SEMCAD X
muzaik.jpg
Hardware Acceleration

 SEMCAD X Hardware Acceleration

SEMCAD X bundled with the latest Acceleware CUDA libraries can run on cutting edge NVIDIA Quadro, Tesla and Fermi technology.  Accelerator cards and Cluster-In-A-Box Solutions are available, featuring larger memory size and bandwidth, higher solver speeds and huge simulation sizes

 

 Fermi GPUs

 

NVIDIA Tesla C2050 or

NVIDIA Tesla C2070 (A50)

The C2050 and C2070 cards bring the performance of a small cluster to the desktop. Both cards include graphic output.

Specifications

1 GPU

C2050 - 3 GB RAM

C2070 - 6 GB RAM

448 Cores

144 GB/s Memory Bandwith

Maximum problem size 150 Mcells                                               
Typical speedup 42x
Form Factor 4.376" x 9.75" Dual Slot
Power Consumption 238 W
Workstation requirements
Click here for more information

top

NVIDIA Tesla S2050 or

NVIDIA Tesla S2070 (Q50)

The S2050 and S2070 cards are rack mounted and do not have video output.

Specifications

4 GPU / 1U Server form factor

4x C2050 - 12 GB RAM

4x C2070 - 24 GB RAM

Maximum problem size 800 Mcells                                       
Typical speedup 170x
Form Factor 1.71'' x 17.425'' x 28.5''
Power Consumption 900 W

top

C-type cluster: C50-X

C-Type Clusters are basically clusters of CIBs (e.g., several quads put together via Infiniband high-speed network, additional hardware and drivers/software). This presentation is mainly based on Tesla S2050 or S2070 nodes, where X is the number of nodes (up to 16)

Specifications

4 GPUs per node

12 or 24 GB RAM per node

Simulation size more than 6 Gcells               
Speed more than 13 Gcells/s

 

 

 

 

top


 Tesla/Quadro GPUs

 

NVIDIA Quadro FX 5800  or

NVIDIA Tesla C1060 (A40)

The Quadro FX 5800 is the next generation of the Quadro FX 5600 (formerly Acceleware v3 card). It features higher memory and solver speeds and is an affordable solution for medium to large simulation sizes. The C1060 card has the same characteristics as the FX 5800, except it does not have video output.

Specifications

1 GPU

4 GB RAM

240 Cores

102 GB/s Memory Bandwith

Maximum problem size 120 Mcells                                               
Typical speedup 35x
Form Factor 4.376'' x 10.50'' Dual Slot
Power Consumption 200 W
Workstation requirements Click here for more information

top

NVIDIA Quadro Plex 2200 Model D2 (D40)

The CIB D40 is the next generation CIB1000 (Quadro Plex 1000 Model IV formerly Acceleware D30) . Its NVIDIA Quadro Plex 2200 Model D2 substation offers improved performance and memory throughput. The CIB D40 is easily maintainable workstation with simulation performance equivalent to a multi-node supercomputer.

Specifications

2 GPU

8 GB RAM

deskside chassis

Maximum problem size 240 Mcells                                                    
Typical speedup 70x
Form Factor Deskside or 3U Rackmount Kit

 

 

 

top

2 NVIDIA Quadro Plex 2200 Model D2 (in parallel) (Q40)

The CIB Q40 is the next generation CIB1500 (2 xQuadro Plex 1000 Model IV formerly Acceleware Q30) . The combined capacity of both NVIDIA Quadro Plex 2200 Model D2 substations is an ideal platform for highly complex industrial applications.

Specifications

4 GPU

16 GB RAM

Maximum problem size 480 Mcells                                  
Typical speedup 140x

 

 

 

 

top

NVIDIA Tesla S1070 (Q40)

The S1070 card is rack mounted and does not have video output.

Specifications

4 GPU / 1U Server form factor

16 GB RAM

Maximum problem size 480 Mcells                                       
Typical speedup 140x
Form Factor 1.73'' x 17.5'' x 28.5''
Power Consumption 700 W

top

C-type cluster: C40-X

C-Type Clusters are basically clusters of CIBs (e.g., several Q40s put together via Infiniband high-speed network, additional hardware and drivers/software). This presentation is mainly based on Tesla S1070 nodes, where X is the number of nodes (up to 16)

Specifications

4 GPUs per node

16 GB RAM per node

Simulation size more than 2 Gcells               
Speed more than 10 Gcells/s

 

 

 

 

top

 
 
UPCOMING EVENTS
 
 
SPEAG Seminars in China & Taiwan, May–June 2012
25/05/2012 - 01/06/2012 (China & Taiwan)
 
SPEAG @ IEEE IMS 2012
17/06/2012 - 22/06/2012 (Montreal, Canada)
 
SEMCAD X Workshop @ APS/URSI 2012
08/07/2012 - 12/07/2012 (Chicago, US)
 
created and designed by R.Ø.S.A.