/sec
25.6 Gpixels/Gtexels/sec
512-bit 1GB to 2GB 2.0GHz GDDR4 Memory
140+ GB/sec Bandwidth (at 2.0 GHz)
1024-Bit Ring Bus
WGF2.0 Unified Shader
etc.
LINK: - Posted today OCT 13, 2006
http://www.theinquirer.net/default.aspx?article=35062 http://www.theinquirer.net/default.aspx?article=38300
All I need to say is that is one hell of a change. Or maybe that is what they had in mind from the begining? Because for them to be able to get a 512-Bit memory interface to work, they must have been working on that puppy for a long wile.
It looks to me that it will be well worth the wait IMO
Where is Ian? He got to see this.