From WikiChip
Difference between revisions of "intel/microarchitectures/ice lake (client)"
< intel‎ | microarchitectures

(SoC)
(Overview)
(37 intermediate revisions by 2 users not shown)
Line 5: Line 5:
 
|designer=Intel
 
|designer=Intel
 
|manufacturer=Intel
 
|manufacturer=Intel
|introduction=2019
+
|introduction=May 27, 2019
 
|process=10 nm
 
|process=10 nm
 
|isa=x86-64
 
|isa=x86-64
Line 48: Line 48:
 
== Process Technology==
 
== Process Technology==
 
{{see also|intel/microarchitectures/cannon lake#Process_Technology|l1=Cannon Lake § Process Technology}}
 
{{see also|intel/microarchitectures/cannon lake#Process_Technology|l1=Cannon Lake § Process Technology}}
Ice Lake will use a second-generation enhanced [[10 nm process]] called "10 nm+". Versus the first generation 10nm which was used for Cannon Lake, 10nm+ will feature higher performance through higher drive current for the same power envelope.
+
Ice Lake is fabricated on Intel's second-generation enhanced [[10 nm process]] called "10 nm+". Versus the first generation 10nm which was used for {{\\|Cannon Lake}}, 10nm+ features higher performance through higher drive current for the same power envelope. Intel says that Ice Lake is built on their learnings from their Cannon Lake products which were largely treated as a learning vehicle. Between Cannon Lake and Ice Lake, a number of changes were made in order to improve the process for their products. One such change was the addition of an extra metal layer (originally said to be 12, is now presumably 13 on Ice Lake) in order to improve the power delivery of the chip. Additionally, they have improved the threshold voltage of the transistors as well as their MIM cap among other changes.
  
 
[[File:intels 10+ and 10++.png|750px]]
 
[[File:intels 10+ and 10++.png|750px]]
Line 83: Line 83:
  
 
== Architecture ==
 
== Architecture ==
Not much is known about Ice Lake's architecture.
+
Ice Lake comprises of {{\\|Sunny Cove}} cores on the {{intel|ring interconnect architecture}} along with {{\\|Gen11}} GPU, and an improved {{intel|System Agent}} with a new display engine and I/O.
  
=== Key changes from {{\\|Cannon Lake}}===
+
=== Key changes from {{\\|Cannon Lake}}/{{\\|Skylake}}===
 
* Enhanced "10nm+" (from "10nm", 2nd gen)
 
* Enhanced "10nm+" (from "10nm", 2nd gen)
* {{\\|Sunny Cove|Sunny Cove core}} (from {{\\|Palm Cove}})
+
* Core
** ''See {{\\|Sunny Cove}} for microarchitectural details and changes''
+
** {{\\|Sunny Cove|Sunny Cove core}} (from {{\\|Palm Cove}})
* {{intel|Gen10|l=arch}} → {{intel|Gen11|l=arch}} graphics
+
*** ''See {{\\|Sunny Cove}} for microarchitectural details and changes''
* {{intel|Gen11|l=arch}} GPUs
+
* Memory
** UHD Graphics 7xx (GT1) '''→''' UHD Graphics 9xx (GT2) (32 Execution Units, 1.3x EUs from {{\\|Cannon Lake}})
+
** 4 32-bit [[LPDDR4X]] channels (from 2 64-bit [[DDR4]] channels)
** UHD Graphics 7xx (GT2) '''→''' Iris Plus Graphics 9xx (GT2) (48-64 Execution Units, 1.2-1.6x EUs from {{\\|Cannon Lake}})
+
** 1.4x higher data rates (3733 MT/s, up from 2666 MT/s)
 +
*** 1.5x higher memory bandwidth (60 GB/s, up from 40 GB/s)
 +
* Graphics
 +
** {{intel|Gen10|l=arch}} → {{intel|Gen11|l=arch}} graphics (''Gen10 was never productized'')
 +
** {{intel|Gen11|l=arch}} GPUs
 +
*** UHD Graphics 6xx (GT1) '''→''' UHD Graphics 9xx (GT2) (24 Execution Units, 2x EUs from {{\\|Gen9}})
 +
*** UHD Graphics 6xx (GT2) '''→''' Iris Plus Graphics 9xx (GT2) (48-64 Execution Units, 2-2.6x EUs from {{\\|Gen9}})
 +
**** 1,024 GFLOPS @ 1 GHz (GT2)
 
* Display
 
* Display
 +
** Gen 11.5 (from Gen9/Gen9.5)
 
** DisplayPort 1.4a with Display Stream Compression(DSC) (from DisplayPort 1.2)
 
** DisplayPort 1.4a with Display Stream Compression(DSC) (from DisplayPort 1.2)
** HDMI 2.0 (from HDMI 1.4)
+
** [[HDMI]] 2.0b (from HDMI 1.4)
 
* IPU
 
* IPU
 
** 4th Gen IPU (from 3rd Gen in {{\\|Skylake (client)|Skylake}})
 
** 4th Gen IPU (from 3rd Gen in {{\\|Skylake (client)|Skylake}})
 +
** More cameras support
 +
** New concurrent image pipeline
 +
** on-die MIPI interface
 +
* New Integration
 +
** New Gaussian Neural Accelerator 1.0 (Added in {{\\|Cannon Lake}} but unclear to what extent)
 
* I/O
 
* I/O
 
** Thunderbolt 3 over Type-C
 
** Thunderbolt 3 over Type-C
 +
* Package
 +
** New Type3, Type4 packages
 +
*** New thin-film magnetic inductors
  
 
{{expand list}}
 
{{expand list}}
  
 
====New instructions ====
 
====New instructions ====
Ice Lake introduced a number of {{x86|extensions|new instructions}}. See {{intel|Sunny cove#New instructions|Sunny Cove § New Instructions|l=arch}} for details.
+
{{see also|intel/microarchitectures/sunny_cove#New_instructions|l1=Sunny Cove § New Instructions}}
 +
Ice Lake introduced a number of {{x86|extensions|new instructions}}.
 +
 
 +
* {{x86|SHA|<code>SHA</code>}} - [[Hardware acceleration]] for SHA hashing operations
 +
* {{x86|CLWB|<code>CLWB</code>}} - Force cache line write-back without flush
 +
* {{x86|RDPID|<code>RDPID</code>}} - Read Processor ID
 +
* {{x86|AVX-512|<code>AVX-512</code>}} (originally introduced in {{\\|Skylake (Server)}} but only now in client)
 +
** {{x86|AVX512F|<code>AVX512F</code>}} - AVX-512 Foundation
 +
** {{x86|AVX512CD|<code>AVX512CD</code>}} - AVX-512 Conflict Detection
 +
** {{x86|AVX512BW|<code>AVX512BW</code>}} - AVX-512 Byte and Word
 +
** {{x86|AVX512DQ|<code>AVX512DQ</code>}} - AVX-512 Doubleword and Quadword
 +
** {{x86|AVX512VL|<code>AVX512VL</code>}} - AVX-512 Vector Length
 +
* Additional {{x86|AVX-512}} extensions:
 +
** {{x86|AVX512VPOPCNTDQ|<code>AVX512VPOPCNTDQ</code>}} -  AVX-512 Vector Population Count Doubleword and Quadword
 +
** {{x86|AVX512VNNI|<code>AVX512VNNI</code>}} -  AVX-512 Vector Neural Network Instructions
 +
** {{x86|AVX512GFNI|<code>AVX512GFNI</code>}} -  AVX-512 Galois Field New Instructions
 +
** {{x86|AVX512VAES|<code>AVX512VAES</code>}} -  AVX-512 Vector AES
 +
** {{x86|AVX512VBMI2|<code>AVX512VBMI2</code>}} -  AVX-512 Vector Bit Manipulation, Version 2
 +
** {{x86|AVX512BITALG|<code>AVX512BITALG</code>}} -  AVX-512 Bit Algorithms
 +
** {{x86|AVX512VPCLMULQDQ|<code>AVX512VPCLMULQDQ</code>}} -  AVX-512 Vector Vector Carry-less Multiply
 +
* {{x86|SSE_GFNI|<code>SSE_GFNI</code>}} - SSE-based Galois Field New Instructions
 +
* {{x86|AVX_GFNI|<code>AVX_GFNI</code>}} - AVX-based Galois Field New Instructions
 +
* Split Lock Detection - detection and cause an exception for split locks
 +
* Fast Short REP MOV
  
