My tweets

• Fri, 12:55: RT @jangray: #FPGA FPGAs in Datacenter? HLS based on C? Naive code yields low perf. Need many pragmas, directives, arch knowledge.
• Fri, 12:57: RT @jangray: #FPGA (BTW MS parallel computing mission: make it easier for mainstream devs to write programs that scale, correctly on parall…
• Fri, 12:58: RT @DavidKenner: Syria tragedy lamented when Syrians reach Europe. Anti-ISIS war launched when ISIS beheads Westerners. Syrians dying in Sy…
• Fri, 12:59: RT @jangray: #FPGA ..., non-von Neumann, and simple to (re)program directly from Bing language model => small compiler-HW semantic gap
• Fri, 12:59: RT @jangray: #FPGA Alternative View of Microsoft Catapult: Catapult uses programmable accelerators -- soft cores and soft FSMs, massively p…
• Fri, 12:59: RT @jangray: #FPGA "Can RC be made as easy as GPU programming?"
• Fri, 12:59: RT @jangray: #FPGA "Can high-level *programming languages* be compiled down to FPGAs?"
• Fri, 13:00: RT @jangray: #FPGA "Can SDEs program FPGAs without learning HW design or getting an EE degree?"
• Fri, 13:00: RT @jangray: #FPGA #FPL2015 FIrst speaker, my old colleague Jim Larus of EPFL, formerly of MSR and Wisconsin: "Catapult the Masses"
• Fri, 13:01: Re all these FPGA retweets; heads up, is going to be really important in the next couple of years for cloud performance and your data centre

( Read more... )