Mobile Power Guidelines ’99
`
`Intel Corporation
`
`Revision 1.00
`
`December 1, 1997
`
`MICROCHIP TECH. INC. - EXHIBIT 1032
`MICROCHIP TECH. INC. V. HD SILICON SOLS. - IPR2021-01265 - Page 001
`
`

`

`Mobile Power Guidelines Rev. 1.00
`
`ACKNOWLEDGEMENTS
`
`We would like to extend special recognition to the following companies for their early and extensive participation in the development
`of the Mobile Power Guidelines: IBM Corporation, Toshiba Corporation, Compaq Corporation, Dell Corporation and NEC
`Corporation.
`
`We would also like to acknowledge the following companies for their contributions to this document:
`Analog Devices, Inc
`Chips and Technologies, Inc
`Cirrus Logic, Inc
`Compaq Corporation
`Dell Corporation
`ESS Technology, Inc
`Hitachi Corporation
`IBM Corporation
`Linear Technology Corporation
`Mfg Components, Inc
`Maxim Integrated Products
`NEC Corporation.
`NeoMagic Corporation
`Rambus Corporation
`S3, Inc
`Sharp Corporation
`Siliconix/Temic Semiconductors
`Toshiba Corporation
`Trident Microsystems, Inc
`
`December 1, 1997
`
`2
`
`MICROCHIP TECH. INC. - EXHIBIT 1032
`MICROCHIP TECH. INC. V. HD SILICON SOLS. - IPR2021-01265 - Page 002
`
`

`

`Mobile Power Guidelines Rev. 1.00
`
`Table of Contents
`
`1. Executive Summary...........................................................................................................................................................................4
`
`2. Introduction .......................................................................................................................................................................................5
`
`3. Notebook System Power Trends.......................................................................................................................................................6
`
`4. Power Reduction Methods................................................................................................................................................................9
`
`5. System Feature and Power Targets ...............................................................................................................................................11
`
`6. Battery Life ......................................................................................................................................................................................29
`
`7. Summary ..........................................................................................................................................................................................33
`
`Appendix A. Power Consumption by Application ..........................................................................................................................34
`
`Appendix B. Power Consumption Trend Data and Assumptions ..................................................................................................35
`
`Appendix C. ACPI Power State Descriptions ..................................................................................................................................36
`
`THIS DOCUMENT IS PROVIDED "AS IS" WITH NO WARRANTIES WHATSOEVER, INCLUDING ANY WARRANTY OF MERCHANTABILITY,
`NONINFRINGEMENT, FITNESS FOR ANY PARTICULAR PURPOSE, OR ANY WARRANTY OTHERWISE ARISING OUT OF ANY PROPOSAL,
`GUIDELINE, SPECIFICATION OR SAMPLE. Intel disclaims all liability, including liability for infringement of any proprietary rights, relating to use of
`information in this specification. No license, express or implied, by estoppel or otherwise, to any intellectual property righ ts is granted herein. Copyright © Intel
`Corporation 1997. *Third-party brands and names are the property of their respective owners.
`
`December 1, 1997
`
`3
`
`MICROCHIP TECH. INC. - EXHIBIT 1032
`MICROCHIP TECH. INC. V. HD SILICON SOLS. - IPR2021-01265 - Page 003
`
`

`

`Mobile Power Guidelines Rev. 1.00
`
`1. Executive Summary
`
`The demand for new features and higher performance in mobile PCs presents a major challenge to the mobile PC industry in the area
`of platform power. Power dissipated in the interior of a full-featured notebook has increased by 90% in the last three years. Looking
`ahead to expected system enhancements, this power consumption could increase to more than 35 watts by 1999. This would create a
`thermal problem since 1999 notebook computers are only expected to be able to dissipate about the same 23 to 25 watts as today’s
`systems.
`
`Intel’s Mobile Power Initiative is a coordinated industry program that addresses these power challenges. It is a comprehensive
`program that spans across all areas that impact power: platform, operating system, and applications. To address application power,
`Intel has created power monitoring and analysis tools that help software developers identify and correct power wasting code. To
`improve operating system power management, Intel, Microsoft and Toshiba authored an industry power management specification
`called the Advanced Configuration and Power Interface (ACPI.) This Mobile Power Guidelines focuses on the platform power issues
`by providing achievable power targets for all system components. Meeting these component power targets ensures that mobile PC
`systems in 1999 will contain all the features and performance that users demand, while holding power to within the thermal limits of
`today’s mobile PCs. Hardware component vendors and original equipment manufacturers will benefit by cost reductions due to lower
`power components, lighter weight thermal solutions, and higher product reliability.
`
`This document also discusses various techniques that can be used to hit the power targets. Among them, voltage reduction is the most
`significant method of reducing power. Table 1.1 shows the key component voltage and power changes targeted for systems shipping
`in mid 1999.
`
`Table 1.1 1999 Key Component Voltage and Power Targets
`
`Component
`
`CPU Core
`Memory Controller
`System Memory
`System Memory Bus
`Graphics Controller
`Graphics Frame Buffer
`Advanced Graphics Port
`Bus (AGP)
`Power Supply
`
`1998 Estimate
`Voltage (Volts)
`3D WinBench
`Power (Watts)
`7.9
`1.7
`1.7
`
`1.6
`3.3
`3.3 (SDRAM)
`3.3
`3.3
`3.3
`3.3
`
`2.0
`0.6
`
`1999 Targets
`Voltage (Volts)
`3D WinBench
`Power (Watts)
`7.9
`1.2
`1.4
`
`1.6 or lower
`1.8
`2.5 (RDRAM(cid:226) *)
`1.8
`2.4
`1.8
`Frame Buffer Integrated into Graphics Controller
`1.5
`
`88% Efficient
`
`90% Efficient
`
`Implementing these voltage and power reduction targets allows the performance improvements customers demand while keeping the
`overall system power within the 23 to 25 watt thermal envelope. Intel is committed to continuing the voltage reduction trend on
`mobile system components it manufactures. We encourage mobile computer component and original equipment manufacturers to
`join this trend. Those who embrace it will reap many benefits including the following:
`
`O Lower power components
`4 Lower cost component packages
`4 More reliable components
`4 Higher performance
`
`O Lower power systems
`4 Lower cost thermal solutions
`4 Lighter weight thermal solutions
`4 More reliable systems
`4 More room for additional features or performance
`
`We call upon industry leaders like yourself to join us in bringing about the changes needed to unlock the opportunities on the mobile
`horizon. Designing your components and systems to meet the Intel Mobile Power Guidelines will benefit both your customers and
`the mobile PC industry.
`
`December 1, 1997
`
`4
`
`MICROCHIP TECH. INC. - EXHIBIT 1032
`MICROCHIP TECH. INC. V. HD SILICON SOLS. - IPR2021-01265 - Page 004
`
`

`

`Mobile Power Guidelines Rev. 1.00
`
`2. Introduction
`New technologies and feature rich applications are promising unprecedented opportunities for mobile computer system, hardware
`component, and software vendors. Users expect the same high performance from their mobile computers that they enjoy on their
`desktops. At the same time mobile computer system design is becoming more challenging.
`
`As we look to the future, we see these opportunities bounded by the thermal realities of mobile computer system design. New
`systems cannot increase performance or add new features if they can not dissipate the heat. To address this issue, new systems can
`either find ways to dissipate more heat or find ways to reduce power consumption. In the next few years, we feel there are significant
`opportunities and benefits in reducing power consumption.
`
`To support this strategy Intel is significantly reducing voltages on mobile system components it manufactures such as
`microprocessors and chip sets. However, enabling higher performance systems with new features will require lowering the voltage
`of many other system components to keep the overall system thermally manageable. This will require an industry wide effort to
`lower bus interface voltages and establish power guidelines. To facilitate this cooperative industry effort, Intel is proposing these
`"Mobile Power Guidelines."
`
`2.1 Objective
`The objective of this document is to unify the mobile computing industry with a common set of guidelines to enable continued
`feature and performance enhancements in a thermally manageable system.
`
`This document describes the challenges mobile system designers will face in the next few years if nothing is done to manage system
`power consumption and sets bus interface and component voltage and power targets for 1999 systems. This document describes
`high-end, full-size, full-featured mobile platforms expected to be in production in mid 1999.
`
`December 1, 1997
`
`5
`
`MICROCHIP TECH. INC. - EXHIBIT 1032
`MICROCHIP TECH. INC. V. HD SILICON SOLS. - IPR2021-01265 - Page 005
`
`

`

`Mobile Power Guidelines Rev. 1.00
`
`3. Notebook System Power Trends
`Notebook computers are becoming increasingly more sophisticated incorporating many of the features of high performance desktop
`systems. These include such features as high performance processors, second level cache, better graphics, DVD drives, and new 1/0
`busses like USB. As new notebook computer generations are introduced the power required to implement these new features keeps
`increasing. Projecting this trend forward over the next few years to include features such as larger LCD screens, accelerated 3D
`graphics, more system memory, and 1394 based 1/0 devices shows system power requirements increasing significantly.
`
`Table 3.1 shows tl1e growth in full featured notebook interior power since 1994 and projects notebook interior power through 1999 if
`nothing is done to manage power consmnption. Notebook interior power includes all components inside the notebook main chassis.
`It excludes power for the LCD panel and external connections such as USB and 1394 since they do not add to the thermal energy
`dissipated in tl1e main notebook chassis.
`
`40....---------------------------,
`
`Source: Intel Corp.
`
`35
`
`30
`
`25
`
`20
`Watts
`15
`
`10
`
`5
`
`0 -1---"--....,_---1
`1994
`
`_
`
`_.__.__--l-__._.....__-l--_.__..__---1----J'-----1,-4--....L.-..._----I
`1995
`1996
`1997
`
`1998
`Est.
`
`1999
`Est.
`
`■ Other Electronics
`■ cooling
`
`□ Audio
`□ clocks
`■ Keyboard Controller
`■ PCMCIA/Card Bus Controlle
`□ super 10
`■ Power Supply
`□ Drives
`■ Graphics
`
`□ Chipset
`□ Memory
`
`■ L2Cache
`□ CPU
`Average power while rnnning worst
`case application. Excludes display
`
`Table 3.1 Unmanaged Notebook Power Trends
`
`This trend shows power dissipated in tl1e notebook interior increasing by 90% between 1994 and 1997 and projects power increasing
`by another 85% between 1997 and 1999. (See appendix B for power trend data and assumptions.) If this trend in notebook power
`consumption goes unmanaged, it will soon outpace notebook thermal capabilities limiting future system features and performance.
`
`3.1 Notebook Cooling Capabilities
`While notebook power consumption has been increasing rapidly, cooling techniques to remove the heat generated witllin the
`notebook have been evolving more slowly. Current notebook systems measuring 8 ½ x 11 inches witl1 base units (excluding lid
`thickness) between 0.75 inches and 1.5 inches can dissipate about 23 to 25 Watts. 1 Tllis thermal envelope makes the following
`assumptions:
`• Ambient room te1nperature is approximately 25 °C.
`• The bottom surface of the notebook is insulated and the keyboard surface is used as a radiator.
`• The notebook skin is made of ABS/PC plastic and has an isotl1ermal skin temperature.
`• The notebook uses a combined heat spreader plate, heat pipe, remote heat exchanger, and fan cooling method.
`The heat from inside a notebook is dissipated by warming the outside surfac•e (skin) of the plastic over the ambient air and by using a
`fan and heat exchanger. Most fan designs can remove about 4 to 6 Watts of heat and the rest is passively dissipated by natural
`
`1 Refer to Intel Application Note 584 "Notebook Thennal Design Guide for High-Powered Microprocessors" (order# 243321-001)
`for a detailed discussion of notebook tl1ermal considerations.
`December 1, 1997
`
`6
`
`MICROCHIP TECH. INC. - EXHIBIT 1032
`MICROCHIP TECH. INC. V. HD SILICON SOLS. - IPR2021-01265 - Page 006
`
`

`

`Mobile Power Guidelines Rev. 1.00
`
`convection and radiation from the outside surface. If there are no limits on the outside surface temperature, the passive heat
`dissipation limit is unbounded. Of course, such a notebook with unconstrained high skin temperatures will be unusable. Due to
`ergonomic constraints limiting notebook computer skin temperature to approximately 15 °C over ambient air temperature, systems in
`1999 are likely to still be constrained within the same thennal envelope as today.
`
`Table 3.2 shows the change in skin temperature for a given notebook size, ambient temperature and total power dissipated in the base
`of the notebook. The dotted lines show the approximate power dissipation limits for an 8.5 x 11 size notebook with a base height
`between 0.75 inches to 1.5 inches. The total heat dissipation power in the figure does not include the display power. This thermal
`constraint shows the projected system power is unmanageable and requires action to reduce future notebook system power.
`
`Total Heat Dissipation From An Idealized Notebook Base
`For Notebook Operation at Tair = 25 •c, Aecom mended
`fl T between NB surface and air tern perature = 15 •c
`35 "T"""---------,r---------r--;r---------r-----2"'~---,
`8.5 in x 11 in x 1.5 in
`8.5 in x 11 in x 1.0 in
`■
`en 30 -----------+------------1
`8.5 in x 11 in x 0.75 in
`::::
`
`~ -
`
`a:
`
`...... ---------------------1
`10+---------------------
`40
`45
`30
`35
`Notebook Surface Temperature (°C)
`
`50
`
`Table 3.2 Heat Dissipation Limits From a Notebook Base
`
`December 1, 1997
`
`7
`
`MICROCHIP TECH. INC. - EXHIBIT 1032
`MICROCHIP TECH. INC. V. HD SILICON SOLS. - IPR2021-01265 - Page 007
`
`

`

`Mobile Power Guidelines Rev. 1.00
`
`3.1.1 Notebook System Power Targets
`To reduce notebook system power to thermally manageable levels we are proposing power targets for each subsystem. Implementing
`these subsystem power targets will bring the overall notebook power back within the 23 to 25 watt thermal limit. Table 3.3 shows the
`notebook power targets achievable with industry coordination and illustrates how much power can be saved over the urunanaged
`system power projections. Subsystem power targets are described in detail in section 5.
`
`40-,--------------------------.
`Source: Intel Corp.
`
`35
`
`30
`
`25
`
`20
`
`15
`
`10
`
`Watts
`
`5
`0 _ ___._ _ _,_ __
`
`....___._-+-__.-..J-____,,----L.._.,__
`
`_ __, _ __._ __
`
`--L.._....__---1
`
`1994
`
`1995
`
`1996
`
`1997
`
`1998
`Est.
`
`1999
`Est.
`
`Table 3.3 Notebook Power Targets with Industry Coordination
`
`■ Other Electronics
`■ Cooling
`□Audio
`□ Clocks
`■ Keyboard Controller
`■ PCMCIA/Card Bus Controlle
`□ Super 10
`■ Power Supply
`□ Drives
`
`■ Graphics
`□ Chipset
`□ Memory
`■ L2 Cache
`□ CPU
`Average power while running worst
`case application. Excludes display.
`
`December 1, 1997
`
`8
`
`MICROCHIP TECH. INC. - EXHIBIT 1032
`MICROCHIP TECH. INC. V. HD SILICON SOLS. - IPR2021-01265 - Page 008
`
`

`

`Mobile Power Guidelines Rev. 1.00
`
`4. Power Reduction Methods
`Meeting the power targets and improving notebook system performance can be accomplished through voltage reduction, component
`optimization, improvements in device power management and software optimization.
`
`4.1 Voltage Reduction
`Power consumed in CMOS circuits is proportional to frequency and voltage as shown in the following equation:
`
`Power (cid:181)
`
` (Capacitance *Frequency * (Voltage)2) + (Voltage * (DC Current + Leakage Current))
`
`The DC and leakage currents tend to be small for CMOS circuits which makes the CFV2 term dominate this equation. Reducing the
`frequency reduces power linearly. However, it has the disadvantage of reducing performance linearly as well. Reducing the voltage
`has the advantage of reducing power as the square of the voltage. Furthermore, components optimized for lower voltage operation
`also benefit from improved performance since they can operate at higher speeds. Therefore, voltage reduction yields the biggest
`component power improvement.
`
`Intel is improving system performance and holding down power consumption by significantly reducing voltages on microprocessors
`and chipsets. Additionally, Intel is actively promoting low voltage interfaces such as 1.8 volt cache and direct RDRAM(cid:228) * signaling,
`and 1.5 volt Advanced Graphics Port (AGP) signaling. To add new features and increase performance in other parts of the system,
`more system components need to move to lower voltages.
`
`4.1.1 Component Core Voltages
`Table 4.1 shows Intel’s proposed mobile component core voltage targets through the year 1999.
`
`Table 4.1 Proposed Mobile System Component Core Voltages
`
`CPU Core
`L2 Cache
`Memory Controller
`System Memory
`Graphics controller
`Graphics Frame Buffer
`I/O Controller
`LCD Logic
`Backlight inverter
`FLASH EPROM
`Audio (Digital)
`Audio CODEC
`Super I/O
`1394 Controller
`CardBus Controller
`LAN
`MODEM DSP
`MODEM CODEC / DAA
`
`1997
`
`1.8 V
`3.3 V
`3.3 V
`3.3 V (EDO / SDRAM)
`3.3 V
`3.3 V (EDO / SDRAM)
`3.3 V
`5 V
`Battery voltage
`3.3 V
`3.3 V
`5 V
`3.3 V
`None
`3.3 V
`3.3 V (PC Card or dock)
`3.3 V
`5 V
`
`1998
`
`1999
`
`1.6 V or lower
`1.6 V
`3.3 V
`3.3 V
`1.8 V
`3.3 V
`2.5 V (RDRAM)
`3.3 V (EDO / SDRAM)
`1.8 V
`2.5 V
`3.3 V (EDO / SDRAM / RDRAM) 3.3 V or Integrated
`3.3 V
`2.5 V
`5 V or 3.3 V
`3.3 V
`Battery voltage
`Battery voltage
`3.3 V
`3.3 V
`3.3 V
`2.5 V
`5 V
`3.3 V
`3.3 V
`2.5 V
`None
`3.3 V
`3.3 V
`2.5 V
`3.3 V (PC Card or dock)
`3.3 V
`3.3 V
`Soft Modem
`5 V
`3.3 V
`
`December 1, 1997
`
`9
`
`MICROCHIP TECH. INC. - EXHIBIT 1032
`MICROCHIP TECH. INC. V. HD SILICON SOLS. - IPR2021-01265 - Page 009
`
`

`

`Mobile Power Guidelines Rev. 1.00
`
`4.1.2 Bus Signaling Voltages
`Table 4.2 shows Intel’s proposed bus voltage targets through the year 1999.
`
`Table 4.2 - Proposed Bus Interface Voltages
`
`CPU Host Bus
`CPU Cache Bus
`Memory Bus
`Graphics Interface
`Graphics Frame Buffer Bus
`
`1997
`2.5 V (CMOS)
`NONE
`3.3 V (EDO / SDRAM)
`3.3 V (PCI)
`3.3 V (EDO / SDRAM)
`
`LCD Interface
`PCI Bus
`ISA subset Bus
`IDE
`Floppy Drive Interface
`USB Signaling
`CardBus
`1394 Link Interface
`1394 Physical Interface
`SMBUS
`Parallel Port
`Serial Port
`
`5 V
`3.3 V
`3.3 V
`3.3 V / 5 V safe
`3.3 V / 5 V safe
`3.3 V
`3.3 V
`NONE
`NONE
`3.3 V
`3.3 V (5 V safe)
`+/-12 V RS232
`
`1998
`
`1.6 V (GTL)
`1.8 V
`3.3 V (EDO / SDRAM)
`3.3 V (AGP)
`3.3 V (EDO / SDRAM / RDRAM
`or integrated)
`0.7 V (LVDS or panel link)
`3.3 V
`3.3 V
`3.3 V / 5 V safe
`3.3 V / 5 V safe
`3.3 V
`3.3 V
`NONE
`NONE
`3.3 V
`3.3 V (5 V safe)
`+/-12 V RS232
`
`1999
`1.6 V or lower (GTL)
`1.8 V
`1.8 V (Direct RDRAM)
`1.5 V (AGP)
`integrated
`(or EDO / SDRAM / RDRAM)
`0.7 V (LVDS or panel link)
`3.3 V
`NONE
`3.3 V / 5 V safe
`3.3 V / 5 V safe
`3.3 V
`3.3 V
`2.5 V
`3.3 V
`3.3 V
`3.3 V (5 V safe)
`+/-12 V RS232
`
`4.2 Component Optimization
`In addition to reducing component voltages, integrated circuits can be made more power efficient through design optimization.
`During integrated circuit design, benchmarks can be developed to identify where power is being consumed on a unit by unit basis.
`Once the power distribution profile is characterized, low power design techniques such as gating the clock to unused sections of
`circuitry or reducing the number of nodes toggling for a given operation can be implemented. Buffer design can also be optimized to
`reduce power by not overdriving signals. Implementing these design optimization techniques can significantly reduce both idle and
`active power.
`
`4.3 Device Power Management
`Intel, Microsoft, and Toshiba defined a power management interface called the Advanced Configuration and Power Interface
`(ACPI). ACPI defines system and device power states and a consistent register interface for device configuration and control. This
`allows better component power management allowing components to spend more time in low power states. This will help reduce the
`system thermal load and will improve battery life for typical applications. The amount of power saved will be influenced significantly
`by application and usage patterns. Refer to Appendix C for a brief description of the defined ACPI system and device power states.
`
`4.4 Software Optimization
`Software can be optimized to conserve power by eliminating loops and software sequences that prevent the system from going into
`low power idle states and by optimizing software design. Intel has developed guidelines and tools to assist software developers in
`making their programs more power friendly. This includes a software tool called the Intel Power Monitor (IPM). This tool monitors
`and displays system activity allowing software vendors to identify programs that are wasting power. If the power monitor detects
`specific power wasting code loops, it can demonstrate the impact of fixing the software by temporarily updating the running code to
`be more power friendly. The Intel power monitor is available on the world wide web at http://www.intel.com/ial/ipm.
`
`December 1, 1997
`
`10
`
`MICROCHIP TECH. INC. - EXHIBIT 1032
`MICROCHIP TECH. INC. V. HD SILICON SOLS. - IPR2021-01265 - Page 010
`
`

