Recent changes to Performance

WikiPage Performance modified by Shane Saxon

Shane Saxon — Tue, 17 Jul 2012 23:50:40 -0000

--- v16
+++ v17
@@ -38,7 +38,7 @@
 -----
 Things to Consider
 
-Stereoscopic 3D with OpenGL requires [nVidia Quadro](http://www.nvidia.com/object/workstation-solutions.html) or [ATI FirePro](www.amd.com/firepro) card.
+Stereoscopic 3D with OpenGL requires [nVidia Quadro](http://www.nvidia.com/object/workstation-solutions.html) or [ATI FirePro](http://www.amd.com/firepro) card.
 
 Important to have a good power supply and lots of cooling.

WikiPage Performance modified by Shane Saxon

Shane Saxon — Tue, 17 Jul 2012 23:40:34 -0000

--- v15
+++ v16
@@ -70,9 +70,12 @@
 -------
 Near Future:
 
-Projected future potential with code optimizations for multiple cores and using current 4-way SLI with a pair of [GeForce GTX 690](http://www.geforce.com/hardware/desktop-gpus/geforce-gtx-690/specifications) cards).
+Projected future performance based on code optimizations to improve the efficiency of OpenGL draw routines and take advantage of multi-GPU configurations. 
 
-500,000 records displayed as 3D objects in space (pin nodes)
+Estimate based on a 4-way SLI config using a pair of [GeForce GTX 690](http://www.geforce.com/hardware/desktop-gpus/geforce-gtx-690/specifications) cards.
+
+500,000 records displayed as 3D objects (pin nodes)
+- 30 distinct parameters displayed per record object
 - 500,000 x 30 = 15 million visualized parameters
 
 10Gb/s realtime data
@@ -81,7 +84,7 @@
 Cluster Visual Environments:
 We plan to implement the [Equalizer](http://www.equalizergraphics.com/) C++ library to help with synchronization of cluster based visual environments, such as VR Caves. ANTz is currently based on GLUT which allows for porting to systems such as the [Star Cave at Calit2](http://ivl.calit2.net/wiki/index.php/Infrastructure).
 
-Genlock is a crucial component to cluster based tiled displays that is often overlooked, (or ignored due to cost / performance of commodity game cards.) When operating in a high-speed environment, it is often deemed 'necessary' to synchronize the vertical refresh across all systems. Both the [nVidia Quadro G-Sync](http://www.nvidia.com/page/quadrofx_gsync.html) card and [ATI FirePro S400 Synchronization Module.](http://www.amd.com/us/PRODUCTS/WORKSTATION/GRAPHICS/S400/Pages/s400.aspx) are available solutions.
+Genlock is a crucial component to cluster based tiled displays that is often overlooked, (or ignored due to cost / performance of commodity game cards.) When operating in a high-speed environment, it is often deemed 'necessary' to synchronize the vertical refresh across all systems. Available solutions include the [nVidia Quadro G-Sync](http://www.nvidia.com/page/quadrofx_gsync.html) card and [AMD ATI FirePro S400 Synchronization Module.](http://www.amd.com/us/PRODUCTS/WORKSTATION/GRAPHICS/S400/Pages/s400.aspx)
 
 To help with physics interactions and data processing we plan to use the [OpenCL](http://www.khronos.org/opencl/) library by the Khronos Group. It allows writing C code to run across multicore GPU, Cell and CPU processors.

WikiPage Performance modified by Shane Saxon

Shane Saxon — Tue, 17 Jul 2012 23:25:55 -0000

--- v14
+++ v15
@@ -2,16 +2,18 @@
 
 ANTz - minimize latency, while maximizing bandwidth and processing power.
 
+also see [Performance_Tips] on how to choose hardware and optimize content.
+ 
+updated on 2012-07-17
+
 -----
-Performance as of 2012-07-17
+Performance
 
 <33ms latency for most user operations at full frame rate (60FPS.) 
 
-User commands change the scene state and typically take 1-2 cycles to be displayed. Some functions such as file loading can take seconds or minutes. Internal calculations are usually 1 cycle, live external IO is 1-2 cycles, (keyboard, mouse, etc....) Video latency can only be reduced to 2 cycles by using either [nVidia Quadro SDI Capture](http://www.nvidia.com/object/product_quadro_sdi_capture_us.html) or [AMD FirePro SDI-Link](http://www.amd.com/us/products/workstation/graphics/amd-sdi-link/Pages/amd-sdi-link-platform.aspx). 
+User commands typically effect the displayed scene state within 1-2 cycles. However, some functions can take seconds or minutes, (such as loading a large dataset.) Internal calculations are mostly within 1 cycle, live external IO is 1-2 cycles, (keyboard, mouse, audio, etc....) Video latency is typically 3-7 frames or more depending on compression codec. Low-latency video of 2-3 frames is possible with either [nVidia Quadro SDI Capture](http://www.nvidia.com/object/product_quadro_sdi_capture_us.html) or [AMD FirePro SDI-Link](http://www.amd.com/us/products/workstation/graphics/amd-sdi-link/Pages/amd-sdi-link-platform.aspx). 
 