 
=== Block Diagram ===
 
=== Block Diagram ===
Line 116: Line 155:
 
==== Gen11 Graphics ====
 
==== Gen11 Graphics ====
 
See {{intel|Gen11#Block Diagram|Gen11 Graphics § Block Diagram|l=arch}}.
 
See {{intel|Gen11#Block Diagram|Gen11 Graphics § Block Diagram|l=arch}}.
 +
 +
== Overview ==
 +
[[File:ice lake overview.svg|right|500px]]
 +
The Ice Lake [[system on a chip]] is a [[10-nanometer]] SoC that is aimed at the mainstream to premium mobile and the thin-and-light market. The microprocessor consists of five major components: CPU cores, LLC, {{intel|ring interconnect}}, {{intel|system agent}}, and {{\\|Gen11}} graphics. While a lot of what Ice Lake provides is inherited from the prior generations, Intel claims that every IP on Ice Lake has been enhanced in one way or another. A major enhancement in Ice Lake over the prior generation is the integration of up to four {{\\|Sunny Cove}} cores which provide a significant uplift in IPC. Those cores also bring {{x86|AVX-512}} support for the client market. Those cores, along with the system agent and the GPU, are linked over Intel's {{intel|ring interconnect}}. The chip is fed through a new [[integrated memory controller]] that supports quad-channel 32-bit LPDDR4X memory, providing bandwidths in the range of 50-60 GB/s. Ice Lake has a new integrated GPU which is based on their {{\\|Gen11}} microarchitecture which provides a large improvement in graphics performance.
 +
 +
The system architecture in Ice Lake has been redesigned. Intel added a new Gaussian Neural Accelerator (GNA) for the acceleration of inference applications. There is a new 4th-generation [[image processing unit]] (IPU). There is a new Thunderbolt 3 integration. Ice Lake has four Thunderbolt 3 ports. All four ports have the same capabilities and can be used simultaneously at full performance at up to 40 Gbps per port. Intel also upgraded the display engine to {{\\|Gen11}} with an improved display pipe that has a new Adaptive Sync and HDR-capable display pipes that support HDR 3 and DisplayPort 1.4, supporting error correction and compression.
 +
 +
Ice Lake chips integrate the {{intel|PCH}} die on-package communicating over the on-package interconnect (OPI). The new PCH  The PCH has an expanded I/O support for PCIe, USB, SATA, as well as audio DSP for load power voice processing. It also integrates WiFi 6+ {{intel|CNVi}}.
 +
 +
On the platform level, there is a new integrated power delivery (FIVR) on both the PCH and the CPU which Intel says allows them to save on platform area by about 15% and it reduces the power delivery rails for the OEMs by roughly half.
 +
 +
== Core ==
 +
{{empty section}}
 +
 +
== Integration ==
 +
=== GNA ===
 +
Ice Lake introduced a new low-power [[neural processor]] called the '''Gaussian Neural Accelerator v1.0''' ('''GNA''') which is integrated on the SoC and runs at very low power even when the GPU and CPUs are turned off. The GNA can be used for long-running tasks (e.g., real-time meeting transcription). The GNA can operate while the remaining parts of the SoC are in idle in order to have minimal impact on performance.
 +
 +
=== IPU ===
 +