`

`Mobile Power Guidelines Rev. 1.00
`
`5. System Feature and Power Targets
`This section describes system feature assumptions and power targets for 1999 notebook computers. 1998 feature and power figures
`are included for comparison. Section 5.1 shows the system configuration assumptions. Section 5.2 proposes power targets for each
`subsystem and summarizes the total power dissipated in the notebook interior. Section 5.3 provides more detail on the subsystem
`power targets and shows utilization assumptions. It also provides implementation ideas and design considerations.
`
`5.1 System Configurations
`Table 5.1 shows the system configuration assumptions for mid 1998 and mid 1999 systems. This only describes high-end, full-
`featured systems implementing the latest technology since these will be the systems pushing the thermal envelope. It is assumed that
`solving the thermal problem for high-end systems will automatically solve it for full-sized mid-range and entry level systems.
`
`Table 5.1 - 1998 and 1999 System Configuration Assumptions
`
`1998 System Configuration
`• Mobile Pentium(cid:210)
` II processor
`•
`512 KB pipeline burst level two cache
`• Graphics controller
` AGP 1x interface
` 4 MB frame buffer
` 1024 x 768 x 18 bit / pixel resolution
` MPEG2 H/W assist (motion compensation, YUV 4:2:0)
` 3D acceleration
` LCD and CRT dual screen support
` LVDS or Panel Link display interface
`
`•
`
`•
`13.3” Color TFT LCD display
`• Memory Subsystem
` 32MB EDO or SDRAM memory
`I/O Subsystem
` I/O controller with integrated timers, counters, etc.
` 3.3v 33 MHz PCI bus
` Fast IR
` System Management Bus controller
` Universal Serial Bus controller
` Parallel and serial ports
` Keyboard controller
` ACPI microcontroller with SM Bus system battery interface
` 3.3 / 5v ISA subset bus
`
`•
`•
`
`1 USB port
`Storage media
`Floppy drive
` IDE hard drive
` IDE DVD-ROM
`• CardBus
`2 slots
`• Audio
` Sound blaster H/W compatible
` Wavetable synthesis (down loadable samples)
` 3D positional sound
` AC3 / MPEG2 decode
`Full duplex G.723.1 encode/decode with acoustic echo cancel.
`Host controller
`• Data, fax, voice modem including V.80 support
`•
`PCI Docking
`•
`LAN on CardBus card
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`1999 System Configuration
`• Mobile Pentium(cid:210)
` II processor or next generation
`•
`512 KB pipeline burst level two cache
`• Graphics controller
` AGP 2x interface
` 4 MB frame buffer
` 1024 x 768 x 24 bit / pixel resolution
` MPEG2 H/W assist (motion compensation, YUV 4:2:0)
` Enhanced 3D acceleration
` LCD and CRT dual screen support
` LVDS or Panel Link display interface
` TV output
`•
`13.3” Color TFT LCD display
`• Memory Subsystem
` 64MB RDRAM memory
`I/O Subsystem
` I/O controller with integrated timers, counters, etc.
` 3.3v 33 MHz PCI bus
` Fast IR
` System Management Bus controller
` Universal Serial Bus controller
` Parallel and serial ports
` Keyboard controller
` ACPI microcontroller with SM Bus system battery interface
`
`•
`
`•
`•
`•
`
`1 1394 walk-up port (S400)
`1 USB port
`Storage media
`Floppy drive
` IDE hard drive
` IDE DVD-ROM in swap bay
`• CardBus
`2 power managed slots
`• Audio
` Sound blaster S/W emulation
` Wavetable synthesis (down loadable samples)
` 3D positional sound
` AC3 / MPEG2 decode
`Full duplex G.723.1 encode/decode with acoustic echo cancel.
`Host controller
`Software modem data pump
`PCI Docking
`LAN on motherboard
`
`•
`•
`•
`
`5.2 System Power Targets
`This section proposes power targets for each subsystem and summarizes the total power dissipated in the notebook interior.
`December 1, 1997
`
`11
`
`MICROCHIP TECH. INC. - EXHIBIT 1032
`MICROCHIP TECH. INC. V. HD SILICON SOLS. - IPR2021-01265 - Page 011
`
`