 ANTz minimizes latency by running nearly all operations in a pseudo-state-machine fashion. Most operations are completed within the cycle and multi-cycle logic is avoided at all costs. However, occasional system hiccups (on MS Windows & OSX) consume about 50-100ms at a random interval. This can be minimized through careful system management or simply use Linux. For details see [State_Machine].
-
-Also see [Performance_Tips] on how to optimize content and hardware.
 
 -----
 Hardware Benchmarks
@@ -40,24 +42,26 @@
 
 Important to have a good power supply and lots of cooling.
 
-GPU is the primary performance factor. ANTz currently does NOT support SLI or CrossFireX multi-GPU architecture. So currently, the fastest single chip GPU solution is best (GTX 680.)
+GPU is the primary performance factor. ANTz currently does NOT support SLI or CrossFireX multi-GPU architecture. So currently, the fastest single chip GPU solution is best.
 
 CPU clock rate is a secondary factor, the more GHz the better.
 
-RAM should be fast enough to match the CPU(s). To run at full speed, multi-core CPU's typically require 3 or 4 sticks of DD3 RAM per chip.
+RAM should be fast enough to match the CPU(s). To run at full speed, multi-core CPU's require multiple sticks of RAM, based on the number of cores.
 
-ANTz has been tested to work on netbooks and older systems running Windows XP, Linux and Mac OSX 10.5.8 with Intel and G4 PPC.
+ANTz has been tested to work on netbooks and older systems running Windows XP, Linux and Mac OSX 10.5.8 on Intel and G4 PowerPC systems.
 
-Systems w/o GPU perform poorly (integrated / on-board video chip.)
+Systems w/o a separate GPU perform poorly, aka: integrated / on-board video chip.
 
 -----
-4Gbps realtime IO
+Realtime IO Channels
 
-Approximate estimate based on related tests using Fibre Channel to the [Violin Memory 1010](http://www.violin-memory.com/) FLASH cache.
+Approximate performance estimate based on related tests using Fibre Channel to the [Violin Memory 1010](http://www.violin-memory.com/) FLASH cache.
 
-Size of the external dataset is a factor of the network infrastructure.
-- multi-TeraByte datasets will likely require FLASH cache
-- 100,000 *IOPS per ANTz system require a minimum of 4TB SSA storage, (Solid State Array based on FLASH.)
+4Gbps of IO - node parameters to/from track data.
+
+Maximum size of the external dataset is a factor of the network infrastructure.
+- multi-TeraByte datasets are accessible in realtime when using FLASH cache.
+- 100,000 *IOPS per ANTz system requires a minimum of 4TB SSA storage, (Solid State Array based on FLASH.)
 
 *IOPS = IO operations per second (4kB records.)

WikiPage Performance modified by Shane Saxon

Shane Saxon — Tue, 17 Jul 2012 22:26:05 -0000

--- v13
+++ v14
@@ -2,74 +2,84 @@
 
 ANTz - minimize latency, while maximizing bandwidth and processing power.
 
-Latency is <33ms for most user operations when running at full frame rate 
-(60FPS.) User commands change the scene state and typically take 1-2 cycles to be displayed. Some functions can take seconds or minutes, such as file loading. Internal calculations are usually 1 cycle, live external IO is 1-2 cycles, (keyboard, mouse, etc....) Video latency can only be reduced to 2 cycles by using either [nVidia Quadro SDI Capture](http://www.nvidia.com/object/product_quadro_sdi_capture_us.html) or [AMD FirePro SDI-Link](http://www.amd.com/us/products/workstation/graphics/amd-sdi-link/Pages/amd-sdi-link-platform.aspx). 
+-----
+Performance as of 2012-07-17
+
+<33ms latency for most user operations at full frame rate (60FPS.) 
+
+User commands change the scene state and typically take 1-2 cycles to be displayed. Some functions such as file loading can take seconds or minutes. Internal calculations are usually 1 cycle, live external IO is 1-2 cycles, (keyboard, mouse, etc....) Video latency can only be reduced to 2 cycles by using either [nVidia Quadro SDI Capture](http://www.nvidia.com/object/product_quadro_sdi_capture_us.html) or [AMD FirePro SDI-Link](http://www.amd.com/us/products/workstation/graphics/amd-sdi-link/Pages/amd-sdi-link-platform.aspx). 
 
