Industry-Standard Benchmarks for Embedded Systems
EEMBC, an industry alliance, develops benchmarks to help system designers select the optimal processors and understand the performance and energy characteristics of their systems. EEMBC has benchmark suites targeting cloud and big data, mobile devices (for phones and tablets), networking, ultra-low power microcontrollers, the Internet of Things (IoT), digital media, automotive, and other application areas. EEMBC also has benchmarks for general-purpose performance analysis including CoreMark, MultiBench (multicore), and FPMark (floating-point).

GrinderBench

GrinderBench Scores

Grinderbench is no longer supported. It has been replaced by AndEBench for Android.

 

  Chess Crypto kXML Parallel PNG GrinderMark™
Android prototype 35349  54228  45348  74696  45926 49534.0
Audiovox SMT 5600 2117  9864  3233  4235  5002 4276.4
Benq-Siemens E71 4876  8900  4872  8896  7579 6773.2
HTC Hero 3030  1230  2030  3030  2030 2155.5
HTC Touch Diamond 2119  2018  3536  2102  1899 2270.6
HTC Touch Diamond 16341  32381  28416  20707  28829 24581.9
iMate SPL 6949  10037  10952  8320  7540 8631.8
Motorola Atrix 4G 33900  43776  35736  28349  37223 35446.9
Motorola E680i 13638  23128  12766  8382  19835 14626.6
Motorola Razr V3 159  113  115  124  153 131.4
Motorola V180E 150  135  141  146  141 142.5
Motorola V600 147  141  134  142  139 140.5
Nokia 5120B 123  129  129  118  103 120.0
Nokia 6230 1297  1757  1118  1611  1608 1458.5
Nokia 6600 3304  5641  2074  4960  4301 3830.5
Nokia 6610I 129  131  149  122  105 126.4
Nokia 6820B 130  129  145  121  104 125.1
Nokia 6822 128  130  144  120  109 125.7
Nokia N-gage QD 540  1084  551  720  954 739.8
Orange SPV C600 2978  10067  4206  4528  5867 5069.9
Palm Treo 650 1310  2237  1863  1704  1477 1688.9
Palm Zire72 1466  2220  2074  1746  1604 1800.2
Samsung Blackjack 2 9770  12293  10261  12124  9503 10726.3
Samsung Gts-3850 4896  4260  4686  2348  2313 3507.4
Samsung Omnia i900 22001  37892  34840  25813  34116 30309.1
Samsung s8300 1064  14636  13369  12795  10591 7764.0
Samsung SGH-Z500V 2564  3251  3239  1509  2956 2607.0
Samsung SGHX426 33  30  35  33  27 31.5
Sanyo RL4920 53  58  57  69  60 59.2
Sharp GX-31 257  282  226  267  301 265.4
Siemens S65 753  1646  1987  1551  2855 1612.6
Sony Ericcson T616 78  67  48  92  68 69.0
Sony Ericsson k750i 1924  2731  2215  4946  2237 2642.2
Sony Ericsson K750i 1956  2790  2222  4983  2233 2667.0
Sony Ericsson K750i 1890  2705  2164  4818  2137 2578.2
Sony Ericsson P910i 4598  9234  2997  7777  6202 5722.7
Sony Ericsson S700i 1116  2056  930  4238  1815 1750.0
Sony Ericsson V600i 962  1243  1299  1200  1130 1160.7
Sony Ericsson W550i 3341  5741  6772  4484  4671 4863.3
Sony Ericsson W550i 3359  5793  6803  4559  4734 4911.1
Toshiba 902T, TS803 4624  6151  3663  6753  6268 5356.6

About the Grindermark™
The Grindermark™ is a single-number score that EEMBC provides, in addition to scores based on individual benchmark applications within its GrinderBench suite, to enhance the presentation of comparative data on Java platform performance. The bigger the Grindermark score, the better. The Grindermark is the geometric mean of the five benchmark tests included in the GrinderBench suite. Grindermark numbers are intended to provide a first-order representation of Java platform performance in tasks related to Java applications. The detailed scores on individual benchmarks will continue to offer the highest value to system designers, allowing comparison of the individual applications that are specific to their designs.

Note: The GrinderBench scores appearing on this page and elsewhere on grinderbench.com are not directly comparable with the EEMBC benchmark scores published in March 2004 for the Sun Microsystems Connected Limited Device Configuration (CLDC) Reference Implementation and CLDC HotSpot Implementation nor with any other scores that were generated with EEMBC's first-generation mobile Java benchmarks. EEMBC's first-generation mobile Java benchmarks used a benchmarking methodology which is substantially different from the methodology used in the current GrinderBench benchmarks.