macdawg
Daemon Poster
- Messages
- 813
The R600 itself can consume up to 300 watts! There goes the power bill, not only will I wait for dx10 cards, I will wait until they get more effective per watt!
* The R600 is estimated to be the largest GPU ever made. Production based on an 80 nm process with ATI having 65 nm clearly in sights. [1]
* TSMC has only just begun 65 nm of simple designs.[2],
* The R600 might have over a 1000 MHz clock, making it the highest clocked GPU ever made.[3]
* The R600 will be the first graphics card faster than a X1800GTO able to do dongleless Crossfire (no master card). The X1800XT, X1900GT, X1900XT, and X1900XTX all require a slightly slower master card to run Crossfire. [4]
* 64 unified shader pipelines, 32 TMUs, and 32 ROPs.
* It shares a similar design of the Xbox 360's GPU "Xenos". However, the R600 will not have the 10 MB daughter embedded DRAM framebuffer of the "Xenos".
* The R600 will support the upcoming GDDR4 memory interface with 512 MB running higher than the X1950XTX's memory clock speeds of 1.0 GHz (2.0 GHz effective).
* The current target release date is December 2006. [citation needed]
* Designed "from the ground up," according to sources at ATI, for DX10, however it will be backwards compatible with DX9.
* Designed for Windows Vista
* PCI-E Interface
* HDMI connector (may have support for Displayport )
* Anandtech reports that the next generation of GPUs will range in power consumption from 130W to 300W. This increase in power consumption will make for higher-wattage Power Supply Units (to the 1 kW-1.2 kW range) and/or the addition of internal, secondary PSUs solely for powering the GPUs.[5]
* The G80 is expected to be released anywhere from October to November of 2006.
* It will be the first Direct3D 10 and Shader Model 4.0 card to be released.
* The G80 will be a 32 pipeline card with 32 pixel shaders, 16 vertex shaders, and 16 geometry shaders per core. The vertex shader pool and geometry shader pool may be unified.
* There will be 2 cards released at launch, the flagship 8800GTX and the slightly slower 8800GT. The 8800GTX will have 512 MB of ram while the 8800GT will have from 256 MB-512 MB of RAM and will also have slower core and memory clock speeds, and possibly fewer pipelines, than the 8800GTX.
* The launch price of the 8800GTX will be $499-599 while the 8800GT will initially cost $299-399. [1]
* Following the introduction of the flagship 8800 range, NVidia will introduce cheaper, slower versions of the card with one or more of the following reductions in capability: (a) reduced core clock speed; (b) reduced memory clock speed; (c) reduced memory bus width; (d) reduced pipeline count; (e) reduced chip count (if the 8800 is, in fact, a dual chip part); and / or (f) reduced complement of memory. If previous the NVidia naming convention is followed, mid-range products will be identified by model number 8600, and low-end parts will be named 8300.
* The G80 will not have unified shaders, but will still be DirectX10 compliant.
* The memory interface is said to be Samsung's new GDDR4 specification.
* The series is also rumored to have fast memory and core clocks, but will still be built on 90nm technology to avoid unnecessary risks.
* Anandtech reports that the next generation of GPUs will range in power consumption from 130W to 300W. This increase in power consumption will make for higher-wattage Power Supply Units (in the 800 to 1100 Watt range) and/or the addition of internal, secondary PSUs solely for powering the GPUs. [2]
* The G80 should also feature UDI connections with full HDMI support.
Anyone understand this about the G80? Its DX10 compliant but technically still a DX9 card?????
* The R600 is estimated to be the largest GPU ever made. Production based on an 80 nm process with ATI having 65 nm clearly in sights. [1]
* TSMC has only just begun 65 nm of simple designs.[2],
* The R600 might have over a 1000 MHz clock, making it the highest clocked GPU ever made.[3]
* The R600 will be the first graphics card faster than a X1800GTO able to do dongleless Crossfire (no master card). The X1800XT, X1900GT, X1900XT, and X1900XTX all require a slightly slower master card to run Crossfire. [4]
* 64 unified shader pipelines, 32 TMUs, and 32 ROPs.
* It shares a similar design of the Xbox 360's GPU "Xenos". However, the R600 will not have the 10 MB daughter embedded DRAM framebuffer of the "Xenos".
* The R600 will support the upcoming GDDR4 memory interface with 512 MB running higher than the X1950XTX's memory clock speeds of 1.0 GHz (2.0 GHz effective).
* The current target release date is December 2006. [citation needed]
* Designed "from the ground up," according to sources at ATI, for DX10, however it will be backwards compatible with DX9.
* Designed for Windows Vista
* PCI-E Interface
* HDMI connector (may have support for Displayport )
* Anandtech reports that the next generation of GPUs will range in power consumption from 130W to 300W. This increase in power consumption will make for higher-wattage Power Supply Units (to the 1 kW-1.2 kW range) and/or the addition of internal, secondary PSUs solely for powering the GPUs.[5]
* The G80 is expected to be released anywhere from October to November of 2006.
* It will be the first Direct3D 10 and Shader Model 4.0 card to be released.
* The G80 will be a 32 pipeline card with 32 pixel shaders, 16 vertex shaders, and 16 geometry shaders per core. The vertex shader pool and geometry shader pool may be unified.
* There will be 2 cards released at launch, the flagship 8800GTX and the slightly slower 8800GT. The 8800GTX will have 512 MB of ram while the 8800GT will have from 256 MB-512 MB of RAM and will also have slower core and memory clock speeds, and possibly fewer pipelines, than the 8800GTX.
* The launch price of the 8800GTX will be $499-599 while the 8800GT will initially cost $299-399. [1]
* Following the introduction of the flagship 8800 range, NVidia will introduce cheaper, slower versions of the card with one or more of the following reductions in capability: (a) reduced core clock speed; (b) reduced memory clock speed; (c) reduced memory bus width; (d) reduced pipeline count; (e) reduced chip count (if the 8800 is, in fact, a dual chip part); and / or (f) reduced complement of memory. If previous the NVidia naming convention is followed, mid-range products will be identified by model number 8600, and low-end parts will be named 8300.
* The G80 will not have unified shaders, but will still be DirectX10 compliant.
* The memory interface is said to be Samsung's new GDDR4 specification.
* The series is also rumored to have fast memory and core clocks, but will still be built on 90nm technology to avoid unnecessary risks.
* Anandtech reports that the next generation of GPUs will range in power consumption from 130W to 300W. This increase in power consumption will make for higher-wattage Power Supply Units (in the 800 to 1100 Watt range) and/or the addition of internal, secondary PSUs solely for powering the GPUs. [2]
* The G80 should also feature UDI connections with full HDMI support.
Anyone understand this about the G80? Its DX10 compliant but technically still a DX9 card?????