Ice Lake incorporates 4th generation [[image processing unit]] (IPU). The IPU was first introduced with {{\\|Skylake (client)|Skylake}} mobile SoCs (note that those were 3rd gen). The 4th Gen IPU found in Ice Lake introduces a number of new enhancements. It introduces new support for 4K video capture at 30fps. There is also new hardware support for better de-noising which supports up to 16 megapixels stills in low light conditions. In addition for support more camera simultaneously, the IPU incorporates a new concurrent image pipeline, supporting multiple different processing from the same camera stream, allowing a single camera to take the functionality of multiple sensors. A common example of that is devices with both IR and RGB cameras in the laptop bezel which can now be changed to a single camera. Intel says they are exposing more registers from the IPU to software in order to provide more flexibility for applications that make use of that for machine learning. It’s also worth noting that Intel integrated the MIPI interface onto the processor as well. Previously that was found on the chipset. The change significantly improves the latency, a required attribute needed for more advanced ML-specific applications. Some of those changes are designed to form the foundation for future generations of improvements.
 +
 +
== Clock domains ==
 +
Ice Lake is divided into a number of [[clock domains]], each controlling the clock frequency of their respective unit in the processor. All clock domains are some multiple of the [virtual] bus clock ([[BCLK]]).
 +
 +
* '''BCLK''' - Bus/Base Clock - The system bus interface frequency (once upon a time referred to the actual [[FSB]] speed, it now serves as only a base clock reference for all other clock domains). The base clock is 100 MHz.
 +
* '''Core Clock''' - The frequency at which the core and the [[L1]]/[[L2]] caches operate at. (Frequency depends on the model and is represented as a multiple of BCLK).
 +
* '''Ring Clock''' - The frequency at which the ring interconnect and [[L3$|LLC]] operate at. Data from/to the individual cores are read/written into the L3 at a rate of 32B/cycle operating at Ring Clock frequency.
 +
* '''IGP Clock''' - The frequency at which the [[integrated graphics]] ({{\\|Gen11}} GPU) operates at. Data from/to the GPU are read/written into the LLC at a rate of 64B/cycle operating at this frequency as well.
 +
* '''IPU''' - The frequency at which the [[image processing unit]] operates at
 +
* '''MemClk''' - Memory Clock - The frequency at which the system DRAM operates at. DRAM data is transferred at a rate of 8B/cycle operating at MemClk frequency.
 +
 +
[[File:ice lake soc clock domain block diagram.svg|850px]]
 +
 +
== Packaging ==
 +
Ice Lake comes in two packagings.
 +
 +
{| class="wikitable"
 +
|-
 +
! Package !! Type3 !! Type4
 +
|-
 +
| Core || {{intel|Ice Lake U|l=core}} || {{intel|Ice Lake Y|l=core}}
 +
|-
 +
| TDP || 15 W || 9 W
 +
|-
 +
| Dimensions || 50 x 25 x 1.3 mm || 26.5 x 18.5 x 1.0 mm
 +
|-
 +
| Balls || 1526 balls || 1377 balls
 +
|-
 +
| Ball Pitch || 0.65 mm || 0.43 mm
 +
|-
 +
| Package (Front) || [[File:ice lake u (front).png|250px]] || [[File:ice lake y (front).png|200px]]
 +
|-
 +
| Package (Back) ||  [[File:ice lake u (back).png|250px]] || [[File:ice lake y (back).png|200px]]
 +
|}
 +
 +
=== Thin-film magnetic inductor ===
 +
Interestingly the new packages include a thin-film magnetic inductor array on the landing side. Those are said to have higher efficiency at lower power but also support the fully processor dynamic frequency range. They can be distinctly seen on the back of the chip.
  