 ANTz minimizes latency by running nearly all operations in a pseudo-state-machine fashion. Most operations are completed within the cycle and multi-cycle logic is avoided at all costs. However, occasional system hiccups (on MS Windows & OSX) consume about 50-100ms at a random interval. This can be minimized through careful system management or simply use Linux. For details see [State_Machine].
 
 Also see [Performance_Tips] on how to optimize content and hardware.
 
 -----
-Performance:
-
-Our core development principle is to minimize latency, while maximizing bandwidth and processing power.
-
--------
-Current Performance:
+Hardware Benchmarks
 
 ....
 Laptop - (MacBook Pro - Early 2011 - 2.2Ghz i7, 8GB, Radeon HD 6750M 1GB)
 
-25,000 records rendered in realtime (15-30fps) as 3D objects (nodes.)
-- a) approx. 30 visual/spatial parameters per node
-- b) 25,000 x 30 = 750,000 visualized parameters
-
+25,000 nodes rendered in realtime (15-30fps) as 3D objects (nodes.)
 
 .....
-Desktop - [ASUS Rampage IV Formula](http://www.asus.com/Motherboards/Intel_Socket_2011/Rampage_IV_Formula/), 3.6GHz Intel i7 3820 CPU, 8GB, 
-[nVidia GeForce GTX 680](http://www.geforce.com/hardware/desktop-gpus/geforce-gtx-680/specifications))
-ASUS Rampage IV Extreme (or IV Formula motherboard)
+Desktop - ([ASUS Rampage IV](http://www.asus.com/Motherboards/Intel_Socket_2011/Rampage_IV_Extreme/), 3.6GHz i7 3820, 16GB 4X4GB, [GeForce GTX 680](http://www.geforce.com/hardware/desktop-gpus/geforce-gtx-680/specifications))
 
-.....
-Workstation - ([ASUS Z9PE-D8 WS](http://www.asus.com/Motherboards/Intel_Socket_2011/Z9PED8_WS/), Dual Xeon E5-2600 series CPUs, 16GB, 
-[nVidia GeForce GTX 680](http://www.geforce.com/hardware/desktop-gpus/geforce-gtx-680/specifications))
+Workstation - ([ASUS Z9PE-D8 WS](http://www.asus.com/Motherboards/Intel_Socket_2011/Z9PED8_WS/), Dual 3.3GHz Xeon E5-2643, 32GB 8X4GB, [GeForce GTX 680](http://www.geforce.com/hardware/desktop-gpus/geforce-gtx-680/specifications))
 
-100,000 records rendered in realtime (15-30fps) as 3D objects (nodes.)
-- a) approx. 30 visual/spatial parameters per node
-- b) 100,000 x 30 = 3 million visualized parameters
+100,000 nodes rendered in realtime (15-30fps) as 3D objects.
 
-Stereoscopic 3D requires a nVidia Quadro or ATI Fire Pro card.
-[nVidia Quadro 6000](http://www.nvidia.com/object/workstation-solutions.html))
+- approx. 30 visual/spatial parameters per node
+- 100,000 x 30 = 3 million visualized parameters
 
-4Gbps realtime data
-- reasonable expectation... based on related tests using Fibre Channel to the [Violin Memory 1010](http://www.violin-memory.com/) FLASH cache.
+Desktop and Workstation performance are similar. Future upgrades to the code will likely change this by utilizing more cores and enable improved IO.
 
-.....
-size of the external dataset is a factor of the network infrastructure.
-- dataset size = multi-TeraByte (not cheap and likely requires FLASH cache..)
-- 100,000 *IOPS per ANTz workstation using minimum of a 4TB SSA storage, (Solid State Array based on FLASH.)
+-----
+Things to Consider
+
+Stereoscopic 3D with OpenGL requires [nVidia Quadro](http://www.nvidia.com/object/workstation-solutions.html) or [ATI FirePro](www.amd.com/firepro) card.
+
+Important to have a good power supply and lots of cooling.
+
+GPU is the primary performance factor. ANTz currently does NOT support SLI or CrossFireX multi-GPU architecture. So currently, the fastest single chip GPU solution is best (GTX 680.)
+
+CPU clock rate is a secondary factor, the more GHz the better.
+
+RAM should be fast enough to match the CPU(s). To run at full speed, multi-core CPU's typically require 3 or 4 sticks of DD3 RAM per chip.
+
+ANTz has been tested to work on netbooks and older systems running Windows XP, Linux and Mac OSX 10.5.8 with Intel and G4 PPC.
+
+Systems w/o GPU perform poorly (integrated / on-board video chip.)
+
+-----
+4Gbps realtime IO
+
+Approximate estimate based on related tests using Fibre Channel to the [Violin Memory 1010](http://www.violin-memory.com/) FLASH cache.
+
+Size of the external dataset is a factor of the network infrastructure.
+- multi-TeraByte datasets will likely require FLASH cache
+- 100,000 *IOPS per ANTz system require a minimum of 4TB SSA storage, (Solid State Array based on FLASH.)
 