`

`Mobile Power Guidelines Rev. 1.00
`
`To define system power targets it is necessary to identify the applications that generate the worst case system thermal power that is
`likely to occur in normal use. Appendix A shows system power measurements for several applications measured on an Intel
`evaluation board in 1997. These tests identified MPEG2 video playback as having the worst case thermal power of the applications
`we tested. Extrapolating this data and factoring in significant system improvements such as multimedia enhancements, we expect the
`worst case power applications in 1998 and 1999 to be MPEG2 video playback and 3D game play. While these applications are not
`expected to be the most commonly used applications they are expected to be the worst case power applications since they both are
`CPU intensive requiring a lot of video and audio decompression, high resolution, fast action graphics, DVD drive and memory
`accesses, and audio playback.
`
`The 3D game scenario assumes scenes are stored on an IDE DVD drive, the hard drive is accessed occasionally, and audio is active
`at 30% amplifier power. A CardBus LAN card is connected in the 1998 scenario and in 1999 an onboard LAN is connected. The
`LAN is assumed to be predominantly idle. In 1999 the CardBus slots are assumed empty.
`
`The MPEG2 movie scenario assumes the movie is playing from an IDE DVD drive, hard drive is powered off, and audio is active at
`30% amplifier power. The CardBus slots are unused. In 1999 the onboard LAN is assumed connected but mostly idle.
`
`In this document power is specified in terms of peak power, idle power, and average power while running the 3D game and MPEG2
`applications. Since it is difficult to benchmark power consumption during 3D game play, the power targets for most subsystems are
`specified while running the benchmark program 3D WinBench*. 3D WinBench power is assumed to be similar to 3D game power
`consumption.
`
`Peak power refers to the highest power a subsystem will draw at nominal Vcc and maximum performance (i.e. CPU doing floating
`point calculations at top frequency with most data hitting the level 1 cache, DRAM reading at peak bandwidth, etc.) This peak
`performance assumes the subsystem is utilized at 100% capacity and that it draws maximum power at maximum performance. Idle
`power refers to the power a subsystem will draw at nominal Vcc when it is not processing data. This is defined as 0% utilization.
`During normal system operation most subsystems are utilized at some percentage of their maximum capacity and maximum power.
`This document assumes that for most subsystems power consumption scales linearly with utilization between idle power and peak
`power. This is illustrated in Table 5.2. For some components, such as an audio accelerator, power does not scale linearly with
`utilization. For these components, power was estimated individually and summed with the rest of the subsystem. For components
`where power is approximately linear with utilization, the average thermal power for a subsystem running a particular application can
`be estimated with the following equation:
`
`Subsystem Average Thermal Power = (Subsystem Peak Power * Utilization) + (Subsystem Idle Power * (1 - Utilization))
`
`Subsystem
`Peak Power
`
`Subsystem average
`thermal power
`for application
`
`Subsystem
`Idle Power
`
`Subsystem utilization
`for a particular
`application
`
`Table 5.2 Subsystem Average Thermal Power
`
`0%
`
`Utilization
`
`100%
`
`The worst case system thermal power is the sum of the average thermal power for each subsystem while running the worst case
`power application. Table 5.3 shows a summary of the proposed 1999 power targets for the 3D game and MPEG 2 movie scenarios.
`1998 power targets are included for comparison.
`
`December 1, 1997
`
`12
`
`MICROCHIP TECH. INC. - EXHIBIT 1032
`MICROCHIP TECH. INC. V. HD SILICON SOLS. - IPR2021-01265 - Page 012
`
`

