ECOSCALE (@ecoscale_h2020) 's Twitter Profile
ECOSCALE

@ecoscale_h2020

ID: 3982968016

calendar_today17-10-2015 23:21:28

261 Tweet

96 Followers

126 Following

European Commission (@eu_commission) 's Twitter Profile Photo

You have the power to shape Europe. 10 European citizens’ initiatives have already been registered in 2019! Our Forum is here to help you drive your own initiative to success. Learn more → collab.ec.europa.eu/wiki/eci/ #EUTakeTheInitiative

RISC-V International (@risc_v) 's Twitter Profile Photo

.All About Circuits spoke with Ted Marena "Trust Ted", interim director of the CHIPSAlliance, to discuss how the Alliance aims to “collaboratively create hardware and open source development tools for the masses.” buff.ly/2YNouzD

DARPA (@darpa) 's Twitter Profile Photo

An industry & program first! Under our 3DSoC program, the first monolithic 3D device structures produced in a fab were unveiled yesterday at our #ERISummit by Dr. Max Shulaker of Massachusetts Institute of Technology (MIT). go.usa.gov/xyyGJ

An industry &amp; program first! Under our 3DSoC program, the first monolithic 3D device structures produced in a fab were unveiled yesterday at our #ERISummit by Dr. Max Shulaker of <a href="/MIT/">Massachusetts Institute of Technology (MIT)</a>. 

go.usa.gov/xyyGJ
Exascale Computing Project (@exascaleproject) 's Twitter Profile Photo

ECP's Center for Efficient Exascale Discretizations helps applications leverage future architectures by developing state-of-the-art discretization algorithms that better exploit the hardware and deliver a performance gain over conventional methods. bit.ly/2xU25EY

ECP's Center for Efficient Exascale Discretizations helps applications leverage future architectures by developing state-of-the-art discretization algorithms that better exploit the hardware and deliver a performance gain over conventional methods. bit.ly/2xU25EY
Altera (@alterafpga_) 's Twitter Profile Photo

The 3-A Intel Enpirion EN6338 PowerSoC is faster, smaller, and more efficient. Learn more in this tutorial: intel.ly/2XU9puX

The 3-A Intel Enpirion EN6338 PowerSoC is faster, smaller, and more efficient. Learn more in this tutorial: intel.ly/2XU9puX
Exascale Computing Project (@exascaleproject) 's Twitter Profile Photo

ICYMI: Watch 👀 the video of the latest in the monthly #HPC Best Practices Webinar Series from the IDEAS Productivity project for ECP: Software Management Plans in Research Projects. Presenter Shoaib Ahmed Sufi of the Software Sustainability Institute bit.ly/2Z7Ehss

ICYMI: Watch 👀  the video of the latest in the monthly #HPC Best Practices Webinar Series from the IDEAS Productivity project for ECP: Software Management Plans in Research Projects. Presenter Shoaib Ahmed Sufi of the Software Sustainability Institute bit.ly/2Z7Ehss
RISC-V International (@risc_v) 's Twitter Profile Photo

.DARPA recently unveiled its first SSITH prototype that uses open source RISC-V International cores. The program aims to release 15 different prototypes ranging from low-end IoT devices to high-end systems. Check out this article at @TechTarget to learn more: buff.ly/30u3crO

Altera (@alterafpga_) 's Twitter Profile Photo

The Intel AIB is a die-to-die interconnect standard that speeds development and improves performance. Here’s how it’s promoting chiplet adoption: intel.ly/2HAbjf4

The Intel AIB is a die-to-die interconnect standard that speeds development and improves performance. Here’s how it’s promoting chiplet adoption: intel.ly/2HAbjf4
IEEE Spectrum (@ieeespectrum) 's Twitter Profile Photo

Neural nets for AI are limited by slow processing and their need for extensive computing resources. Now two groups have demoed all-optical neural networks that do all processing with light, making neural nets faster and more energy efficient. #ai buff.ly/32imilj

HPC Guru (on an extended break) (@hpc_guru) 's Twitter Profile Photo

MIT Researchers built a 16-bit carbon nanotube-based microprocessor that can execute the full RISC-V instruction set The 14,000-transistor chip executed a classic Hello World program news.mit.edu/2019/carbon-na…

MIT Researchers built a 16-bit carbon nanotube-based microprocessor that can  execute the full RISC-V instruction set

The 14,000-transistor chip executed a classic Hello World program 

news.mit.edu/2019/carbon-na…
HiPEAC (@hipeac) 's Twitter Profile Photo

Congratulations @Appentra! Parallelware Trainer aims to make parallel programming open to everyone... not just ninja programmers😊

Upscale Project (@upscale_project) 's Twitter Profile Photo

The Upscale Project faces the challenge of demonstrating that #AI can also deal with real industrial simulation models, such as full size vehicles, aerodynamics or crash to enhance the design process of #ElectricVehicles! Discover how 👉bit.ly/2Pdvu8B #WAIC2019 #ML #EVs

EU processor (@euprocessor) 's Twitter Profile Photo

August is almost over, so onto some more intricate stuff, instead of #SummerReading 😃 A chapter by EPI authors in the Proceedings of the 5th CESA, An Approach for a Future #Automotive eHPC #Semiconductor Platform Faculty of Electrical Engineering and Computing @EB_Automotive BMW Group Atos France Springer Nature

August is almost over, so onto some more intricate stuff, instead of #SummerReading 😃

A chapter by EPI authors in the Proceedings of the 5th CESA, An Approach for a Future #Automotive eHPC #Semiconductor Platform

<a href="/fer_unizg/">Faculty of Electrical Engineering and Computing</a> @EB_Automotive <a href="/BMWGroup/">BMW Group</a> <a href="/AtosFR/">Atos France</a> <a href="/SpringerNature/">Springer Nature</a>
ECOSCALE (@ecoscale_h2020) 's Twitter Profile Photo

2x Baseboards (the final ECOSCALE platform) powered up! 64 FPGAs tightly interconnected & ready to perform acceleration tasks. Each baseboard hosts 8x QFDBs, so the platform gathers 16 QFDBs, with a total of 64 Xilinx Zynq Ultrascale+ MPSoCs and 1 TByte of distributed DDR memory!

2x Baseboards (the final ECOSCALE platform) powered up! 64 FPGAs tightly interconnected &amp; ready to perform acceleration tasks. Each baseboard hosts 8x QFDBs, so the platform gathers 16 QFDBs, with a total of 64 Xilinx Zynq Ultrascale+ MPSoCs and 1 TByte of distributed DDR memory!