 
== Die ==
 
== Die ==
 
=== System Agent ===
 
=== System Agent ===
{{empty section}}
+
* System Agent
 +
** 4th Gen IPU
 +
** Gen11 Display
 +
** Thunderbolt 3 over Type-C I/O subsystem
 +
 
 +
 
 +
:[[File:ice lake die sa.png|700px]]
 +
 
 +
 
 +
:[[File:ice lake die sa (annotated).png|700px]]
 +
 
 +
==== IPU ====
 +
:[[File:ice lake die ipu.png|500px]]
 +
 
 +
 
 +
:[[File:ice lake die ipu 2.png|500px]]
 +
 
 +
==== Display engine ====
 +
:[[File:ice lake die display engine.png|500px]]
 +
 
 +
 
 +
:[[File:ice lake die display engine 2.png|500px]]
 +
 
 +
==== Thunderbolt 3 I/O subsystem ====
 +
:[[File:ice lake die tb3 io subsystem.png|500px]]
 +
 
 +
 
 +
:[[File:ice lake die tb3 io subsystem 2.png|500px]]
 +
 
 
=== Core ===
 
=== Core ===
{{empty section}}
+
{{see also|intel/microarchitectures/sunny_cove#Die|l1=Sunny Cove § Die}}
 +
* ~6.91 mm² die size
 +
** ~3.5 mm x ~1.97 mm
 +
 
 +
:[[File:ice lake die core.png|400px]]
 +
 
 +
 
 +
:[[File:ice lake die core (annotated).png|400px]]
 +
 
 +
 
 +
:[[File:ice lake die core 2.png|500px]]
 +
 
 
=== Core group ===
 
=== Core group ===
{{empty section}}
+
{{see also|intel/microarchitectures/sunny_cove#Die|l1=Sunny Cove § Die}}
 +
* ~30.73 mm² die size
 +
** ~7.86 mm x ~3.91 mm
 +
 
 +
 
 +
:[[File:ice lake die core group.png|700px]]
 +
 
 +
 
 +
:[[File:ice lake die core group (annotated).png|700px]]
 +
 
 +
 
 +
:[[File:ice lake die core group 2.png|700px]]
 +
 
 +
 
 
=== Integrated graphics ===
 
=== Integrated graphics ===
{{empty section}}
+
* {{\\|Gen11}} GPU
 +
* 64 EUs
 +
* ~41.1 mm² silicon area
 +
** ~5.22 mm x ~7.86 mm
 +
:[[File:ice lake die gpu.png|700px]]
 +
 
 +
 
 +
:[[File:ice lake die gpu (annotated).png|700px]]
 +
 
 +
 
 +
:[[File:ice lake die gpu 2.png|800px]]
 +
 
 
=== SoC ===
 
=== SoC ===
 
* [[10 nm process]]
 
* [[10 nm process]]
 +
* ~122.52 mm² die size
 +
** ~11.44 mm x ~10.71 mm
 
* 4 {{\\|Sunny Cove}} [[big cores]]
 
* 4 {{\\|Sunny Cove}} [[big cores]]
 
* 64-EU {{\\|Gen11}} GPU
 
* 64-EU {{\\|Gen11}} GPU
Line 133: Line 294:
  
  
:[[File:ice lake die (quad core).png|700px]]
+
:[[File:ice lake die (quad core).png|class=wikichip_ogimage|700px]]
  
  
 
:[[File:ice lake die (quad core) (annotated).png|700px]]
 
:[[File:ice lake die (quad core) (annotated).png|700px]]
 +
 +
 +
:[[File:ice lake die.png|800px]]
 +
 +
=== PCH ===
 +
* [[14 nm process]]
 +
* ~53.76 mm² die size
 +
** ~5.69 mm x 9.45 mm
 +
 +
 +
:[[File:ice lake pch die.png|700px]]
  
 
== All Ice Lake Chips ==
 
== All Ice Lake Chips ==

Revision as of 22:14, 4 June 2019

Edit Values
Ice Lake (client) µarch
General Info
Arch TypeCPU
DesignerIntel
ManufacturerIntel
IntroductionMay 27, 2019
Process10 nm
Instructions
ISAx86-64
Cache
L1I Cache32 KiB/core
8-way set associative
L1D Cache48 KiB/core
12-way set associative
L2 Cache512 KiB/512 KiB
12-way set associative
L3 Cache2 MiB/core
16-way set associative
Cores
Core NamesIce Lake Y,
Ice Lake U
Succession
Contemporary
Ice Lake (server)

Ice Lake (ICL) Client Configuration is Intel's successor to Cannon Lake, a 10 nm microarchitecture for mainstream workstations, desktops, and mobile devices.