 *IOPS = IO operations per second (4kB records.)
 
-**In general the above numbers are conservative estimates based on real-world testing of various core methods that ANTz builds upon... using state-of-the-art hardware such as Enterprise level FLASH cache by TMS and VM. The Fore-mentioned manufactures claim spec's 2-5X more, which are 'not-to-exceed' specs.
+**In general the above numbers are conservative estimates based on real-world testing of various core methods that ANTz builds upon... using state-of-the-art hardware such as Enterprise level FLASH cache by TMS and VM. The Fore-mentioned manufactures claim spec's 2-5X more, which are 'not-to-exceed' specs.
 
 -------
 Near Future:
 
-future potential (using the same hardware with code optimizations):
+Projected future potential with code optimizations for multiple cores and using current 4-way SLI with a pair of [GeForce GTX 690](http://www.geforce.com/hardware/desktop-gpus/geforce-gtx-690/specifications) cards).
 
 500,000 records displayed as 3D objects in space (pin nodes)
 - 500,000 x 30 = 15 million visualized parameters
 
-10Gb/s realtime data
+10Gb/s realtime data
 - core technology tests done using the [Texas Memory RAMSAN 610](http://www.ramsan.com/) and [Violin Memory 1010.](http://www.violin-memory.com/)
 
 Cluster Visual Environments:
-We plan to implement the [Equalizer](http://www.equalizergraphics.com/) C++ library to help with synchronization of cluster based visual environments, such as VR Caves. ANTz is currently based on GLUT which allows it to easily be ported to systems such as the [Star Cave at Calit2](http://ivl.calit2.net/wiki/index.php/Infrastructure).
+We plan to implement the [Equalizer](http://www.equalizergraphics.com/) C++ library to help with synchronization of cluster based visual environments, such as VR Caves. ANTz is currently based on GLUT which allows for porting to systems such as the [Star Cave at Calit2](http://ivl.calit2.net/wiki/index.php/Infrastructure).
 
-Genlock is a crucial component to cluster based tiled displays that is all too often overlooked, (or ignored due to cost.) When operating in a high-speed environment, it is 'necessary' to synchronize the vertical refresh across all systems. Both the [nVidia Quadro G-Sync](http://www.nvidia.com/page/quadrofx_gsync.html) card and [ATI FirePro S400 Synchronization Module.](http://www.amd.com/us/PRODUCTS/WORKSTATION/GRAPHICS/S400/Pages/s400.aspx) are available solutions.
+Genlock is a crucial component to cluster based tiled displays that is often overlooked, (or ignored due to cost / performance of commodity game cards.) When operating in a high-speed environment, it is often deemed 'necessary' to synchronize the vertical refresh across all systems. Both the [nVidia Quadro G-Sync](http://www.nvidia.com/page/quadrofx_gsync.html) card and [ATI FirePro S400 Synchronization Module.](http://www.amd.com/us/PRODUCTS/WORKSTATION/GRAPHICS/S400/Pages/s400.aspx) are available solutions.
 
-To help with physics interactions and data processing we plan to use the [OpenCL](http://www.khronos.org/opencl/) library by the Khronos Group It allows writing C code to run on powerful GPU, Cell and multi-core CPU processors.
+To help with physics interactions and data processing we plan to use the [OpenCL](http://www.khronos.org/opencl/) library by the Khronos Group. It allows writing C code to run across multicore GPU, Cell and CPU processors.
 
 --------
 Ultimate Goal:
@@ -80,4 +90,4 @@
 - IO bandwidth on par with the spatial and temporal resolution of the nervous systems external senses.
 - Processing power must be sufficient to create the illusion of a real-world physical environment.
 
