Skip to main content

Resultados relacionados

Resultados totales (Incluyendo duplicados): 5
Encontrada(s) 1 página(s)

RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publication Contribución de congreso (ConferenceOutput)

Combining RAM technologies for hard-error recovery in L1 data caches working at very-low power modes

  • Lorente Garcés, Vicente Jesús
  • Sahuquillo Borrás, Julio|||0000-0001-8630-4846
  • Petit Martí, Salvador Vicente|||0000-0003-2426-4134
  • López Rodríguez, Pedro Juan|||0000-0003-4544-955X
  • Valero Bresó, Alejandro
  • Canal, Ramón
  • Duato Marín, José Francisco
Low-power modes in modern microprocessors rely on low frequencies and low voltages to reduce the energy budget. Nevertheless, manufacturing induced parameter variations can make SRAM cells unreliable producing hard errors at supply voltages below Vccmin. Recent proposals provide a rather low fault-coverage due to the fault coverage/overhead trade-off. We propose a new faulttolerant L1 cache, which combines SRAM and eDRAM cells in L1 data caches to provide 100% SRAM hard-error fault coverage. Results show that, compared to a conventional cache and assuming 50% failure probability at low-power mode, leakage and dynamic energy savings are by 85% and 62%, respectively, with a minimal impact on performance., This work was supported by the Spanish MICINN (TIN2010-18368) with the Consolider-Ingenio 2010 Programme co-funded by the European Commission FEDER funds (CSD2006-00046) and co-funded with the Plan E funds (TIN2009-14475-C04-01). Additionaly, it was supported by Generalitat de Catalunya (2009SGR1250), by FP7 program of the European Commission (TRAMS-248789), and by Spanish MINECO (TIN2012-38341-C04-01).
UPCommons. Portal del coneixement obert de la UPC
Publication Contribución de congreso (ConferenceOutput)

Fault-tolerant nanoscale architecture based on linear threshold gates with redundancy

  • Aymerich Capdevila, Nivard
  • Rubio Sola, Jose Antonio|||0000-0003-1625-1472
One of the main objectives of the data computing and memory industry is to keep and ever accelerate the increase of component density reached in nowadays integrated circuits in future technologies based on ultimate CMOS and new emerging research devices. The worldwide-accepted predictions with these technologies indicate a remarkable reduction of the components quality, because of the manufacturing process complexity and the erratic behavior of devices, causing a drop in the system reliability if we maintain the same design rules than today. Together with the introduction of new devices, new architectural design paradigms have to be included. Fault tolerant techniques are considered necessary and relevant in this scenario. In this paper we present a Fault-Tolerant Nanoscale architecture based on the implementation of logic systems with averaging cells linear threshold gates (AC-LTG). The sensitivity of the gates in relation with manufacturing and environment deviation is investigated and compared with the well known NAND multiplexing concept, showing that the AC-LTG is a valuable alternative in specific nanoscale conditions., Peer Reviewed
UPCommons. Portal del coneixement obert de la UPC
Publication Contribución de congreso (ConferenceOutput)

Combining RAM technologies for hard-error recovery in L1 data caches working at very-low power modes

  • Lorente, Vicente
  • Valero, Alejandro
  • Sahuquillo, Julio
  • Petit, Salvador
  • Canal Corretger, Ramon|||0000-0003-4542-204X
  • López, Pedro
  • Duato, José
Low-power modes in modern microprocessors rely on low frequencies and low voltages to reduce the energy budget. Nevertheless, manufacturing induced parameter variations can make SRAM cells unreliable producing hard errors at supply voltages below Vccmin. Recent proposals provide a rather low fault-coverage due to the fault coverage/overhead trade-off. We propose a new fault- tolerant L1 cache, which combines SRAM and eDRAM cells in L1 data caches to provide 100% SRAM hard-error fault coverage. Results show that, compared to a conventional cache and assuming 50% failure probability at low-power mode, leakage and dynamic energy savings are by 85% and 62%, respectively, with a minimal impact on performance., Peer Reviewed
Recercat. Dipósit de la Recerca de Catalunya
Publication Contribución de congreso (ConferenceOutput)

Combining RAM technologies for hard-error recovery in L1 data caches working at very-low power modes

  • Lorente, Vicente
  • Valero, Alejandro
  • Sahuquillo, Julio
  • Petit, Salvador
  • Canal Corretger, Ramon
  • López, Pedro
  • Duato, José
Low-power modes in modern microprocessors rely on low frequencies and low voltages to reduce the energy budget. Nevertheless, manufacturing induced parameter variations can make SRAM cells unreliable producing hard errors at supply voltages below Vccmin. Recent proposals provide a rather low fault-coverage due to the fault coverage/overhead trade-off. We propose a new fault- tolerant L1 cache, which combines SRAM and eDRAM cells in L1 data caches to provide 100% SRAM hard-error fault coverage. Results show that, compared to a conventional cache and assuming 50% failure probability at low-power mode, leakage and dynamic energy savings are by 85% and 62%, respectively, with a minimal impact on performance., Peer Reviewed, Postprint (published version)
Recercat. Dipósit de la Recerca de Catalunya
Publication Contribución de congreso (ConferenceOutput)

Fault-tolerant nanoscale architecture based on linear threshold gates with redundancy

  • Aymerich Capdevila, Nivard
  • Rubio Sola, Jose Antonio
One of the main objectives of the data computing and memory industry is to keep and ever accelerate the increase of component density reached in nowadays integrated circuits in future technologies based on ultimate CMOS and new emerging research devices. The worldwide-accepted predictions with these technologies indicate a remarkable reduction of the components quality, because of the manufacturing process complexity and the erratic behavior of devices, causing a drop in the system reliability if we maintain the same design rules than today. Together with the introduction of new devices, new architectural design paradigms have to be included. Fault tolerant techniques are considered necessary and relevant in this scenario. In this paper we present a Fault-Tolerant Nanoscale architecture based on the implementation of logic systems with averaging cells linear threshold gates (AC-LTG). The sensitivity of the gates in relation with manufacturing and environment deviation is investigated and compared with the well known NAND multiplexing concept, showing that the AC-LTG is a valuable alternative in specific nanoscale conditions., Peer Reviewed, Postprint (published version)