HPC-related projects

2003

Personal cluster UCUBE
4 nodes 2 CPU(AMD) each, Dolphinics SCI — design, engineering, implementation

2004

SKIT-1 cluster Institute of Cybernetics of Ukraine
24 nodes
2xP4 CPU InfiniBand, 180/255GFlops,
Lustre 16TB)
17th in Top50.ru

design, engineering, implementation, tuning, administration
SKIT-2 cluster installed in the Ukrainian Institute of Cybernetics
32 nodes, 2xItanium-2 CPU, Dolphinics SCI, 278/358GFlops, Lustre 16TB, 8th in Top50.ru — design, engineering, implementation, tuning, administration

2006

SKIT-3 cluster in Ukrainian Institute of Cybernetics
74 node, 2xi5160 CPU, Infiniband, 2.223/2.976TFlops, Lustre 40TB) 5th in Top50.ru — design, engineering, implementation, tuning, sysOS, administration
KPI cluster in Kiev Polytechnical Institute
84 nodes, 2xi5160, InfiniBand, 3.130/4.032TFlops, Lustre 40TB) top50.ru — 6 place design, installation, tuning, sysOS, training of administrators

2007

Successful release of Cluster Automatic Emergency Shutdown System 1.0 (AESS).
Optimized linux kernel for compute node & CNL generation infrastructure(CoLoS)
Cluster-wide monitoring system (SiMon)
Cluster administration portal.
Computing cluster for Ukrainian Goverment
24 nodes, 2xi5160,
InfiniBand, Lustre 32TB,

Design, installation, tuning, training of administrators
Upgrading of SKIT-3+ cluster in Ukrainian Institute of Cybernetics

125 nodes, 2xi5160,
InfiniBand, 5.317/7.354TFlops, Lustre 120TB
9th in Top50.ru

Design, engineering, implementation, tuning, administration, integration into CERN GRID & UA.GRID programs. Software HPC Stack: MS AESS + ColoS + SiMon+ Administration Portal
Upgrade of KPI+ cluster in Kiev Polytechnical Institute
112 nodes, 2xE5440,
InfiniBand, 5.667/7.249TFlops, Lustre 60TB
16th in Top50.ru

Design, installation, tuning, training of administrators, integration into CERN GRID & UA.GRID programs

2008

Cluster Automatic Emergency Shutdown System (AESS/ClusterSafe) 2.0 released
Now cluster-wide real-time monitoring system integrated with AESS (Argos).
Cluster for National Academy of Sciences of Ukraine
16 nodes, 2xE5472
Design, installation works, tuning, training of administrators, integration into CERN GRID & UA.GRID programs
In co-operation with T-Platforms:
Cluster in Tomsk University
288 nodes, 2xX5120,
InfiniBand, PANASAS, 9/12TFlops
System OS tuning, AESS/ClusterSafe
1st in Top50.ru, 199th in Top500.org
Cluster “Chebishev/T60” in Moscow State University
Monitoring and safety solutions: MS Argos, MS AESS/ClusterSafe 1st in Top50.ru, 36th in Top500.org
Cluster in South Ural State University
166 nodes, 2xE5470, InfiniBand, PANASAS, 12.360/15.936TFlops)
OS tuning, MS AESS/ClusterSafe
8th in Top50.ru
VLSU cluster
64 nodes 2xE5345; 6 nodes 2xCELL/PPC, PANASAS, 3.756/4.771TFlops,
Configuration, system software tuning, MS AESS/ClusterSafe
16th in Top50.ru

2009

First vendor specific version of Clustrx - Clustrx TP Edition
Additional support of hardware specific introduced by T-Blade lines of T-Platforms, Russia
Cluster in National Academy of Sciences of Belorussia,
Minsk 52 nodes, 2xE5472; 16 nodes, 2xCELL/PPC, PANASAS, 4.027/4.992TFlops
Configuration, system software tuning, MS AESS/ClusterSafe
26th in Top50.ru
Moscow State University (MSU), “Lomonosov” T500
4420 nodes, 2xX5570, 350/414TFlops, InfiniBand, Lustre 2PB)
1st in Top50.ru, 12th in Top500.org

Storage installation, system installation, tuning, training of administrators
Runs HPC OS Clustrx TP Edition v1.0 (with additional support of global interrupt network, global timer network, barrier synchronisation network)

2010

Engineering and developing the software software solution for HOPSA project (HOlistic Performance System Analysis) in collaboration with Moscow State University & Jülich Supercomputing Centre
Continues in 2011
Upgrade of MSU's “Lomonosov” T500
5100 nodes, 2xX5570; 2xX5670; 397/510TFlops, InfiniBand, Lustre 8PB)
MS Clustrx TP Edition,
Storage installation, installation works, tuning, training of administrators
1st in Top50.ru

2011

Cluster in Tumen State University
164 nodes, 984 cores; 11.5TFlops, Infiniband
MS AESS/ClusterSafe installation and tuning
40th in Top50.ru
Cluster in Central Institute of Scientific Research "KURS"
Hybrid architecture (CPU+GPU), Nvidia Tesla S2050, Infinibad
MS AESS/ClusterSafe installation and tuning
Institute of Mathematical Research of Complex Systems in Moscow State University
12 nodes, 1.3 TFlop, Infiniabnd, 208GB RAM
MS AESS/ClusterSafe installation and tuning