64b RISC-V Code Size (RV64GC vs. ARM Cortex M0） linux-2.4.23-pre3-testplatform arch/ testplatform/kernel/init_task.c 192% =19948:10372 Initialize task structure

lwip-0.5.3.preproc src/core/netif.c 232% = 982 : 424 add TCP/IP network interface to LwiP net interfaces

teem-1.6.0-src src/ell/mat 38% = 9502 : 25024 Linear algebra using single and double float

mpeg2dec-0.3.1: libvo/yuv2rgb.c

20% = 847 : 4204

Convert YUB to RGB

CSiBE = Code Size Benchmark URL: http://szeged.github.io/csibe/ Compiled with GCC

2016/11/24

(c) SHC Sensi2ve Material Rev 3.1

Average = 115%

1

64b RISC-V Code Size (RV64GC vs. ARM Cortex M4） linux-2.4.23-pre3-testplatform arch/ testplatform/kernel/s390_ext.c 237.22% =1670:704 Initialize task structure

mpeg2dec-0.3.1: libvo/yuv2rgb.c

lwip-0.5.3.preproc src/core/netif.c 416.10% = 982:236 add TCP/IP network interface to LwiP net interfaces

teem-1.6.0-src src/ell/mat.c 38.57% = 9502:24636 Linear algebra using single and double float

221.63% = 847:3916 Convert YUB to RGB

CSiBE = Code Size Benchmark URL: http://szeged.github.io/csibe/ Compiled with GCC

2016/11/24

(c) SHC Sensi2ve Material Rev 3.1

Average = 119%

2

32b RISC-V Code Size (RV32GC vs. ARM Cortex M4） jpeg-6b wrgif.c 196% = 3560:1816 write output images in GIF format

lwip-0.5.3.preproc src/core/netif.c 247% = 584:236 add TCP/IP network interface to LwiP net interfaces

teem-1.6.0-src src/ell/mat.c 35% = 8624:24636 Linear algebra using single and double float

jpeg-6b jfdctflt.c

56% = 458:816

a floating-point implementation of the forward DCT

CSiBE = Code Size Benchmark URL: http://szeged.github.io/csibe/ Compiled with GCC

2016/11/24

(c) SHC Sensi2ve Material Rev 3.1

Average = 115%

3

32b RISC-V Code Size (RV32GC vs. ARM Cortex M0） jpeg-6b wrgif.c 176% = 3560:2028 write output images in GIF format

jpeg-6b jfdctflt.c 58% = 458:788 a floating-point implementation of the forward DCT

lwip-0.5.3.preproc src/core/netif.c 184% = 236:128 add TCP/IP network interface to LwiP net interfaces.

teem-1.6.0-src src/ell/mat.c 35% = 8624:24636 Linear algebra using single and double float

CSiBE = Code Size Benchmark URL: http://szeged.github.io/csibe/ Compiled with GCC

2016/11/24

(c) SHC Sensi2ve Material Rev 3.1

Average = 111%

4

RISC-V GCC コード効率(SH4比）

SH4 GCC AMS = Google Summer of Coding (Addressing Mode Selection). 2016/11/24

(c) SHC Sensi2ve Material Rev 3.1

5