-It is a reasonable expectation that capable hardware will exist within the not too distant future, (if Moore's Law holds up.) Do not confuse this requirement with that of a full Human Being level AI, this is not required. We have plenty of people to form the 'Intelligence' component of the mixed neural network. We simply need to link human beings in a fully immersive virtual universe.
+It is a reasonable expectation that capable hardware will exist within the not too distant future, (if Moore's Law holds up.) Do not confuse this requirement with that of a full Human Being level AI, this is not required. We have plenty of people to form the 'Intelligence' component of the mixed neural network. We simply need to link human beings in a fully immersive mixed reality.

WikiPage Performance modified by Shane Saxon

Shane Saxon — Tue, 17 Jul 2012 07:31:36 -0000

--- v12
+++ v13
@@ -2,7 +2,8 @@
 
 ANTz - minimize latency, while maximizing bandwidth and processing power.
 
-Latency is <33ms for nearly all operations, (other then file loading.) You can typically expect end-to-end latency between 1-3 ANTz cycles, internal calculations are 1 cycle, most external IO is 1-2 cycles. Video latency can only be reduced to 2 cycles by using either [nVidia Quadro SDI Capture](http://www.nvidia.com/object/product_quadro_sdi_capture_us.html) or [AMD FirePro SDI-Link](http://www.amd.com/us/products/workstation/graphics/amd-sdi-link/Pages/amd-sdi-link-platform.aspx). 
+Latency is <33ms for most user operations when running at full frame rate 
+(60FPS.) User commands change the scene state and typically take 1-2 cycles to be displayed. Some functions can take seconds or minutes, such as file loading. Internal calculations are usually 1 cycle, live external IO is 1-2 cycles, (keyboard, mouse, etc....) Video latency can only be reduced to 2 cycles by using either [nVidia Quadro SDI Capture](http://www.nvidia.com/object/product_quadro_sdi_capture_us.html) or [AMD FirePro SDI-Link](http://www.amd.com/us/products/workstation/graphics/amd-sdi-link/Pages/amd-sdi-link-platform.aspx). 
 
 ANTz minimizes latency by running nearly all operations in a pseudo-state-machine fashion. Most operations are completed within the cycle and multi-cycle logic is avoided at all costs. However, occasional system hiccups (on MS Windows & OSX) consume about 50-100ms at a random interval. This can be minimized through careful system management or simply use Linux. For details see [State_Machine].
 
@@ -19,28 +20,38 @@
 ....
 Laptop - (MacBook Pro - Early 2011 - 2.2Ghz i7, 8GB, Radeon HD 6750M 1GB)
 
-25,000 records rendered in realtime (30fps) as 3D objects (called pin nodes.)
+25,000 records rendered in realtime (15-30fps) as 3D objects (nodes.)
 - a) approx. 30 visual/spatial parameters per node
 - b) 25,000 x 30 = 750,000 visualized parameters
 
+
 .....
-Workstation - ([SuperMicro X8DAH+-F](http://www.supermicro.com/products/motherboard/QPI/5500/X8DAH_-F.cfm), 12GB, [nVidia Quadro 6000](http://www.nvidia.com/object/workstation-solutions.html))
+Desktop - [ASUS Rampage IV Formula](http://www.asus.com/Motherboards/Intel_Socket_2011/Rampage_IV_Formula/), 3.6GHz Intel i7 3820 CPU, 8GB, 
+[nVidia GeForce GTX 680](http://www.geforce.com/hardware/desktop-gpus/geforce-gtx-680/specifications))
+ASUS Rampage IV Extreme (or IV Formula motherboard)
 
-100,000 records rendered in realtime (30fps) as 3D objects (called pin nodes.)
+.....
+Workstation - ([ASUS Z9PE-D8 WS](http://www.asus.com/Motherboards/Intel_Socket_2011/Z9PED8_WS/), Dual Xeon E5-2600 series CPUs, 16GB, 
+[nVidia GeForce GTX 680](http://www.geforce.com/hardware/desktop-gpus/geforce-gtx-680/specifications))
+
+100,000 records rendered in realtime (15-30fps) as 3D objects (nodes.)
 - a) approx. 30 visual/spatial parameters per node
 - b) 100,000 x 30 = 3 million visualized parameters
 
-4Gb/s realtime data
-- reasonable expectation... based on related tests using Fiber Channel to the [Violin Memory 1010](http://www.violin-memory.com/) FLASH cache.
+Stereoscopic 3D requires a nVidia Quadro or ATI Fire Pro card.
+[nVidia Quadro 6000](http://www.nvidia.com/object/workstation-solutions.html))
+
+4Gbps realtime data
+- reasonable expectation... based on related tests using Fibre Channel to the [Violin Memory 1010](http://www.violin-memory.com/) FLASH cache.
 
