MaybeStef Profile Banner
Stefano Markidis Profile
Stefano Markidis

@MaybeStef

Followers
33
Following
85
Media
3
Statuses
15

Joined September 2016
Don't wanna be here? Send us removal request.
@MaybeStef
Stefano Markidis
3 years
Beautiful work by the KTH ScaLab Daniel Medeiros, Gabin Schieffer, and Ivy Peng on pushing molecular docking workload to converged computing systems (HPC + Cloud )with Kubernetes & Airflow. @OpenCUBE_EU @KTHuniversity #ConvergedComputing #MolecularDocking
0
1
1
@Plasma_PEPSC_EU
Plasma-PEPSC - EuroHPC CoE for Plasma Simulations
3 years
ISC BOF 2023: @Plasma_PEPSC_EU discussion focused on addressing all limitations and gaps present, while identifying priorities towards ... "Pushing Flagship Plasma Simulation Codes to Tackle Exascale-Enabled Grand Challenges via Performance Optimisation and Codesign" @ISChpc
0
2
6
@MaybeStef
Stefano Markidis
3 years
An amazing achievement, ScaLab Team, and @OpenCUBE_EU! @KTHuniversity
0
0
5
@OpenCUBE_EU
OpenCUBE
3 years
OpenCUBE at the EU concertation and consultation event #EUCEIevent @EU_CloudEdgeIoT
0
4
4
@Plasma_PEPSC_EU
Plasma-PEPSC - EuroHPC CoE for Plasma Simulations
3 years
More pictures from EuroHPC Summit 2023: @Plasma_PEPSC_EU Project Coordinator and Associate Professor Stefano Markidis at the Parallel Session focusing on Space Physics and Astrophysics Codes.
0
1
2
@EPFL_en
EPFL
4 years
Scientists at EPFL’s Swiss Plasma Center & @DeepMind have developed a new magnetic control method for plasmas based on deep reinforcement learning and applied it to a real-world plasma for the 1st time in the @TCV_tokamak research facility. #nuclearfusion https://t.co/VStZPEbfUP
Tweet card summary image
actu.epfl.ch
Scientists at EPFL’s Swiss Plasma Center and DeepMind have jointly developed a new method for controlling plasma configurations for use in nuclear fusion research.
4
51
230
@MaybeStef
Stefano Markidis
8 years
yesterday at the Sage and NextGenIO workshop in Ljubljana! Still jet-lagged 😴 but many great talks about heterogenous memories and multi-tier storage πŸ˜€
@exdci_eu
EXDCI
8 years
"Parallel I/O is one of the most serious performance bottlenecks in HPC applications" : the ground motivation at the forefront of Stefano Markilis (from @KTHresearch & @SageStorage) talk on exploring scientific application performance using large scale object storage. #EHPCSW
1
0
4
@MaybeStef
Stefano Markidis
8 years
πŸ₯‡Best paper on programmability, performance and precision loss of #Nvidia tensor cores at #IPDPS #ASHES18( https://t.co/NZlkWVXT1y)!!! Celebrating in Vancouver πŸ‡¨πŸ‡¦!!!
0
1
5
@MaybeStef
Stefano Markidis
8 years
Finally back to Serie A! What a magic night!
@1913parmacalcio
𝐏𝐚𝐫𝐦𝐚 π‚πšπ₯𝐜𝐒𝐨 πŸπŸ—πŸπŸ‘
8 years
"Tre anni fa ci davano per morti: si sono sbagliati" Noi siAmo AncorA quA! πŸ’ͺπŸ»πŸ’›πŸ’™ #EhGiΓ  #comenoinessunomAi #ForzaParma
1
0
1
@MaybeStef
Stefano Markidis
8 years
Thank you Adrian
@adrianjhpc
Adrian Jackson
8 years
Stefano Markidis from @KTHinfo presenting improved #tensorflow performance with higher performance communication libraries
0
0
2
@MaybeStef
Stefano Markidis
8 years
Our paper on benchmarking #NVIDIA #Volta #GPU for #HPC shows that we can get up to 83 Tflops/s when running mixed-precision large sgemm and #NVIDIA #Tensor #Cores flops can be used to decrease loss of precision! #GTC18 https://t.co/RnXlYVMPbV
0
1
1
@SageStorage
Sage Storage
8 years
SAGE team at the ISC poster session
0
1
6