|
|
|
| |
Texas Memory Systems, Inc. (TMS) delivers high-performance DSP hardware and high-performance DSP software. In classical computer marketing all computer manufacturers touted the speed of their hardware. Now as hardware gets incredibly fast the marketing emphasis is put on achieving this speed in real world applications. This usable performance is directly related to the hardware-software interface efficiency.
Despite four generations of TMS DSP products that have raised the performance bar by a cumulative factor of 600, TMS has maintained a uniform architecture for the user. This compatibility has allowed the user to reuse legacy code on next generation TMS hardware. TMS hardware is designed with current generation software in mind. No hardware can be sold without efficient software and an efficient hardware/software interface. TMS has always delivered high-performance hardware that has efficient, usable, and compatible software. |
|
|
|
|
|
The DSP Chip
TMS always strives to have the best cost-effective hardware in the DSP arena. It all comes down to the best DSP chip available. Traditionally, TMS designs special DSP chips to fill the needs of our customers. For over two years, TMS has been developing our next generation DSP chip and it is finally here. The TM-100 is running at top-speed (100 GFLOPS) for real-world applications. The TM-100 is software compatible with the previous generation DSP products, so it is executing user application programs written for the XP-30 and older cards. The first available PCI Express DSP Accelerator card with the TM-100 is the XP-100, available now.
|
|
TM-100 Function Timing
|
| Function |
Size |
TM-44 |
TM-100 |
Faster |
| Real FFT |
4K |
20 us |
2.3 us |
x8 |
| Real FFT |
64K |
410 us |
36 us |
x11 |
| Real FFT |
1M |
7,884 us |
787 us |
x10 |
| Complex FFT |
4k |
31 us |
3 us |
x10 |
| Complex FFT |
64K |
655 us |
49 us |
x13 |
| Complex FFT |
1M |
13,107 us |
1180 us |
x11 |
| Function |
NF |
Size |
TM-44 |
TM-100 |
Faster |
| Real CONV |
16 |
4k |
20 us |
1.5 us |
x13 |
| Real CONV |
64 |
4k |
81 us |
6.2 us |
x13 |
| Complex CONV |
64 |
4k |
328 us |
24 us |
x13 |
| CONV_DEC |
64 |
4k |
328 us |
24 us |
x13 |
| CONV_COL |
64 |
4k |
328 us |
24 us |
x13 |
| Function |
Size |
TM-44 |
TM-100 |
Faster |
| MMUL |
64x64 |
163 us |
6 us |
x26 |
| MMUL |
1024x1024 |
699,050 us |
25,200 us |
x26 |
|
|
TM-100 (100 GFLOPS) Architecture
The TM-100 is the latest generation DSP chip from Texas Memory Systems. Architecturally, it is very similar to the previous generation TM-44 chip except faster, wider, and deeper. It has a 333-MHz processor clock with a 667-MHz external memory clock. It has a dual core architecture that again doubles the processing power as compared to the TM-44. Finally, it has twice as many floating point processing units per core as compared to the TM-44. These additional processing units allow a 256-point radix FFT to be processed in one pass instead of two passes. Combined, these improvements provide a 10-12x application performance boost over the TM-44 based chip. With the dual core architecture, the XP-100 (one chip) looks like the XP-30 board (two chips).
As with all DSP chips, chip bandwidth is a very important parameter. The TM-100 has two front-side busses (I/O) and eight back-side busses (Mem) that run at 5-GB/s. These busses have been designed to complement the processing power of the TM-100 not to limit it.
|
|
XP-100
The XP-100 DSP Accelerator for PCI-Express is available now. It implements the same DSP math library that has been a TMS standard for all previous generations of DSP products. The TM-100 chip operates at 333 MHz with a dual core design. Each core is programmed independently with separate data memories. At TMS, we provide the total solution. While our DSP hardware is ranked at the top of the high-performance list, our included DSP software has also been received well and it is easy to use. The XP-100 comes with a mature DSP math library and many valuable software development tools to make programming the XP-100 faster and easier.
|
XP-30/XP-100 Application Comparison |
4k
FFT
|
Application |
XP-30
|
XP-100
|
Units |
2-ch TMIC |
17 |
61 |
MS/s |
Combiner |
18 |
66 |
MS/s |
2-ch SA-TMIC |
30 |
105 |
MS/s |
2-ch SA (50%) |
33 |
119 |
MS/s |
2-ch SA (25%) |
35 |
127 |
MS/s |
1-ch TMIC |
40 |
146 |
MS/s |
1-ch SA-TMIC |
80 |
289 |
MS/s |
|
For Complex Data, multiply the above rates by 0.58
|
|
64k
FFT
|
Application |
XP-30
|
XP-100
|
Units |
2-ch TMIC |
16 |
61 |
MS/s |
Combiner |
17 |
66 |
MS/s |
2-ch SA-TMIC |
28 |
105 |
MS/s |
2-ch SA (50%) |
31 |
119 |
MS/s |
2-ch SA (25%) |
32 |
127 |
MS/s |
1-ch TMIC |
36 |
146 |
MS/s |
1-ch SA-TMIC |
73 |
289 |
MS/s |
|
For Complex Data, multiply the above rates by 0.53
|
|
|
20 Years of DSP Experience
For 20 years TMS has been providing advanced DSP systems. During this time we have created many valuable hardware/software tools to ease the process of quickly designing custom solutions. TMS specializes in developing high-performance solutions in DSP co-processing, DSP parallel processing, real-time data acquisition, and real-time synchronization. We have unique insights into the issues of processing extremely large volumes of data quickly and efficiently. For years, TMS has developed software tools and techniques to harness the power of our products and to hide the complexity of large-scale multi-processor programming without resorting to overly complex parallel programming methods. Existing XP-30 programs can be re-compiled for the XP-100 for an immediate performance improvement. Additional performance improvement is available by using the new features. |
|