 .....
 size of the external dataset is a factor of the network infrastructure.
-- dataset size = multi-PetaByte (not cheap and requries FLASH cache..)
-- 100,000 *IOPS per ANTz workstation using 2-24TB FLASH cache, (capacity depends on $$...)
+- dataset size = multi-TeraByte (not cheap and likely requires FLASH cache..)
+- 100,000 *IOPS per ANTz workstation using minimum of a 4TB SSA storage, (Solid State Array based on FLASH.)
 
 *IOPS = IO operations per second (4kB records.)
 
-**In general the above numbers are conservative estimates based on real-world testing of various core methods that ANTz builds upon... using state-of-the-art hardware such as Enterprise level FLASH cache by TMS and VM. The Fore-mentioned manufactures claim spec's 2-5X that of the above, but are more like 'not-to-exceed' specs.
+**In general the above numbers are conservative estimates based on real-world testing of various core methods that ANTz builds upon... using state-of-the-art hardware such as Enterprise level FLASH cache by TMS and VM. The Fore-mentioned manufactures claim spec's 2-5X more, which are 'not-to-exceed' specs.
 
 -------
 Near Future:

WikiPage Performance modified by Shane Saxon

Shane Saxon — Sat, 21 Apr 2012 20:42:25 -0000

--- v11 
+++ v12 
@@ -6,6 +6,8 @@
 
 ANTz minimizes latency by running nearly all operations in a pseudo-state-machine fashion. Most operations are completed within the cycle and multi-cycle logic is avoided at all costs. However, occasional system hiccups (on MS Windows & OSX) consume about 50-100ms at a random interval. This can be minimized through careful system management or simply use Linux. For details see [State_Machine].
 
+Also see [Performance_Tips] on how to optimize content and hardware.
+
 -----
 Performance:

WikiPage Performance modified by Shane Saxon

Shane Saxon — Sat, 21 Apr 2012 20:12:27 -0000

--- v10 
+++ v11 
@@ -2,63 +2,69 @@
 
 ANTz - minimize latency, while maximizing bandwidth and processing power.
 
-Latency for nearly all operations (other the file loading) is typically less then 1 frame cycle (FPS.) So at full frame rate (60Hz) typical latency is <33ms. However, occasional system hiccups (on MS Windows & OSX) consume about 50-100ms at a random interval. Hiccups can be minimized through careful system management or simply use Linux. ANTz minimizes latency by running nearly all operations in a pseudo-state-machine fashion. In other words it does everything within the cycle and multi-cycle logic is avoided at all costs and hence nearly non-existent. For details see [State_Machine].
-
+Latency is <33ms for nearly all operations, (other then file loading.) You can typically expect end-to-end latency between 1-3 ANTz cycles, internal calculations are 1 cycle, most external IO is 1-2 cycles. Video latency can only be reduced to 2 cycles by using either [nVidia Quadro SDI Capture](http://www.nvidia.com/object/product_quadro_sdi_capture_us.html) or [AMD FirePro SDI-Link](http://www.amd.com/us/products/workstation/graphics/amd-sdi-link/Pages/amd-sdi-link-platform.aspx). 
+
+ANTz minimizes latency by running nearly all operations in a pseudo-state-machine fashion. Most operations are completed within the cycle and multi-cycle logic is avoided at all costs. However, occasional system hiccups (on MS Windows & OSX) consume about 50-100ms at a random interval. This can be minimized through careful system management or simply use Linux. For details see [State_Machine].
+
 -----
 Performance:
 
 Our core development principle is to minimize latency, while maximizing bandwidth and processing power.
 
 -------
 Current Performance:
 
 ....
 Laptop - (MacBook Pro - Early 2011 - 2.2Ghz i7, 8GB, Radeon HD 6750M 1GB)
 
 25,000 records rendered in realtime (30fps) as 3D objects (called pin nodes.)
 - a) approx. 30 visual/spatial parameters per node
 - b) 25,000 x 30 = 750,000 visualized parameters
 
 .....
 Workstation - ([SuperMicro X8DAH+-F](http://www.supermicro.com/products/motherboard/QPI/5500/X8DAH_-F.cfm), 12GB, [nVidia Quadro 6000](http://www.nvidia.com/object/workstation-solutions.html))
 
 100,000 records rendered in realtime (30fps) as 3D objects (called pin nodes.)
 - a) approx. 30 visual/spatial parameters per node
 - b) 100,000 x 30 = 3 million visualized parameters
 
 4Gb/s realtime data
 - reasonable expectation... based on related tests using Fiber Channel to the [Violin Memory 1010](http://www.violin-memory.com/) FLASH cache.
 
