Next:
Systems under development
Up:
Overview of Recent
Previous:
The Silicon Graphics Origin series.
As already stated in the introduction the list of systems is not
complete. On one hand this is caused by the sheer number of systems
that are presented to the market and are often very similar to systems
described above (for instance, the Volvox system not listed is very
similar but not equivalent to the listed C-DAC system and there are
numerous other examples). On the other hand there many systems that are
still in operation around the world, often in considerable quantities
that for other reasons are excluded. The most important reasons are:
- The system is not marketed anymore. This is generally for one of two
reasons:
- The manufacturer is out of business.
- The manufacturer has replaced the system by a newer model of the same
type or even of a different type.
- The system has become
technologically obsolete in comparison to others of the same type. Therefore,
listing them is not sensible anymore.
Below we present a table of systems that fall into one of the categories
mentioned above. We think this may have some sense to those who come across
machines that are still around but are not the latest in their fields. It may
be interesting at least to have an indication how such systems compare to the
newest ones and to place them in context.
It is good to realise that although systems have disappeared
from the section above they still may exist and are actually sold. However,
their removal stems in such cases mainly from the fact that they are not
serious candidates for high-performance computing anymore.
The table is, again, not complete and admittedly somewhat arbitrary.
The data are in a highly condensed form: the system name, system type,
theoretical maximum performance of a fully configured system, and the
reason for their disappearance is given. The arbitrariness lies partly
in the decision which systems are still sufficiently of interest to
include and which are not.
For the first time we include the year of introduction and the year of exit
of the systems when they were readily accessible. These timespans could
give a hint of the dynamics that governs this very dynamical branch of
the the computer industry.
Machine: The Alex AVX 2.
Year of introduction: 1992.
Year of exit: 1997.
Type: RISC-based distributed-memory multi-processor.
Theoretical Peak performance: 3.84 Gflop/s.
Reason for disappearance: System is obsolete, there is no
new system planned.
Machine: Alliant FX/2800.
Year of introduction: 1989.
Year of exit: 1992.
Type: Shared memory vector-parallel, max. 28 processors.
Theoretical Peak performance: 1120 Mflop/s.
Reason for disappearance: Manufacturer out of business.
Machine: The AxilSCC.
Year of introduction: 1996.
Year of exit: 1997.
Type: RISC-based distributed-memory system, max. 512 processors.
Theoretical Peak performance: 76.8 Gflop/s.
Reason for disappearance: System is not marketed anymore by Axil.
Machine: BBN TC2000.
Year of introduction: ---
Year of exit: 1990.
Type: Virtual shared memory parallel, max. 512 processors.
Theoretical Peak performance: 1 Gflop/s
Reason for disappearance: Manufacturer has discontinued marketing
parallel computer systems.
Machine: Cambridge Parallel Processing DAP Gamma.
Year of introduction: 1986.
Year of exit: 1995.
Type: Distributed memory processor array system.
Theoretical Peak performance: 1.6 Gflop/s (32-bit)
Reason for disappearance: replaced by newer
Gamma II Plus series.
Machine: C-DAC PARAM 9000/SS.
Year of introduction: 1995.
Year of exit: 1997.
Type: RISC-based distributed-memory system, max. 200 processors.
Theoretical Peak performance: 12.0 Gflop/s.
Reason for disappearance: replaced by newer
PARAM OpenFrame series.
Machine: Convex SPP-1000/1200/1600.
Year of introduction: 1995 (SPP-1000).
Year of exit: 1996 (SPP-1600).
Type: Distributed memory RISC based system, max. 128 processors.
Theoretical Peak performance: 25.6 Gflop/s
Reason for disappearance: replaced by newer
SPP-2000 series.
Machine: Cray Computer Corporation Cray-2.
Year of introduction: 1982.
Year of exit: 1992.
Type: Shared memory vector-parallel, max. 4 processors.
Theoretical Peak performance: 1.95 Gflop/s
Reason for disappearance: Manufacturer out of business.
Machine: Cray Computer Corporation Cray-3.
Year of introduction: 1993.
Year of exit: 1995.
Type: Shared memory vector-parallel, max. 16 processors.
Theoretical Peak performance: 16 Gflop/s
Reason for disappearance: Manufacturer out of business.
Machine: Cray Research Inc. APP.
Year of introduction: 1994.
Year of exit: 1996.
Type: Shared memory RISC based system, max. 84 processors.
Theoretical Peak performance: 6.7 Gflop/s
Reason for disappearance: Product line discontinued, gap expected to
be filled by
Cray J90.
