D. Lavenier. SAMBA: Systolic Accelerator for Molecular Biological Applications. Technical Report RR 2845, INRIA, 1996.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....in the domain of general purpose supercomputers than cost effective parallel co processors. Second, for many applications, data buffering and recirculating among a small number of highspeed processors can be an effective solution. This idea has been adapted by a number of specialpurpose machines [4, 24, 40]. 2 Third, programmable processing elements that require several instructions to compute the basic function can be used. This solution, used by Kestrel, essentially allows the addition of programmability into the processing elements with little penalty over the single purpose system. If, for ....

.... BioSCAN(est) Merc1 Alpha5 Paragon Samba Decy 1 Decy 15 FDF 3 Merc2 Machines: BISP (single purpose, not completed) 5] BioSCAN (single purpose) 36] 15 board DecypherII (FPGA, 173k) 40] 5 board FDF (single purpose, 50k) 16,536 PE MasPar, 34] SAMBA (single purpose) [24], 1 and 2 board Mercury (single purpose) 4] Biocellerator (FPGA, about 50k) 7] 1 board Decypher II ( 18k) 32 node Paragon [31] 1024 PE MasPar, 5 DEC Alpha 3000 workstations [31] B SYS [16] and Sun Ultrasparc 140. Figure 3. Database search time as a function of protein query ....

[Article contains additional citation context not shown here]

D. Lavenier, "SAMBA: Systolic accelerators for molecular biological applications," Tech. Rep. 988, IRISA, 35042 Rennes Cedex, France, Mar. 1996.

....necessary to run the variety of algorithms required for analyzing DNA, RNA, and proteins. P NAC was the first such machine, and computed edit distance over a four character alphabet [22] More recent examples, better tuned to the needs of computational biology, include BioScan, BISP and Samba [29, 5, 20]. Reconfigurable systems are based on programmable logic such as field programmable gate arrays (FPGAs) or custom designed arrays. They are generally slower and have far lower PE densities than single purpose VLSI but can be faster than supercomputers. They are flexible, but the configuration ....

Dominique Lavenier. SAMBA: Systolic accelerators for molecular biological applications. Technical Report 988, IRISA, 35042 Rennes Cedex, France, March 1996.

.... a hardware implementation of the full Smith Waterman algorithm, though a complete system was not reported (Chow et al. 1991) Mercury and SAMBA, both currently in working systems, follow the BISP approach by implementing gapped sequence analysis and profiles in hardware (Brutlag et al. 1995; Lavenier, 1996). SAMBA uses PAM, mentioned below, as a host interface. The commercial Fast Data Finder (FDF) is primarily geared to text search, but has been adapted to sequence analysis, and a server is available for Fast Data Finder profile search (http: ulrec3.unil.ch software FDFGP form.html) The 5 board ....

Lavenier, D. (1996). SAMBA: Systolic accelerators for molecular biological applications. Technical Report 988 IRISA 35042 Rennes Cedex, France.

....of the data base, the pairwise comparison is performed in l Q l DB i 1 systolic cycles, instead of l Q l DB i steps on a sequential processor. The linear systolic structure has been implemented on a prototype called samba (Systolic Accelerator for Molecular Biological Applications) [4,5]. The machine houses 128 dedicated full custom Vlsi processors. The array is connected to a standard workstation through a Fpga interface which has the major role of managing the partitioning of the computation at the clock rate of the array. As a matter of fact, and as explained above, comparing ....

D. Lavenier. SAMBA: Systolic Accelerator for Molecular Biological Applications. Technical Report RR 2845, INRIA, 1996.

....of the data base, the pairwise comparison is performed in l Q l DB i Gamma 1 systolic cycles, instead of l Q Theta l DB i steps on a sequential processor. The linear systolic structure has been implemented on a prototype called samba (Systolic Accelerator for Molecular Biological Applications) [4,5]. The machine houses 128 dedicated full custom Vlsi processors. The array is connected to a standard workstation through a Fpga interface which has the major role of managing the partitioning of the computation at the clock rate of the array. As a matter of fact, and as explained above, comparing ....

D. Lavenier. SAMBA: Systolic Accelerator for Molecular Biological Applications. Technical Report RR 2845, INRIA, 1996.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC