Using Endurance Predictive Software Using the Microchip Endurance Predictive Software INTRODUCTION TOTAL ENDURANCE PREDICTIVE SOFTWARE
Endurance, as it applies to non-volatile memory, refersto the number of times an individual memory cell can be
The predictive software described here originally was
erased and/or written (some architectures do not erase
being developed as a tool for determining endurance
before every write). Advances in process technology
levels of Microchip non-volatile devices. Upon seeing
have made it possible to increase these limits and for
the potential of the software as a design aid, it was
Microchip to offer new concepts - Total Endurance™ and
decided to make a version that could be purchased by
a split architectural design for variable endurance. These
the engineering community. The benefit gained from this
concepts lead to more reliable products with more bits
software is the ability to predict endurance capability of
per dice, such as the 24C32 and 24C65.
Microchip’s EEPROM devices under various operatingconditions. Prior to this tool becoming available, the only
TOTAL ENDURANCE ™
way to assemble this type of data would be to doextensive life testing in the target system. It should be
When defining endurance, we need to look at a few
noted that this predictive model applies only to Microchip
common definitions and possible misconceptions. En-
Technology Inc. non-volatile devices.
durance with respect to EEPROMs is defined in numberof Erase/Write (E/W) Cycles and is the most common
The program uses an iterative statistical model devel-
rating referred to when discussing or specifying endur-
oped by Microchip Technology Inc. physicists. The
ance. E/W ratings are based on the environmental and
model was first used in a DOS-based text program as a
operating conditions of voltage, temperature, cycling
proof of concept and for developing the exhaustive
mode and rate (for each byte in the application not on the
database needed for such a tool (included on the pro-
number of opcodes or control byte commands) and is
gram disk as enddos.exe). This model was then im-
never based on any read functions whether they be a
ported to a Windows™-based software package with full
data read or configuration read. If a part is rated at 100K
GUI capabilities and all the normal cut, paste, print,
E/W cycles, then each individual byte can be erased and
viewing properties. The model actually operates as a
written 100,000 times. This is probably the most com-
mathematical function which is called from within the
mon misinterpretation made by system designers. En-
Windows Visual Basic shell and is passed all of the
durance is thus an interactive application-specific reli-
pertinent operational, process, and device information.
ability parameter. It is not a typical data sheet specifica-
The model then, after calculating the essential data
tion, such as a parametric AC/DC specification with
points, returns this information to the main program to be
benchmark standards for measurement. Microchip
formatted and displayed both textually and graphically.
has done extensive predictive laboratory studies on
Applying the predictive data to the high endurance block
Microchip 2- and 3-wire Serial EEPROMs. These stud-
of the 24C65, using the 24LC04 which has similar
ies led to the concept of using the computer to predict the
characteristics, and assuming the following:
theoretical wear out of the floating gate and ultimately to
project the point in time of a product’s life cycle when the
• an expected E/W cycles of 10 times per day
first non-volatile memory bit or periphery failure should
occur. After many man-years of data collecting, predict-ing and verifying the results, Microchip feels confident inpublishing and offering for the general technical commu-nity this predictive model in the form of IBM® PC-compatible software. Microchip has a patent pending onthis predictive mathematical model. Using Endurance Predictive Software Operational specifications:
endurance of a particular EEPROM device within a setof application parameters. Trade-off analysis can be
painfully time-consuming and only marginally accurate
without specific knowledge of the behavior of the deviceunder different conditions of use.
The Microchip Total Endurance Software allows the
designer to trade off voltage, temperature, write cycles,
number of bytes written, number of writes per day, PPMand FIT rates, and years of use in order to optimize the
system and accurately predict product lifetime and reli-
The following is an example using the Endurance Soft-
The 4K HE block with 1 M E/W cycles typical, in this
ware to aid in the design of an electronic phone book/
application, should yield the following results:
The auto-dialer may have new numbers added orchanged several times per day; but how can the
manufacturer specify the life of the unit, and at whatrate of update of the phone numbers? First, thedesigner must make some assumptions. If we as-
sume that the average user will change or add 50phone numbers per day, and the manufacturer iswilling to live with a 0.1% failure rate (1,000 PPM) after
24LC04, 25C, 5V, 11bytes, 10cycles/day, RANDOM, BYTE
10 years of use, then we have almost enough informa-tion to verify whether we are in the ball park given the
physics of the EEPROM device which will store thenumbers. We also need to know the operating voltage
and temperature of the application; we will say that a3.3V lithium button battery is powering the unit and the
temperature range is limited to that for which the LCDdisplay will function: 0°C to 70°C. End-of-life voltage
for the battery is approximately 2.0V; assuming thatthe ASIC or microcontroller in the application willoperate down to 2.5V, the EEPROM also has a 2.5V
requirement. The designer would like to be able tostore 100 phone numbers of 16 bytes each, which
results in a 1.6K byte requirement for the SerialEEPROM. Because 1.6K bytes is equal to 12.8K bits,a 16K bit 2-wire Serial EEPROM will more than
The results shown are predictive in nature and should
suffice. Specifically, Microchip’s 24LC16B will oper-
reflect an accurate representation of the expected re-
ate down to 2.5V and even includes a write-protect
sults. For a more detailed description of endurance, see
feature which can be used to block inadvertent writes
the related application notes AN536 and AN537 con-
tained elsewhere in this volume. All operation param-eters, along with the process technology, effect the
effective endurance of a non-volatile device. The volt-
age, temperature, cycles per, bytes per cycle, and eventhe number of times written per day (time between write
cycles) all have an effect on the oxide breakdown or
periphery failure rate of a particular non-volatile process.
Endurance is not a well-defined concept within the
semiconductor industry. The number of erase/writecycles which a particular EEPROM can endure is de-
pendent not only upon the design of the device but alsoupon the application environment in which it is used. Therefore, blanket claims such as “1 million erase/writecycles typical” can only be validated based upon thespecific parameters of each application. Yet until now,there has been no tool available for predicting the
Using Endurance Predictive Software
Once these values are entered into the Total Endurance
for the application life of the product given a 1000 PPM
Now we have some more options: (1) specify the
Both of the lists above were copied directly from the
product life at 5 years or (2) trade-off other parameters
Total Endurance program output to the Microsoft® Win-
of the application such as voltage or temperature, or (3)
dows clipboard and pasted into this document (the Total
decide which is more important – a 10-year product
Endurance program has a handy menu click to make this
lifetime, or the ability to change 50 numbers every single
day. Maybe this analysis has caused our designer to re-
Unfortunately for our designer, the desired 0.1% failure
evaluate the 50 cycle-per-day requirement. Will the user
rate has almost doubled to 0.18% (1842 PPM). But
really change or add that many numbers per day – half
fortunately for the designer, the Total Endurance pro-
of the unit’s total capacity? Maybe 20 or even 10 is a
gram makes trade-off analysis very simple and fast. At
more practical figure. Realistically, a user may enter or
this point there are at least three options: (1) live with
change quite a few numbers the first week or two of the
almost 2000 PPM, or (2) look at the endurance plot and
application, and after that the unit will be used mostly for
check whether there is a reasonable number of E/W
cycles which will provide a 1000 PPM failure rate, or (3)
Changing the number of erase/write cycles to 20 per day
specify a PPM rate to the Total Endurance program and
let it crank out the number of cycles it will take.
Below is the endurance plot, again pasted directly from
You can see that by reducing the number of cycles fromthe 182,500 which resulted from our first trial to about
100,000, we can achieve a PPM rate of about 1000
(0.1%). But how does 100,000 cycles translate intoapplication life or cycles per day?
By switching the Total Endurance program mode to a
PPM request mode instead of application life mode, we
can query the program for this information. Let’s ask it
Using Endurance Predictive Software
The new PPM rate of 625 gives our triumphant designermore than 30% margin on his PPM target of 1000.
This example shows the significant reduction in time for
design trade-off analysis and time-to-market which can
be achieved with a useful tool like the Microchip Total
Wow! Reducing the number of cycles per day not only
Endurance Disk. In addition, it demonstrates the in-
brought us back to a 10-year life, it gave us some margin
crease in robustness of the system design by providing
on that, too. Keeping all the other parameters the same
known quantities and readily accessible handles to
and forcing a 10-year lifetime gives us the following final
modify those quantities in the trade-off analysis. This
tool can literally reduce weeks of effort into a few minutesof point and click.
Richard J. FisherMemory Products Division
WORLDWIDE SALES & SERVICE AMERICAS AMERICAS (continued) Corporate Office United Kingdom ASIA/PACIFIC
Tel: 44 0 1628 851077 Fax: 44 0 1628 850259
Web: http://www.mchip.com/microhip
Hong Kong
Tel: 33 1 69 53 63 20 Fax: 33 1 69 30 90 79
Arizona Microchip Technology GmbHGustav-Heinemann-Ring 125
Tel: 49 89 627 144 0 Fax: 49 89 627 144 44
Centro Direzionale ColleoniPalazzo Pegaso Ingresso No. 2
Singapore
Tel: 39 039 689 9939 Fax: 39 039 689 9883
Los Angeles Microchip Technology Inc. 18201 Von Karman, Suite 455 Irvine, CA 92715 Tel: 714 263-1888 Fax: 714 263-1338 New York Microchip Technology Inc. 150 Motor Parkway, Suite 416 Hauppauge, NY 11788 Tel: 516 273-5305 Fax: 516 273-5335
All rights reserved. 1995, Microchip Technology Incorporated, USA.
Information contained in this publication regarding device applications and the like is intended through suggestion only and may be superseded by updates. No representation or warrantyis given and no liability is assumed by Microchip Technology Incorporated with respect to the accuracy or use of such information, or infringement of patents or other intellectual propertyrights arising from such use or otherwise. Use of Microchip’s products as critical components in life support systems is not authorized except with express written approval by Microchip. No licenses are conveyed, implicitly or otherwise, under any intellectual property rights. The Microchip logo and name are registered trademarks of Microchip Technology Inc. All rightsreserved. All other trademarks mentioned herein are the property of their respective companies.
j o u r n a l o f d e n t i s t r y x x x ( 2 0 0 6 ) x x x – x x xa v a i l a b l e a t w w w . s c i e n c e d i r e c t . c o mj o u r n a l h o m e p a g e : w w w . i n t l . e l s e v i e r h e a l t h . c o m / j o u r n a l s / j d e nThe bleaching of teeth: A review of the literatureUnilever Oral Care, Quarry Road East, Bebington, Wirral, CH63 3JW, UKObjectives: To review c
International Morning News Wednesday, 14 July 2010 Market Commentary: U.S. stocks surged on Tuesday, extending the longest rally for the S&P 500 in 3 months, as improving outlook from Al- coa bolstered confidence in the economy. Alcoa gained 1.2% as sales and profit exceeded analysts’ estimates and the company forecast stronger global demand. Chevron led gains in energy companies,