Machine: Cray T3D.
Year of introduction: 1994.
Year of exit: 1996.
Type: Distributed memory RISC based system, max. 2048 processors.
Theoretical Peak performance: 307 Gflop/s
Reason for disappearance: replaced by newer
Cray T3E.
Machine: Cray Research Inc. Cray Y-MP, Cray Y-MP M90.
Year of introduction: 1989 (Cray Y-MP).
Year of exit: 1994 (Cray Y-MP M90)..
Type: Shared memory vector-parallel, max. 8 processors.
Theoretical Peak performance: 2.6 Gflop/s
Reason for disappearance: replaced by newer
T90.
Machine: Cray Y-MP C90.
Year of introduction: 1994.
Year of exit: 1996.
Type: Shared memory vector-parallel, max. 16 processors.
Theoretical Peak performance: 16 Gflop/s
Reason for disappearance: replaced by newer
T90.
Machine: Digital Equipment Corp. Alpha farm.
Year of introduction: ---
Year of exit: 1994.
Type: Distributed memory RISC based system, max. 4 processors.
Theoretical Peak performance: 0.8 Gflop/s
Reason for disappearance: replaced by newer
AlphaServer clusters.
Machine: Fujitsu AP1000.
Year of introduction: 1991.
Year of exit: 1996.
Type: Distributed memory RISC based system, max. 1024 processors.
Theoretical Peak performance: 5 Gflop/s
Reason for disappearance: replaced by the
Fujitsu AP3000.
Machine: Fujitsu VPP500 series.
Year of introduction: 1993.
Year of exit: 1995.
Type: Distributed memory multi-processor vectorprocessors, max.
222 processors.
Theoretical Peak performance: 355 Gflop/s
Reason for disappearance: replaced by the
VPP300/700 series.
Machine: Fujitsu VPX200 series.
Year of introduction: ---
Year of exit: 1995.
Type: Single-processor vectorprocessors.
Theoretical Peak performance: 5 Gflop/s
Reason for disappearance: replaced by the
VPP300/700 series.
Machine: Hitachi SR2001 series.
Year of introduction: 1994.
Year of exit: 1996.
Type: Distributed memory RISC based system, max. 128 processors.
Theoretical Peak performance: 23 Gflop/s
Reason for disappearance: Replaced by the newer
SR2201.
Machine: HP/Convex C4600.
Year of introduction: 1994.
Year of exit: 1997.
Type: Shared memory vector-parallel, max. 4 processors (C4640).
Theoretical Peak performance: 3.2 Gflop/s
Reason for disappearance: The C4600 is not marketed by HP/Convex anymore.
Machine: IBM ES/9000 series.
Year of introduction: 1991.
Year of exit: 1994.
Type: Shared memory vector-parallel system, max. 6 processors.
Theoretical Peak performance: 2.67 Gflop/s
Reason for disappearance: IBM does not pursue high-performance computing
by this product line anymore.
Machine: IBM Power/4.
Year of introduction: 1992.
Year of exit: 1994.
Type: Shared memory RISC based system, max. 4 processors.
Theoretical Peak performance: 336 Mflop/s
Reason for disappearance: Product line discontinued, gap expected to
be filled by
SP2.
Machine: IBM SP1 series.
Year of introduction: 1992.
Year of exit: 1994.
Type: Distributed memory RISC based system, max. 64 processors.
Theoretical Peak performance: 8 Gflop/s
Reason for disappearance: Replaced by the newer
SP2.
Machine: Intel iPSC/860.
Year of introduction: 1989.
Year of exit: 1992.
Type: Distributed memory RISC based system, max. 128 processors.
Theoretical Peak performance: 7.7 Gflop/s.
Reason for disappearance: was replaced by the Intel Paragon XP (MP)
series (but see below).
Machine: Intel Paragon XP.
Year of introduction: 1992.
Year of exit: 1996.
Type: Distributed memory RISC based system, max. 4000 processors.
Theoretical Peak performance: 300 Gflop/s.
Reason for disappearance: Except for a non-commercial research system
(the ASCI Option Red system at Sandia National Labs.) Intel is not in the
business of high-performance computing anymore.
Machine: Kendall Square Research KSR2.
Year of introduction: 1992.
Year of exit: 1994.
Type: Virtually shared memory parallel, max. 1088 processors.
Theoretical Peak performance: 400 Gflop/s
Reason for disappearance: Kendall Square has terminated its business.
Machine: Kongsberg Informasjonskontroll SCALI.
Year of introduction: 1996.
Year of exit: 1997.
