The upgraded Jaguar system at Oak Ridge National Laboratory took the No. 1 spot from Roadrunner with 1.75 PF linpack performance.
Roadrunner is now No. 2 with reduced linpack performance of 1.04 PF (1.105 PF in June) after repartitioning of the system.
The upgraded Kraken system at the National Institute for Computational Sciences/University of Tennessee is then second Cray XT5 in the TOP3 with 832 TFlop/s.
The most powerful system outside the U.S. is an IBM BlueGene/P system at the German Forschungszentrum Juelich (FZJ) at No. 4
The new Tianhe-1 system installed at the National Super Computer Center in Tianjin, China is #5 and the highest ranked Chinese system ever.
Intel dominates the high-end processor market with 80.4 percent of all systems and 76 percent of quad-core based systems.
Intel Core i7 (Nehalem-EP) increases its presence in the list with 95 systems compared with 33 in the last list.
Quad-core processors are used in 85.4 percent of the systems. Their use accelerates performance growth at all levels.
Other notable systems are:Hewlett-Packard kept a narrow lead in market share by total systems from IBM, but IBM still stays ahead by overall installed performance.
After a close race last November when Roadrunner was almost surpassed by the second petaflop/s system ever - the Jaguar system installed at the DOE’s Oak Ridge National Laboratory, Jaguar now comes back with new processors bringing the total peak capability of that machine to 2.3 petaflop/s and nearly a quarter of a million cores. Jaguar is now at No. 1 with 1.75 petaflop/s.
The Roadrunner system at DOE’s Los Alamos National Laboratory (LANL) was built by IBM and in June 2008 was the first system ever to break the petaflop/s Linpack barrier. Roadrunner is now at No. 2 with 1.04 petaflop/s, downgraded from 1.105 petaflop/s after it was re-partitioned. Roadrunner is based on the IBM QS22 blades which are built with advanced versions of the processor in the Sony PlayStation 3. These nodes are connected with a commodity InfiniBand network.
The TOP10 features 2 new systems and some upgrades (Jaguar) and downgrades (Roadrunner)
In the TOP10 only the No. 4 and 5 systems are installed outside the U.S. – in this case in Germany and China.
The No. 5 system called “Tianhe-I” (meaning River in Sky) was built by NUDT for NSCC-TJ. The TH-1 supercomputer system will be a key node linked into the national grid of China. The TH-1 system will be used to provide high performance computing service for the Tianjin area and the northeast of China. NSCC-TJ plans to use this system to solve the computing problems in data processing for petroleum exploration and the simulation of large aircraft designs.
The No. 10 system called "Red Sky" is a new Sun Blade system, installed at the Sandia National Lab with a Linpack performance of 423 Tflop/s.
Quad-core processor based systems have taken over the TOP500 quite rapidly. Already 426 systems are using them. 59 systems are using dual-core processors, and only four systems still use single core processors. Already six systems use IBMs advanced Sony PlayStation 3 processor with 9 cores and three systems are using the new six-core Shanghai AMD Opteron processors. The Linpack benchmark can utilize multi-core processors very well, which led to performance levels increasing above average across the whole list.
The entry level to the list moved up to the 20 Tflop/s mark on the Linpack benchmark, compared to 17.1 Tflop/s six months ago.
The last system on the newest list would have been listed at position 336 in the previous TOP500 just six months ago. This turnover rate is just below average after the TOP500 recorded the highest turnover in its history one year ago.
Total combined performance of all 500 systems has grown to 27.6 Pflop/s, compared to 22.6 Pflop/s six months ago and 16.9 Pflop/s one year ago.
The entry point for the top 100 increased in six months from 39.58 Tflop/s to 47.72 Tflop/s.
The average concurrency level in the TOP500 is 9,174 cores per system up from 8,210 six month ago and 6,240 one year ago.
A total of 402 systems (80.4 percent) are now using Intel processors. This is slightly up from six months ago (399 systems, 79.8 percent). Intel continues to provide the processors for the largest share of TOP500 systems.
The IBM Power processors are the second most common used processor family with 52 systems (10.4 percent), down from 55.
They are followed by the AMD Opteron family with 42 systems (8.4 percent), down from 43.
Multi-core processors are the dominant chip architecture. The most impressive growth showed the number of systems using the Intel Harpertown, Clovertown, and Nehalem-EP quad core chips, which grew from 336 last November to 336 systems in June and now 380 systems with the addition of the Nehalem-EP processor to Intel’s quad core lineup.