Codenames

Core Abbrev Description Graphics Target
Ice Lake Y ICL-Y Extremely low power 2-in-1s detachable, tablets, and computer sticks
Ice Lake U ICL-U Ultra-low Power Light notebooks, portable All-in-Ones (AiOs), Minis, and conference room
Ice Lake H ICL-H High-performance Graphics Ultimate mobile performance, mobile workstations
Ice Lake S? ICL-S Performance-optimized lifestyle Desktop performance to value, AiOs, and minis

Process Technology

See also: Cannon Lake § Process Technology

Ice Lake is fabricated on Intel's second-generation enhanced 10 nm process called "10 nm+". Versus the first generation 10nm which was used for Cannon Lake, 10nm+ features higher performance through higher drive current for the same power envelope. Intel says that Ice Lake is built on their learnings from their Cannon Lake products which were largely treated as a learning vehicle. Between Cannon Lake and Ice Lake, a number of changes were made in order to improve the process for their products. One such change was the addition of an extra metal layer (originally said to be 12, is now presumably 13 on Ice Lake) in order to improve the power delivery of the chip. Additionally, they have improved the threshold voltage of the transistors as well as their MIM cap among other changes.

intels 10+ and 10++.png

Compiler support

Support for Ice Lake was added in LLVM Clang 6.0 and GCC 8.0.

Compiler Arch-Specific Arch-Favorable
ICC -march=icelake -mtune=icelake
GCC -march=icelake -mtune=icelake
LLVM -march=icelake -mtune=icelake
Visual Studio /? /tune:?

CPUID

Core Extended
Family
Family Extended
Model
Model
U, Y 0 0x6 0x7 0xE
Family 6 Model 126
 ? 0 0x6  ?  ?
Family 6 Model ?

Architecture

Ice Lake comprises of Sunny Cove cores on the ring interconnect architecture along with Gen11 GPU, and an improved System Agent with a new display engine and I/O.

Key changes from Cannon Lake/Skylake

  • Enhanced "10nm+" (from "10nm", 2nd gen)
  • Core
  • Memory
    • 4 32-bit LPDDR4X channels (from 2 64-bit DDR4 channels)
    • 1.4x higher data rates (3733 MT/s, up from 2666 MT/s)
      • 1.5x higher memory bandwidth (60 GB/s, up from 40 GB/s)
  • Graphics
    • Gen10Gen11 graphics (Gen10 was never productized)
    • Gen11 GPUs
      • UHD Graphics 6xx (GT1) UHD Graphics 9xx (GT2) (24 Execution Units, 2x EUs from Gen9)
      • UHD Graphics 6xx (GT2) Iris Plus Graphics 9xx (GT2) (48-64 Execution Units, 2-2.6x EUs from Gen9)
        • 1,024 GFLOPS @ 1 GHz (GT2)
  • Display
    • Gen 11.5 (from Gen9/Gen9.5)
    • DisplayPort 1.4a with Display Stream Compression(DSC) (from DisplayPort 1.2)
    • HDMI 2.0b (from HDMI 1.4)
  • IPU
    • 4th Gen IPU (from 3rd Gen in Skylake)
    • More cameras support
    • New concurrent image pipeline
    • on-die MIPI interface
  • New Integration
    • New Gaussian Neural Accelerator 1.0 (Added in Cannon Lake but unclear to what extent)
  • I/O
    • Thunderbolt 3 over Type-C
  • Package
    • New Type3, Type4 packages
      • New thin-film magnetic inductors

This list is incomplete; you can help by expanding it.