Type: Distributed memory RISC based system, max. 512 processors.
Theoretical Peak performance: 76.8 Gflop/s
Reason for disappearance: Kongsberg does not market the system anymore.
Machine: MasPar MP-1, MP-2.
Year of introduction: 1991 (MP-1).
Year of exit: 1996.
Type: Distributed memory processor array system, max. 16384 processors.
Theoretical Peak performance: 2.4 Gflop/s (64-bit, MP-2)
Reason for disappearance: Systems are not marketed anymore.
Machine: Matsushita ADENART.
Year of introduction: 1991.
Year of exit: 1997.
Type: Distributed memory RISC based system, 256 processors.
Theoretical Peak performance: 2.56 Gflop/s.
Reason for disappearance: Machine is obsolete and no new systems
are developed in this line.
Machine: Meiko CS-1 series.
Year of introduction: 1989.
Year of exit: 1995.
Type: Distributed memory RISC based system.
Theoretical Peak performance: 80 Mflop/s per processor
Reason for disappearance: Replaced by the newer
CS-2.
Machine: nCUBE 3.
Year of introduction: --- (see remarks at
nCUBE 2S).
Year of exit: ---
Type: Distributed memory system, max. 10244 processors.
Theoretical Peak performance: 1 Tflop/s.
Reason for disappearance: Was announced several times but was
never finished. Development has been abandoned but may be revived some time.
nCUBE 2S is the available product.
Machine: NEC SX-2.
Year of introduction: 1990.
Year of exit: 1993.
Type: Single-processor vector processors.
Theoretical Peak performance: 1.3 Gflop/s
Reason for disappearance: replaced by newer
SX-4 series.
Machine: NEC SX-3R.
Year of introduction: 1993.
Year of exit: 1996.
Type: Shared memory multi-processor vector processors, max. 4 processors.
Theoretical Peak performance: 25.6 Gflop/s
Reason for disappearance: replaced by newer
SX-4 series.
Machine: Parsys SN9000 series.
Year of introduction: 1993.
Year of exit: 1995.
Type: Distributed memory RISC based system, max. 2048 processors.
Theoretical Peak performance: 51.2 Gflop/s
Reason for disappearance: Replaced by the newer TA9000 (but see below).
Machine: Parsys TA9000 series.
Year of introduction: 1995.
Year of exit: 1996.
Type: Distributed memory RISC based system, max. 512 processors.
Theoretical Peak performance: 119.3 Gflop/s
Reason for disappearance: Parsys does not offer complete system
anymore. Instead it sells node cards based on the TA9000 for embedded
systems.
Machine: Parsytec GC/Power Plus.
Year of introduction: 1993.
Year of exit: 1996.
Type: Distributed memory RISC based system.
Theoretical Peak performance: 266.6 Mflop/s per processor.
Reason for disappearance: System has been replaced by the
Parsytec CC systems.
Machine: Siemens-Nixdorf VP2600 series.
Year of introduction: ---
Year of exit: 1995.
Type: Single-processor vectorprocessors.
Theoretical Peak performance: 5 Gflop/s
Reason for disappearance: replaced by the
VPP300/700 series.
Machine: Silicon Graphics PowerChallenge.
Year of introduction: 1994.
Year of exit: 1996.
Type: Shared memory multi-processor, max. 36 processors.
Theoretical Peak performance: 14.4 Gflop/s
Reason for disappearance: replaced by the SGI
Origin 2000.
Machine: Stern Computing Systems SSP.
Year of introduction: 1994.
Year of exit: 1996.
Type: Shared memory multi-processor, max. 6 processors.
Theoretical Peak performance: 2 Gflop/s
Reason for disappearance: Vendor terminated its business just before
delivering first systems.
Machine: Thinking Machine Corporation CM-2(00).
Year of introduction: 1987.
Year of exit: 1991.
Type: SIMD parallel machine with hypercube structure, max. 64K
processors.
Theoretical Peak performance: 31 Gflop/s
Reason for disappearance: was replaced by the newer CM-5 (but see
below).
Machine: Thinking Machine Corporation CM-5.
Year of introduction: 1991.
Year of exit: 1996.
Type: Distributed memory RISC based system, max. 16K processors.
Theoretical Peak performance: 2 Tflop/s
Reason for disappearance: Thinking Machine Corporation has stopped
manufacturing hardware and hopes to keep alive as a software
vendor.
Next:
Systems under development
Up:
Overview of Recent
Previous:
The Silicon Graphics Origin series.
Aad van der Steen
Wed Feb 18 14:16:31 MET 1998