 .....
 size of the external dataset is a factor of the network infrastructure.
 - dataset size = multi-PetaByte (not cheap and requries FLASH cache..)
 - 100,000 *IOPS per ANTz workstation using 2-24TB FLASH cache, (capacity depends on $$...)
 
-*IOPS = IO operations per second (think in terms of number of 4kB records.)
-
-**In general the above numbers are reasonable guesses based on real-world testing of various core methods that ANTz builds upon....  and using state-of-the-art hardware...  Enterprise level FLASH cache... by TMS and VM
+*IOPS = IO operations per second (4kB records.)
+
+**In general the above numbers are conservative estimates based on real-world testing of various core methods that ANTz builds upon... using state-of-the-art hardware such as Enterprise level FLASH cache by TMS and VM. The Fore-mentioned manufactures claim spec's 2-5X that of the above, but are more like 'not-to-exceed' specs.
 
 -------
 Near Future:
 
-future potential (using above hardware spec with code optimizations):
+future potential (using the same hardware with code optimizations):
 
 500,000 records displayed as 3D objects in space (pin nodes)
 - 500,000 x 30 = 15 million visualized parameters
 
 10Gb/s realtime data
 - core technology tests done using the [Texas Memory RAMSAN 610](http://www.ramsan.com/) and [Violin Memory 1010.](http://www.violin-memory.com/)
 
 Cluster Visual Environments:
-[Equalizer](http://www.equalizergraphics.com/) C library to help with cluster synchronization of visual environments such as VR caves.  It is also worth noting that ANTz is based on GLUT which allows to be easily ported into other systems such as the [Star Cave at Calit2](http://ivl.calit2.net/wiki/index.php/Infrastructure).
-
-Genlock - A crucial component to cluster based environments that is often overlooked (or ignored due to cost,) is genlock to synchronize the vertical refresh across systems. Both the [nVidia Quadro G-Sync](http://www.nvidia.com/page/quadrofx_gsync.html) option and [ATI FirePro S400 Synchronization Module.](http://www.amd.com/us/PRODUCTS/WORKSTATION/GRAPHICS/S400/Pages/s400.aspx)
-
-[OpenCL](http://www.khronos.org/opencl/) - by the Khronos Group allows writing C code to run on powerful GPU, Cell and multi-core CPU processors.
+We plan to implement the [Equalizer](http://www.equalizergraphics.com/) C++ library to help with synchronization of cluster based visual environments, such as VR Caves. ANTz is currently based on GLUT which allows it to easily be ported to systems such as the [Star Cave at Calit2](http://ivl.calit2.net/wiki/index.php/Infrastructure).
+
+Genlock is a crucial component to cluster based tiled displays that is all too often overlooked, (or ignored due to cost.) When operating in a high-speed environment, it is 'necessary' to synchronize the vertical refresh across all systems. Both the [nVidia Quadro G-Sync](http://www.nvidia.com/page/quadrofx_gsync.html) card and [ATI FirePro S400 Synchronization Module.](http://www.amd.com/us/PRODUCTS/WORKSTATION/GRAPHICS/S400/Pages/s400.aspx) are available solutions.
+
+To help with physics interactions and data processing we plan to use the [OpenCL](http://www.khronos.org/opencl/) library by the Khronos Group It allows writing C code to run on powerful GPU, Cell and multi-core CPU processors.
 
 --------
 Ultimate Goal:
 
 The ultimate goal is to close the human-computer cognitive loop.
 
-This requires that latency be below the threshold at which humans perceive interactions in real-time. Need sufficient IO bandwidth to provide full spatial and temporal resolution on par with the nervous systems external senses. The logic engines processing power is responsible for creating the illusion of a real-world physical environment. It is a reasonable expectation that capable hardware will exist within the not too distant future, (if Moore's Law holds up.) Do not confuse this requirement with that of a full Human Being level AI, this is not required. We have plenty of people and simply need to immerse them in a group environment.
+- Latency should be kept below the threshold at which humans perceive interactions in real-time.
+- IO bandwidth on par with the spatial and temporal resolution of the nervous systems external senses.
+- Processing power must be sufficient to create the illusion of a real-world physical environment.
+
+It is a reasonable expectation that capable hardware will exist within the not too distant future, (if Moore's Law holds up.) Do not confuse this requirement with that of a full Human Being level AI, this is not required. We have plenty of people to form the 'Intelligence' component of the mixed neural network. We simply need to link human beings in a fully immersive virtual universe.