New instructions

See also: Sunny Cove § New Instructions

Ice Lake introduced a number of new instructions.

Block Diagram

Entire SoC Overview

ice lake soc block diagram.svg

Individual Core

See Sunny Cove § Block Diagram.

Gen11 Graphics

See Gen11 Graphics § Block Diagram.

Overview

ice lake overview.svg

The Ice Lake system on a chip is a 10-nanometer SoC that is aimed at the mainstream to premium mobile and the thin-and-light market. The microprocessor consists of five major components: CPU cores, LLC, ring interconnect, system agent, and Gen11 graphics. While a lot of what Ice Lake provides is inherited from the prior generations, Intel claims that every IP on Ice Lake has been enhanced in one way or another. A major enhancement in Ice Lake over the prior generation is the integration of up to four Sunny Cove cores which provide a significant uplift in IPC. Those cores also bring AVX-512 support for the client market. Those cores, along with the system agent and the GPU, are linked over Intel's ring interconnect. The chip is fed through a new integrated memory controller that supports quad-channel 32-bit LPDDR4X memory, providing bandwidths in the range of 50-60 GB/s. Ice Lake has a new integrated GPU which is based on their Gen11 microarchitecture which provides a large improvement in graphics performance.

The system architecture in Ice Lake has been redesigned. Intel added a new Gaussian Neural Accelerator (GNA) for the acceleration of inference applications. There is a new 4th-generation image processing unit (IPU). There is a new Thunderbolt 3 integration. Ice Lake has four Thunderbolt 3 ports. All four ports have the same capabilities and can be used simultaneously at full performance at up to 40 Gbps per port. Intel also upgraded the display engine to Gen11 with an improved display pipe that has a new Adaptive Sync and HDR-capable display pipes that support HDR 3 and DisplayPort 1.4, supporting error correction and compression.

Ice Lake chips integrate the PCH die on-package communicating over the on-package interconnect (OPI). The new PCH The PCH has an expanded I/O support for PCIe, USB, SATA, as well as audio DSP for load power voice processing. It also integrates WiFi 6+ CNVi.

On the platform level, there is a new integrated power delivery (FIVR) on both the PCH and the CPU which Intel says allows them to save on platform area by about 15% and it reduces the power delivery rails for the OEMs by roughly half.

Core

New text document.svg This section is empty; you can help add the missing info by editing this page.

Integration

GNA

Ice Lake introduced a new low-power neural processor called the Gaussian Neural Accelerator v1.0 (GNA) which is integrated on the SoC and runs at very low power even when the GPU and CPUs are turned off. The GNA can be used for long-running tasks (e.g., real-time meeting transcription). The GNA can operate while the remaining parts of the SoC are in idle in order to have minimal impact on performance.

IPU

Ice Lake incorporates 4th generation image processing unit (IPU). The IPU was first introduced with Skylake mobile SoCs (note that those were 3rd gen). The 4th Gen IPU found in Ice Lake introduces a number of new enhancements. It introduces new support for 4K video capture at 30fps. There is also new hardware support for better de-noising which supports up to 16 megapixels stills in low light conditions. In addition for support more camera simultaneously, the IPU incorporates a new concurrent image pipeline, supporting multiple different processing from the same camera stream, allowing a single camera to take the functionality of multiple sensors. A common example of that is devices with both IR and RGB cameras in the laptop bezel which can now be changed to a single camera. Intel says they are exposing more registers from the IPU to software in order to provide more flexibility for applications that make use of that for machine learning. It’s also worth noting that Intel integrated the MIPI interface onto the processor as well. Previously that was found on the chipset. The change significantly improves the latency, a required attribute needed for more advanced ML-specific applications. Some of those changes are designed to form the foundation for future generations of improvements.

Clock domains

Ice Lake is divided into a number of clock domains, each controlling the clock frequency of their respective unit in the processor. All clock domains are some multiple of the [virtual] bus clock (BCLK).

  • BCLK - Bus/Base Clock - The system bus interface frequency (once upon a time referred to the actual FSB speed, it now serves as only a base clock reference for all other clock domains). The base clock is 100 MHz.
  • Core Clock - The frequency at which the core and the L1/L2 caches operate at. (Frequency depends on the model and is represented as a multiple of BCLK).
  • Ring Clock - The frequency at which the ring interconnect and LLC operate at. Data from/to the individual cores are read/written into the L3 at a rate of 32B/cycle operating at Ring Clock frequency.
  • IGP Clock - The frequency at which the integrated graphics (Gen11 GPU) operates at. Data from/to the GPU are read/written into the LLC at a rate of 64B/cycle operating at this frequency as well.
  • IPU - The frequency at which the image processing unit operates at
  • MemClk - Memory Clock - The frequency at which the system DRAM operates at. DRAM data is transferred at a rate of 8B/cycle operating at MemClk frequency.

ice lake soc clock domain block diagram.svg

Packaging

Ice Lake comes in two packagings.

Package Type3 Type4
Core Ice Lake U Ice Lake Y
TDP 15 W 9 W
Dimensions 50 x 25 x 1.3 mm 26.5 x 18.5 x 1.0 mm
Balls 1526 balls 1377 balls
Ball Pitch 0.65 mm 0.43 mm
Package (Front) ice lake u (front).png ice lake y (front).png
Package (Back) ice lake u (back).png ice lake y (back).png

Thin-film magnetic inductor

Interestingly the new packages include a thin-film magnetic inductor array on the landing side. Those are said to have higher efficiency at lower power but also support the fully processor dynamic frequency range. They can be distinctly seen on the back of the chip.

Die

System Agent

  • System Agent
    • 4th Gen IPU
    • Gen11 Display
    • Thunderbolt 3 over Type-C I/O subsystem


ice lake die sa.png


ice lake die sa (annotated).png

IPU

ice lake die ipu.png


ice lake die ipu 2.png

Display engine

ice lake die display engine.png


ice lake die display engine 2.png

Thunderbolt 3 I/O subsystem

ice lake die tb3 io subsystem.png


ice lake die tb3 io subsystem 2.png

Core

See also: Sunny Cove § Die
  • ~6.91 mm² die size
    • ~3.5 mm x ~1.97 mm
ice lake die core.png


ice lake die core (annotated).png


ice lake die core 2.png

Core group

See also: Sunny Cove § Die
  • ~30.73 mm² die size
    • ~7.86 mm x ~3.91 mm


ice lake die core group.png


ice lake die core group (annotated).png


ice lake die core group 2.png


Integrated graphics

  • Gen11 GPU
  • 64 EUs
  • ~41.1 mm² silicon area
    • ~5.22 mm x ~7.86 mm
ice lake die gpu.png


ice lake die gpu (annotated).png


ice lake die gpu 2.png

SoC


ice lake die (quad core).png


ice lake die (quad core) (annotated).png


ice lake die.png

PCH


ice lake pch die.png

All Ice Lake Chips

Symbol version future.svg Preliminary Data! Information presented in this article deal with future products, data, features, and specifications that have yet to be finalized, announced, or released. Information may be incomplete and can change by final release.
 List of Ice Lake-based Processors
 Main processorTurbo BoostMemoryGPUFeatures
ModelLaunchedPriceFamilyPlatformCoreCoresThreadsL3$TDPBase1 Core2 Cores4 Cores6 CoresMax MemoryNameBaseBurstTBTHT
Count: 0

Bibliography

  • Intel 2018 Architecture Day.
  • Intel. personal communication. 2019.
codenameIce Lake (client) +
designerIntel +
first launchedMay 27, 2019 +
full page nameintel/microarchitectures/ice lake (client) +
instance ofmicroarchitecture +
instruction set architecturex86-64 +
manufacturerIntel +
microarchitecture typeCPU +
nameIce Lake (client) +
process10 nm (0.01 μm, 1.0e-5 mm) +