The majority of remaining systems uses dual-core processors.
417 systems are labeled as clusters, making this the most common architecture in the TOP500 with a stable share of 83.4 percent.
Gigabit Ethernet is still the most-used internal system interconnect technology (259 systems), due to its widespread use at industrial customers, followed by InfiniBand technology with 181 systems.
IBM and Hewlett-Packard continue to sell the bulk of systems at all performance levels of the TOP500.
HP kept a narrow lead in systems with 210 systems (42 percent) over IBM with 185 systems (37 percent). HP had 212 systems (42.4 percent) six months ago, compared to IBM with 188 systems (37.6 percent).
IBM remains the clear leader in the TOP500 list in performance with 34.8 percent of installed total performance (down from 39.4 percent), compared to HP with 22.8 percent (down from 25.1 percent).
In the system category, Cray, SGI, and Dell follow with 3.8 percent, 3.8 percent and 3.2 percent respectively.
In the performance category, the manufacturers with more than 5 percent are: Cray (15.9 percent of performance) and SGI (6.6 percent), each of which benefits from large systems in the TOP10.
HP (192) and IBM (114) sold together 306 out of 313 systems at commercial and industrial customers and have had this important market segment clearly cornered for some time now.
The U.S. is clearly the leading consumer of HPC systems with 277 of the 500 systems (down from 291). The European share (153 systems – up from 145) is still substantially larger then the Asian share (50 systems – up from 49).
Dominant countries in Asia are China with 21 systems (unchanged), Japan with 16 systems (up from 15), and India with 3 systems (down from 6).
In Europe, UK remains the No. 1 with 45 systems (44 six months ago). Germany and France share the No. 2 spot with 2 systems 27 systems each.
The entry level into the TOP50 is at 94.7 Tflop/s
The U.S. has a lower percentage of systems (44 percent) in the TOP50 than in the TOP500 (55.4 percent).
The dominant architectures are custom-built massively parallel systems MPPs with 60 percent ahead of commodity clusters with 40 percent.
IBM leads the TOP50 with 30 percent of systems and 33.6 percent of performance.
No 2 is Cray with a share of 18 percent of systems and 29.4 percent of performance.
SGI is third with 16 percent of systems and 10.5 percent of performance.
HP has 10 percent of systems and 4.0 percent of performance..
There is only a single system using Gigabit Ethernet in the TOP50.
Cray’s XT is the most-used system family with 9 systems (18 percent), followed by IBM’s BlueGene with 8 systems (16 percent).
Intel processors are used in 38 percent of systems, AMD processors in 30 percent and IBM’s Power processors in 28 percent.
The average concurrency level is 44,338 cores per system – up from 40,871 cores per system six month ago and 30,490 one year ago.
All changes are from June 2009 to November 2009.
The TOP500 list is compiled by Hans Meuer of the University of Mannheim, Germany; Erich Strohmaier and Horst Simon of NERSC/Lawrence Berkeley National Laboratory; and Jack Dongarra of the University of Tennessee, Knoxville.
|1|| Tianhe-2 (MilkyWay-2) - TH-IVB-FEP Cluster, Intel Xeon E5-2692 12C 2.200GHz, TH Express-2, Intel Xeon Phi 31S1P
|2|| Titan - Cray XK7 , Opteron 6274 16C 2.200GHz, Cray Gemini interconnect, NVIDIA K20x
|3|| Sequoia - BlueGene/Q, Power BQC 16C 1.60 GHz, Custom
|4||K computer, SPARC64 VIIIfx 2.0GHz, Tofu interconnect
|5|| Mira - BlueGene/Q, Power BQC 16C 1.60GHz, Custom
|6|| Stampede - PowerEdge C8220, Xeon E5-2680 8C 2.700GHz, Infiniband FDR, Intel Xeon Phi SE10P
|7|| JUQUEEN - BlueGene/Q, Power BQC 16C 1.600GHz, Custom Interconnect
|8|| Vulcan - BlueGene/Q, Power BQC 16C 1.600GHz, Custom Interconnect
|9|| SuperMUC - iDataPlex DX360M4, Xeon E5-2680 8C 2.70GHz, Infiniband FDR
|10|| Tianhe-1A - NUDT YH MPP, Xeon X5670 6C 2.93 GHz, NVIDIA 2050