`

`Mobile Power Guidelines Rev. 1.00
`
`Table 5.3 System Power Target Smmnary (Notebook interior. Excludes Display)
`
`CPU core
`L2 Cache
`Memorv Controller
`System Memory
`Graphics Subsvsterr
`10 Subsvstem
`Audio
`Hard Drive
`DVD Drive
`1394 Controller
`use
`CardBus
`LAN
`Power SUPPIV
`Charging
`Coollng
`Other
`Total
`
`Estimated 1998 Power (Watts)
`3D Game
`MPEG 2 Movie
`7.9
`6.5
`1.4
`1.6
`1.7
`1.7
`1.7
`1.7
`2.6
`2.6
`0.5
`0.7
`2.0
`2.0
`0.7
`0.0
`1.4
`3.0
`0.0
`0.0
`0.0
`0.0
`1.2
`0.2
`0.0
`0.0
`2.8
`2.7
`0.1
`1.0
`0.5
`0.5
`0.5
`0.5
`25.2
`24.5
`
`1999 Power Goals (Watts)
`3D Game
`MPEG 2 Movie
`7.9
`6.5
`1.6
`1.4
`1.2
`0.9
`1.4
`1.4
`2.4
`2.4
`0.5
`0.6
`1.5
`1.5
`0.7
`0.0
`1.4
`3.0
`0.0
`0.0
`0.0
`0.0
`0.1
`0.1
`0.4
`0.4
`2.0
`2.0
`0.1
`1.0
`0.5
`0.5
`0.3
`0.3
`22.0
`22.0
`
`5.3 Subsystem Power Targets
`The following sections elaborate on the power targets, utilization assumptions, and implementation suggestions for each subsystem.
`Subsystem utilization projections are based on measured subsystem power and adjusted for estimated changes due to higher bit rate
`data streams, higher bandwidth busses (AGP, RDRAM, etc.), and higher perfonnance subsystems.
`
`December 1, 1997
`
`13
`
`MICROCHIP TECH. INC. - EXHIBIT 1032
`MICROCHIP TECH. INC. V. HD SILICON SOLS. - IPR2021-01265 - Page 013
`
`

`

`Mobile Power Guidelines Rev. 1.00
`
`5.3.1 CPU and Level Two Cache
`
`5.3.1.1 Features
`The following CPU and level two cache subsystem features are assumed in 1998 and 1999 systems.
`
`1998 Features
`• Mobile Pentium(cid:226)
` II Processor
`•
`512 KByte level two cache
`
`1999 Features
`• Mobile Pentium(cid:226)
` II Processor or next generation
`•
`512 KByte level two cache
`
`5.3.1.2 Power Targets
`The following table shows the CPU and Level two cache power targets and utilization assumptions for 1998 and 1999.
`
`Estimated 1998 CPU & L2 Cache Power
`Peak Power
`Idle Power
`Average 3D
`Average
`(W)
`(W)
`WinBench
`MPEG 2
`Power (W)
`Power (W)
`7.9
`6.5
`1.6
`1.4
`80%
`65%
`24%/50% 19%/40%
`
`0.4
`9.5
`0.6
`2.8
`CPU Utilization
`L2 Utilization (tag / data)
`
`CPU
`L2
`
`1999 CPU & L2 Cache Power Targets
`Peak Power
`Idle Power
`Average 3D
`Average
`(W)
`(W)
`WinBench
`MPEG 2
`Power (W)
`Power (W)
`7.9
`6.5
`1.6
`1.4
`80%
`65%
`24%/50% 19%/40%
`
`0.4
`0.6
`
`9.5
`2.8
`
`Average 3D WinBench power represents the highest sustained power that a real worst-case application will draw from the CPU. The
`typical thermal design power (TDP typical) specification for an Intel processor is derived from sustained power measurements from
`worst case applications such as 3D WinBench with margin added to account for process variations. The peak CPU power guideline
`represents the maximum power consumed while executing the worst case CPU instruction mix at nominal Vcc. This resembles the
`TDP max specification for Intel processors. These are design guidelines and should not be construed as future processor
`specifications.
`
`The CPU average power includes 15% CPU to memory controller bus utilization (0.2 Watts). CPU core thermal design power is held
`at or below 8 watts through 1999.
`
`December 1, 1997
`
`14
`
`MICROCHIP TECH. INC. - EXHIBIT 1032
`MICROCHIP TECH. INC. V. HD SILICON SOL