In the particular case of MATMUL on a RS/6000 F50, we have achieved fully compliant Java performance in excess of 500 Mflops
4) the basic cost of signal processing, MFLOPS
per dollar, is considerably lower than that of comparable microprocessors.
It offers 9.2 MFLOPS
and 29.5 MIPS performance, memory from 8MB to 128MB, internal disk storage from 160MB to 800MB and four I/O slots.
However, the number of floating point operations required for this algorithm becomes prohibitive even at sustained flop rates of several hundred Mflops
(million flops per second) to a few Gflops (billion flops per second).
v sec Mflops
i 40.8 70 [argmax.sub.w] ([M.sub.w]) 35.2 82 [argmin.sub.w] ([T.sub.w]) 6.57 531 7.2 Example 2
Each node has a peak performance of 25 VAX MIPS and 10 double-precision MFLOPS
. Cache coherence within a cluster is maintained at the level of 16-byte lines via a bus-based snoopy protocol.
The 4D-MP graphics superworkstation: Computing + graphics = 40 mips + 40 mflops
+ 100,000 lighted polygons per second.
Early metrics for quantizing microprocessor performance include MIPS and MFLOPS
. "MIPS," million instructions per second, refers to the average number of machine language instructions that a computer can perform in 1 sec.
Each vector node can work at up to 186 MFLOPS
, and each contains 128 MB of very high-speed memory.
This code ran on the NCube at about 1.5 GFLOPS; by comparison a highly vectorized version of ZEPHYR-3D runs at about 220 MFLOPS
on a single Gray Y-MP processor.
Whetsone, Dhrystone, Linpack, SPEC benchmark suite, mflops
, and mips are all names to measure how fast a computer runs a program.
Floating point performance: 1.4 Mflops
. Main memory: 8-16 MB RAM.