WikiPage Performance modified by Shane Saxon

Shane Saxon — Sat, 21 Apr 2012 07:46:30 -0000

--- v9 
+++ v10 
@@ -59,4 +59,6 @@
 --------
 Ultimate Goal:
 
-The ultimate goal is to close the human-computer cognitive loop. That means we must bring latency down to the threshold at which humans perceive interactions as real-time. Bandwidth and resolution should be on par with the nervous systems external senses. The logic engines processing power needs to be sufficient enough to create the illusion of a real-world physical environment. It is a reasonable expectation that capable hardware will exist within the not to distant future, (if Moore's Law holds up.) Do not confuse this requirement with that of a full Human Being level AI, this is not required. We have plenty of people and simply need to immerse them in a group environment.
+The ultimate goal is to close the human-computer cognitive loop.
+
+This requires that latency be below the threshold at which humans perceive interactions in real-time. Need sufficient IO bandwidth to provide full spatial and temporal resolution on par with the nervous systems external senses. The logic engines processing power is responsible for creating the illusion of a real-world physical environment. It is a reasonable expectation that capable hardware will exist within the not too distant future, (if Moore's Law holds up.) Do not confuse this requirement with that of a full Human Being level AI, this is not required. We have plenty of people and simply need to immerse them in a group environment.

WikiPage Performance modified by Shane Saxon

Shane Saxon — Sat, 21 Apr 2012 07:42:29 -0000

--- v8 
+++ v9 
@@ -59,14 +59,4 @@
 --------
 Ultimate Goal:
 
-The ultimate goal is to close the human-computer cognitive loop.
-
-So target performance spec's are as follows:
-
-IO for audio, visual and haptic interfaces that achieve the spatial and temporal resolution (latency) of the human nervous system (whole body.)
-
-1ms visual and haptic latency (need a 1000Hz holographic display at a [resolution suitable to the human eye.](http://clarkvision.com/imagedetail/eye-resolution.html) )
-
-Processing power needs to be many orders of magnitude beyond what is currently available. Even existing technologies in the lab could achieve this. For example, a diamond substrate based holographic quantum processor. Think in terms of using interference patterns of light to run massively parallel multivariable calculations in femto seconds and couple that with the advantage of entangled photons. 
-
-Perhaps more important is to obtain full psychological emersion for a group of users. This is not a matter of hardware but rather software/firmware. We are likely to have computers fast enough within the century, but will we have the code to make them fly.  For some perspective, we could have full fledged 'Lawnmower Man' style VR with the hardware of today. But currently the code does not exist, nor are we even close!  So write code for tomorrows hardware, it will take many coders working together several decades to achieve what we have long dreamed of and could have today, if only thousands of us had just gone for it 30 years ago.
+The ultimate goal is to close the human-computer cognitive loop. That means we must bring latency down to the threshold at which humans perceive interactions as real-time. Bandwidth and resolution should be on par with the nervous systems external senses. The logic engines processing power needs to be sufficient enough to create the illusion of a real-world physical environment. It is a reasonable expectation that capable hardware will exist within the not to distant future, (if Moore's Law holds up.) Do not confuse this requirement with that of a full Human Being level AI, this is not required. We have plenty of people and simply need to immerse them in a group environment.

WikiPage Performance modified by Shane Saxon

Shane Saxon — Fri, 20 Apr 2012 00:40:24 -0000

--- v7 
+++ v8 
@@ -2,7 +2,7 @@
 
 ANTz - minimize latency, while maximizing bandwidth and processing power.
 
-Latency for nearly all operations (other the file loading) is typically less then 1 frame cycle (FPS.) So at full frame rate (60Hz) typical latency is <33ms. However, occasional system hiccups (on MS Windows & OSX) consume about 50-100ms at a random interval. Hiccups can be minimized through careful system management or simply use Linux. ANTz minimizes latency by running nearly all operations in a pseudo-state-machine fashion. In other words it does everything within the cycle and multi-cycle logic is avoided at all costs and hence nearly non-existent. For detail see [State_Machine].
+Latency for nearly all operations (other the file loading) is typically less then 1 frame cycle (FPS.) So at full frame rate (60Hz) typical latency is <33ms. However, occasional system hiccups (on MS Windows & OSX) consume about 50-100ms at a random interval. Hiccups can be minimized through careful system management or simply use Linux. ANTz minimizes latency by running nearly all operations in a pseudo-state-machine fashion. In other words it does everything within the cycle and multi-cycle logic is avoided at all costs and hence nearly non-existent. For details see [State_Machine].
 
 -----
 Performance: