15 лет доступности - karma-group.ru · † Intel may support this extended...

2

15 лет доступности

IOTG is extending the product availability for IOTG roadmap products from a minimum of 7 years to a minimum of 15 years when both processor and chipset are on 22nm and newer process technologies.

- Xeon Scalable (w/ chipsets)

- E3-12xx/15xx v5 and later (w/ chipsets)

- 6th gen Core and later (w/ chipsets)

- Bay Trail (E3800) and later products (Braswell, N3xxx)

- Atom C2xxx (Rangeley) and later

- Не включает в себя Xeon-D (7 лет) и E5-26xx v4 (7 лет)

3

IOTG Product Availability Life-Cycle

L-1 L+1 L+3 L+5 L+7 L+9 L+11 L+13 L+15

7 year product availability

Last Time Order/Ship PeriodsPublication of the PDN will typically be followed by: 12 month last time order period, followed by 6 month last time ship period.

† Intel may support this extended manufacturing using reasonably feasible means deemed by Intel to be appropriate. Future business, manufacturing, or market conditions may change, which may cause Intel to revisit the EOL/PDN timing. Intel may make changes to dates and schedule at any time without notice

L L+2 L+4 L+6 L+8 L+10 L+12 L+14 Years

Introduction of component familyComponent family introduction dates are available from your Intel Sales Representative. Refresh SKU offerings launched after the component family will follow the original component family introduction date.

Last Time Order

La

st

Sh

ip

15 year product availabilityLast Time Order

La

st

Sh

ip

Product Discontinuance Notification (PDN)†

PDNs will typically be issued no later than 13.5 years after component family introduction date. PDNs are published at https://qdms.intel.com/

15 year product availability will start with the following products:• Intel® Xeon® Processor Scalable Family codenamed Skylake-SP and later with associated chipsets• Intel® Xeon® E3-12xx/15xx v5 series (Skylake) and later with associated chipsets• 6th Gen Intel® Core™ processor family (Skylake) and later (includes Intel® Pentium® and Celeron® processors) with

associated chipsets• Intel Pentium processor N3700 (Braswell) and later and Intel Celeron processors N3xxx (Braswell) and J1900/N2xxx

(Bay Trail) and later• Intel® Atom® processor C2xxx (Rangeley) and E3800 family (Bay Trail) and late

4

Какие модели будут доступны 15 лет?

https://ark.intel.com/

https://ark.intel.com/

5

Зачем использовать embedded parts?

- Долгая доступность (15 лет)

- Разные т.н. USE CONDITIONS:

Use Conditions Temp Range Reliability DPM

Automotive• Advanced Driver Assistance Systems (ADAS)• Software Defined Cockpit (SDC)

Automotive Grade 2, Tla = -40⁰ to 105⁰ C15Y Reliability at 11.4% Active,Tj-max = 110o - 125o C,

<10

Industrial CT – Commercial Temp: Tla = 0o to 70o C ET – Extended Temp: Tla = -40o to 85o C

10Y Reliability at 100% active, Tj-max = 100o - 110o CBase Frequency only – No core or graphics turbo

<500

CommunicationsCT – Commercial Temp: Tla = 0o to 70o CET – Extended Temp: Tla = -40o to 85o C

10Y Reliability at 100% active, Tj-max = 100o - 105o C

<500

Embedded Broad Market CT – Commercial Temp: Tla = 0o to 70o C5Y Reliability at 80% Active, Tj-max = 100o - 105o C

<500

Server/Enterprise CT – Commercial Temp: Tla = 0o to 70o C5Y Reliability at 80% Active, Tj-max sku dependent

<500

PC/Client/Tablet CT – Commercial Temp: Tla = 0o to 70o C5Y Reliability up to 30% Active, Tj-max sku dependent,

<500

АНАЛИТИКА,СУПЕРКОМПЬЮТЕРЫ,

УМНЫЕ ДАННЫЕОБЛАКА НОВЫЕ СЕТИ

AI is the fastest growing datacenter workload3

SDN/NFV spend up to$157B through 20202

18% CAGR from2017-20201

1. Source: IDC Q4’16 Cloud Infrastructure Tracker2. Source: Technology Business Research, Sept 20153. Source: Amalgamation of Intel financials, analyst data and Intel analysis, Intel revenue includes FPGAs.

6

7

«Строительные блоки» для решения текущих задач

Intel / Cloudera

Analytics & Technical Computing

SoftwareDevelopmentTools

Instructions

Intel® MPI Library

SW &Libraries

Intel® SHA Extensions

Intel ® TSX Instructions

…

Embree

OpenGL/DX on Intel® HD Graphics …

Intel ® TXT

Intel ® AES-NI Instructions

Intel ® AVX Instructions

Intel® MKL Library Intel® DPDK

Intel® SPDK

Intel® Data Analytics Acceleration Library (DAAL)

Machine Learning Software

Intel Scalable System Framework

Management, Orchestration & Storage

Intel® Parallel Studio XE

Intel® Media Server Studio

Intel® VTune

Intel® Inspector

Intel® CoFluent™ Wind River

Intel® ISA-L

Processors Chipsets EthernetIntel®

Omni-Path Fabric

NetworkASICs

SiliconPhotonics FPGAs SSDs

Intel® QAT Intel® DLIA

Accelerators

Intel® VCA

Intel® Data Center Manager

Intel® Virtual Gateway

Intel® Cache Accelration SW

Intel® RSTe

Intel® Xeon® Scalable ProcessorsФУНДАМЕНТ ДАТАЦЕНТРА БУДУЩЕГО:

ГИБКАЯ И НАДЕЖНАЯ АРХИТЕКТУРА

ПроизводительностьТам, где надо

Оптимизировано под все нагрузки: Compute, storage, Network

ПРОСТОТА ВНЕДРЕНИЯ

В среднем увеличение производительности в 1.6x раза против e5/e7 v46

99

УСКОРЕНИЕ СПЕЦ ВЫЧИСЛЕНИЙ

Application specific instructions and accelerators

▪ Intel® AVX-512 (2X FLOPs/clock) for HPC, Analytics, Database, & Security

▪ Int. Intel® QuickAssist Technology: faster crypto & data compression

▪ Intel® FPGA: Big speedups on server/storage/networking algorithms

ПАМЯТЬ И ШИНА PCI-E

High bandwidth connections for networking and memory performance

▪ More memory bandwidth via added memory channels

▪ Significantly more PCIe IO bandwidth versus prior gen

▪ Richer chipset & IO support options

ПРОИЗВОДИТЕЛЬНОСТЬ ЯДРАFaster application response timeMore choices to optimize software license costs

▪ New core micro-architecture

▪ New lower latency cache memory hierarchy design

ПРОИЗВОДИТЕЛЬНОСТЬ СОКЕТАIncreased multi-tenancy & VM density for cloudApplication performance for parallelizable workloads

▪ More cores

▪ New High-speed Interconnects = much greater inter-CPU bandwidth

* New features in bold

БЕЗОПАСНОСТЬBuilt in performant security for trusted infrastructure

▪ Up to 2X security algorithm performance improvements

▪ Root of trust with Intel® TXT-OTA, Intel® PTT and Intel® Boot Guard

▪ New Key Protection Technology (w/ integrated Intel® QAT & Intel PTT)

• Mode-based Execute Control (MBE): • Timestamp Counter Scaling (TSC):

ГИБКОСТЬOptimized architecture for Hybrid-Cloud world

▪ Enhanced Intel® RunSure Technology

▪ Enhanced Intel® Virtualization Technology (VT-x): MBE & TSC

▪ New Intel® VMD for improved management of NVMe SSD drives

10

«СЕТЕВАЯ» архитектура:

▪ Максимальная производительность

▪ Гарантированно низкие задержки при передачи данных

▪ Оптимизировано для работы с общей памятью

▪ Масшабируется эффективно для 2, 4, & 8+ сокет конфигур

МО

НО

ЛИ

ТН

ОЕ

ЯД

РО

Много вычислительных ядер, быстрая память, малые задержки

Note: Not representative of top bin die

Enhanced PCIe

11

Intel® Xeon® Processor E5Сбалансированные 2х сокет системы с лучшим соотнош performance/watt/$

Intel® Xeon® Processor E7High-end сервера с масшабированием по сокетам и поддержкой RAS

Grantley-EP Platform

E5 v3 E5-2600 v4

Brickland Platform

E7 v3 E7 v4

Purley Platform

Skylake

E5 v3 E5-4600 v4 (4S)

Cascade Lake

2016 2017 2018

Intel Xeon GOLD

Intel® Xeon® PLATINUM

Intel Xeon SILVER

Intel Xeon BRONZE

Общая платформа для построения E5 и e7 – class решений

12

INTEL® XEON® SCALABLE PROCESSORThe Foundation for Agile, Secure Workload-Optimized Hybrid Cloud

MAINSTREAM

BETTER

DEDICATED TASKS

GREAT SCALABLE PERFORMANCE MEMORY PERFORMANCEADVANCED RAS

SCALABLE PERFORMANCE AT LOW POWER ENTRY

SCALABLE PERFORMANCE HARDWARE-ENHANCED SECURITYSTANDARD RASSTANDARD RAS

MODERATE TASKSWORKLOAD-OPTIMIZED

PERFORMANCE FOR GENERAL-PURPOSECOMPUTE, STORAGE AND NETWORKING

TASKSINTEL® TURBO BOOST TECHNOLOGY AND INTEL® HYPER-THREADING TECHNOLOGY

FOR MODERATE WORKLOADSAFFORDABLE, ENTRY PERFORMANCEFOR LIGHT-RANGE WORKLOADS

28 CORESUP TO

SOCKETSUPPORT8

1.5 TB TOPLINE MEMORYCHANNEL BANDWIDTH

3 LINKSUPIUP 2, 4 &

2666DDR4 MHZ

WITHUP TO

TOWITHUP TO

HIGHEST ACCELERATORTHROUGHPUT

INTEL® XEON® PLATINUM 81xx PROCESSORS• Most cores (28C, 56T)• Highest frequency (up to 3.6 GHz, 4C)• Most Socket flexibility (2, 4, 8+)• Elite IO/memory (48 PCIe 3.0 lanes, 6 memory channels)• Most/fastest 3 UPI (Ultra Path Interconnects, 10.4 GT/s)• Fastest Memory (DDR4-2666 MHz)• Highest Intel® AVX-512 with 2 FMA per core• Intel® Turbo Boost and Intel® Hyper-Threading Technology• New Advanced RAS• Node Controller Support to assist in scaled node management

RESULT:• Best workload-optimized performance for general purpose compute across virtualized

environments and hybrid cloud deployments • Best choice for mission-critical, traditional DC apps, real-time analytics, and AI/deep

learning inference workloads; smart path for DL training• Best performance for the most demanding storage and networking workloads

RAS: Reliability, Availability, and Serviceability

13

INTEL® XEON® SCALABLE PROCESSORThe Foundation for Agile, Secure Workload-Optimized Hybrid Cloud

MAINSTREAM

BETTER

DEDICATED TASKS

SCALABLE PERFORMANCE AT LOW POWER ENTRY

SCALABLE PERFORMANCE HARDWARE-ENHANCED SECURITYSTANDARD RASSTANDARD RAS

MODERATE TASKS

INTEL® TURBO BOOST TECHNOLOGY AND INTEL® HYPER-THREADING TECHNOLOGY

FOR MODERATE WORKLOADSAFFORDABLE, ENTRY PERFORMANCEFOR LIGHT-RANGE WORKLOADS

22 CORESUP TO

SOCKETSUPPORT2 & 4

3UP TO UPI LINKS

RELIABILITY, AVAILABILITYAND SERVICEABILITYADVANCED

28 CORESUP TO

SOCKETSUPPORT8


3 LINKSUPIUP 2, 4 &

2666DDR4-

MHZ

WITHUP TO

TOWITHUP TO


INTEL® XEON® GOLD 61xx PROCESSORS

INTEL® XEON® GOLD 51XX PROCESSORS

• 22C (44 threads w/ HT)• 2 and 4S• Up to 3.4 GHz, (6C)• 3 UPI links (10.4GT/s)• DDR4-2666 MHz• Intel® AVX-512 with 2 FMA• Node Controller Support• Enhanced scalability and workload optimized performance for

general purpose/virtualized compute & hybrid cloud• Improved performance for demanding storage and networking

• Up to 14C (28T)• 2 UPI links• DDR4-2400 MHz• Intel® AVX-512 with 1 FMA• Affordable Advanced RAS and 4-socket scalability• Suitable for a wide range of workloads

14

INTEL® XEON® SCALABLE PROCESSOR

MAINSTREAM ENTRYEFFICIENT

28 CORESUP TO

SOCKETSUPPORT8


3 LINKSUPI

22 CORESUP TO

SOCKETSUPPORT2 & 4

3UP TO UPI LINKS

UP 2, 4 &

ADVANCED RELIABILITY, AVAILABILITYAND SERVICEABILITY

12 CORESUP TO

SOCKETSUPPORT2

2666DDR4-

MHZ

WITHUP TO 2400DDR4 M

HZ

INTEL TURBO BOOST TECHNOLOGYHYPER-THREADING TECHNOLOGY

8 CORESUP TO

PCIE3.048

LANES

STANDARD RELIABILITY, AVAILABILITYAND SERVICEABILITY

TOWITHUP TO


SOCKETSUPPORT2

INTEL® XEON® SILVER 41xxPROCESSORS

INTEL® XEON® BRONZE 31XX PROCESSORS

Up to 12C, 24TUp to 2.2 GHz (10C)

DDR4-2400 MHz2 UPI at 9.6 GT/s

Intel® AVX-512 w/ 1 FMAIntel® Turbo Boost and Intel® HT for multi-threaded apps

Suitable for a moderate range of workloads

Up to 8CUp to 1.7GHz (8C)

Entry 2-socket configurationDDR4-2133 MHz2 UPI at 9.6 GT/s

Intel® AVX-512 w/ 1 FMAReliability upgrade versus Intel® Xeon® processor E3 platform

2S Purley Platform Configurations

DDR4 DIMMsDDR4/Apache Pass (Apache Pass support planned for Cascade Lake Server)PCIe* uplink connection for Intel® QuickAssist Technology and Intel® EthernetExample DIMM population shown; please look up Apache Pass customer collateral for specific rules on DDR4/Apache Pass DIMM populations

**

SKL/CLX

Intel®UPI

LBG

DMI

3x16 PCIe*

x4

3x16 PCIe*

1x100G Intel® Omni-Path Fabric

SKL/CLX

**

2S-2UPI Configurationwith integrated Intel® Omni-Path Fabric

15

1x100G Intel® Omni-Path Fabric

4S & 8S Purley Platform Configurations

4S-3UPI Configuration

SKL/CLX

SKL/CLX

LBGLBG

LBGDMI

3x16 PCIe*

SKL/CLX

SKL/CLX

SKL/CLX

SKL/CLX

LBG

LBG

LBG

LBG

DMI

SKL/CLX

SKL/CLX

SKL/CLX

SKL/CLX

SKL/CLX

SKL/CLX

3x16 PCIe*

8S Configuration

**

DDR4 DIMMsDDR4/Apache Pass (Apache Pass support planned for Cascade Lake Server)OptionalPCIe* uplink connection for Intel® QuickAssist Technology and Intel® EthernetExample DIMM population shown; please look up Apache Pass customer collateral for specific rules on DDR4/Apache Pass DIMM populations

LBG

4S-2UPI Configuration

SKL/CLX

SKL/CLX

LBGLBG

LBGDMI

3x16 PCIe*

Intel®UPI

SKL/CLX

SKL/CLX

LBG LBG

** **

Intel®UPI

Intel®UPI

**

16

3x16 PCIe*

3x16 PCIe*

3x16 PCIe*

17

Product Numbering Convention for Intel® Xeon® Scalable Processors

Processor Generation• 1 = 1st Gen (Skylake-SP)

Processor SKU • (ex. 20, 34…)

Integrations and Optimizations (if applicable)

• F = Fabric • T = High Tcase/Extended Reliability

Memory Capacity• No Suffix = 768GB per socket• M = 1.5TB per socket

SKU Level• 8 = Platinum• 6, 5 = Gold• 4 = Silver• 3 = Bronze

8 1 # #

Intel® Xeon® Gold 5 1 # #

6 1 # #

Intel® Xeon® Silver 4 1 # #

Intel® Xeon® Bronze 3 1 # #

processor

processor

processor

processorIntel® Xeon® Gold

a

a

a

a

a

a

a

a

a

a

Intel® Xeon® processor E7 Family (4/8S+)

Intel® Xeon® processor E5 Family (2S, 4S)

processorIntel® Xeon® Platinum

Extended life (10-Year Use) + NEBS-Friendly Thermal Spec

with integrated Intel® Omni-Path Architecture Fabric

12C-24.75M 165W3.2G

24C 160W2.1G 8160F

20C 135W2.0G 6138F

16C 135W2.1G 6130F

28C 173W2.1G 8176F

20C 150W2.4G 6148

28C 205W2.5G 8180

Top Bin

16C 150W2.6G 6142

8C-24.75M 130W3.2G 6134

6C-19.25M 115W3.4G 6128

4C-16.50M 105W3.6G 8156

12C-19.25M 125W2.6G 6126

14C 140W2.6G 6132

18C 165W2.7G 6150

18C 200W3.0G 6154

24C 205W2.7G 8168

8C-24.75M 150W3.5G 6144

8C 70W2.0G 4109T

12C 85W2.1G 4116T

14C 85W1.9G 5119T

14C 105W2.2G 5120T

16C 125W2.1G 6130T

20C 125W2.0G 6138T

24C 150W2.1G 8160T

10C 85W2.2G 4114T

4C-16.50M 105W3.6G 5122

12C-24.75M 150W3.0G 6136

12C-24.75M 150W3.0G 8158

12C-19.25M 125W2.6G 6126T

SK

Us

OP

TIM

IZE

D F

OR

HIG

HE

ST

PE

R-C

OR

E P

ER

FO

RM

AN

CE

16C 160W2.6G 6142F

12C-19.25M 135W2.6G 6126F

20C 160W2.4G 6148F

24C 150W2.1G 8160

26C 150W2.0G 8164

22C 140W2.1G 6152

20C 125W2.0G

6138

18C 140W2.3G

6140

16C 125W2.1G 6130

14C 105W2.2G 5120

12C 105W2.3G 5118

12C 85W2.1G 4116

10C 85W2.2G 4114

8C 85W2.1G 4110

8C 85W1.7G 3106

6C 85W1.7G 3104

8C 85W1.8G

28C 165W2.1G 8176

26C 165W2.1G 8170

4C 85W2.6G 4112

16C 125W2.0G (8S entry) 8153

10C 85W2.4G 5115

• Highly flexible offerings optimized to address broadest array of workloads

• Four primary SKU lineups:

1. Optimized for per core performance

2. Balanced, energy efficient perf/W

3. Extended life

4. Integrated Intel® Omni-Path

• New 205W SKUs:

• e.g. 8180 hits 2.5 GHz at 28 cores

• Great high frequency options:

• e.g. 6144 w/ 8 cores at 3.5 GHz base

• SKUs w/ add’l cache per core vs. default

Core Count TDP

Base Non-AVX Core Frequency

SKU #“M” SKU available (1.5 TB/Sck)

OP

TIM

IZE

D F

OR

BA

LA

NC

ED

EN

ER

GY

EF

FIC

IEN

T P

ER

F/W

Core Count TDP

Base Non-AVX Core Frequency

SKU #“M” SKU available (1.5 TB/Skt)

21

4108

12C-24.75M 165W3.2G

6146

18

Volume Management Device (VMD)

EthernetControllerw/ crypto accelerator & RDMA

20

Intel® Ethernet 700 Series • One Architecture. Multiple Speeds

• int. Intel Ethernet Connection X722 w/ iWARP RDMA (4x10GbE)

• Advanced features • Optimized SW (DPDK) for NFV

Int. Intel® QuickAssist Technology• Crypto & Data Compression

Intel® Omni-Path Architecture (Fabric)• High-bandwidth & low-latency fabricfor HPC clusters (100 Gb / sec)

2.3 ZB GlobalIP TrafficAnnually by 2020

26B IP networked devices/Connectionsby 2020 up to 2.55x higher

IPSec (Gbps)

Приватность данных

up to 2.49x higher

TLS Web Proxy (Gbps)

Шифрация передачи

Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components,

SW, operations and functions. Any change to any of those factors may cause the results to vary. Consult other information/performance tests to assist you in fully evaluating your contemplated purchases, including perfoof that product when

combined with other products. For more information go to http://www.intel.com/performance Intel does not control or audit the design or implementation of third party benchmark data or Web sites referenced in this document. Intel encourages all of its customers to visit the referenced Web sites or others where similar performance benchmark data are reported and confirm whether referenced benchmark data are accurate/reflect performance of systems available for purchase.

Новая платформа для 5g-ready сетей

http://www.intel.com/performance

21

NC-SI

SMBus

4x10GbE/1GbEwith RDMA(iWARP)

Integrated IO

PCIe* with SR-IOV

Queue Mgmt & Scheduling

Qm

In-Band

Mgmt.

PF0 PF1 PFn…

...

4x10G MACs

VF0VF0VF0VF0VF0VF1

VF0VF0… VF0VF0VFm

MCTP

VEB & DCB Traffic Classifiers

Protocol Engine

Q0 Q1

Feature Support Customer Value

Quad Port 10GbE MAC/PHY• Based on Intel’s 10GbE solution• Interfaces: 10G (KR, SFI, XFI), 1G (KX)

• Optimized for networking capability in Cloud, Comms, and Storage

• Single network driver on Intel® platform

Remote Direct Memory Access • iWARP

• Routable and scalable RDMA ideal for large segmented networks in private and public clouds

Network Virtualization Offloads• Flexible Filters (ATR, Flow Director)• NVGRE, IPinGRE, VXLAN, MACinUDP

• Abstract the network for cloud flexibility

• Enhanced programmability and application affinity

Standards Based Virtualization • SR-IOV: 4 Physical/128 Virtual Function• VEB (Virtual Ethernet Bridge)

• Broad OS enablement

Power Management • Energy Efficient Ethernet

Manageability• BMC Pass-Through (control & network)• Interfaces: NC-SI, SMBus, and MCTP

• Common transport for LAN-to-BMC

RDMA is a network performance optimization

•Enables direct app-to-app communication across nodes

•Bypasses the OS stack & kernel

•Provides direct channel for remote memory application access

RDMA offers lower latency, high throughput by:

•Avoiding application context switching

•Placing data directly in application buffers (zero-copy DMA)

•Moving protocol processing off the CPU

Intel® Ethernet Connection X722 is featured with iWARP (Internet Wide-area RDMA Protocol) RDMA:

•Recommended for easy deployment and configuration, scalability and congestion control

RDMA-aware applications or messaging layers are required

•Windows*/Linux* Drivers supported for Storage, Messaging Middleware and HPC application categories

22

Ethernet

iWARP

PCIe

Sys Driver

Application

OS Stack

IO Lib

S/W

H/W

BasicNetwork

ControllerFlow

TCP/IP

Kernel

User

AvoidCopies

& Kernel

iWARPNetworkControllerFlow

WithoutiWARP

WithiWARP

Intel Confidential 23

Intel® Ethernet Architecture Evolution

Fixed Function

Customizable

500 Series 700 Series 800 Series 900 SeriesNiantic Fortville Columbiaville** Gemini Mountain**

Fixed Pipeline

PartiallyProgrammable Pipeline

Programmable Pipeline

ExtensiblePipeline

10GbE 10/25/40GbE 10/25/50/100GbE 10/25/50/100/200GbE

FPGA CPU Cores

Port Speed

Internal Fabric

Pro

gra

mm

ab

ilit

y

**These programs are in the planning stage. Features are subject to change.

Laying the Foundation for supporting a broader portfolio of network protocols

Wo

rklo

ad

Highly Advanced/Proprietary

Traditional (General IT)

Intel Confidential

2015 2016 2017Lewisburg Integrated

4x10GbE with RDMA

2014

XL710 (40GbE QSFP+)Network VirtualizationOverlays Acceleration

X520 10GbE SFP+World’s Best Selling 10GbE CNA

X540 10GBASE-TWorld’s 1st

Single Chip 10GBASE-T

. . .

XXV710 (25GbE SFP28)Cloud and Network Virtualization Overlays

X550 10GBASE-T2nd Generation Single Chip 10GBASE-T

Broadwell DE Integrated2x10GbE

Denverton Integrated4x10GbE

Avoton Integrated4x1/2.5GbE

X710 (10GBASE-T)Quad-Port 10GBASE-T

X710 (10GbE SFP+)Cloud and Network Virtualization Overlays

Columbia Park2x100GbE with RDMA

E810 (100GbE QSFP28)Enhanced Virtualizationand Network Overlays

2019

Ethe

rnet

Adap

ters

IP / S

oCIntel® Ethernet Adapter Roadmap Summary

Intel Ethernet 700 Series

24

Intel Ethernet 800 Series

2018

X722 (10GbE SFP+)RDMA and Network Virtualization Overlays

25

Intel® VMD is a CPU-integrated device to aggregate NVMe SSDs into a storage volume and enables other storage services such as RAID

• Intel® VMD is an “integrated end point” that stops OS enumeration of devices under it

• Intel® VMD maps entire PCIe* trees into its own address space (a domain)

• Intel® VMD driver sets up and manages the domain (enumerate, event/error handling), but out of fast IO path

Eliminates additional components to provide a full-feature storage solution

IntelVMD

IntelVMD

IntelVMD

Intel Confidential – For Use Under NDA OnlyDatacenter Group

ПОДДЕРЖКА 24 NVMe на примере 2U 2S платформы

GRANTLEY to PURLEY CHANGE

2U 8x25 Combo HSBP now supports NVMe in all 8 Drive Slots

OCuLink being used for NVMe vs. miniSAS HD

More options for PCIe SSD AIC

Intel Confidential – For Use Under NDA OnlyData Center Group

SPDK Architecture

Drivers

StorageServices

StorageProtocols

iSCSI Target

NVMe-oF*Target

SCSI

vhost-scsiTarget

NVMe

NVMe Devices

Blobstore

NVMe-oF*

Initiator

Intel® QuickDataTechnology

Driver

Block Device Abstraction (BDEV)

Ceph RBD

Linux Async IO

Logical Volumes

3rd Party

NVMe

NVMe*

PCIe Driver

Released

New release 17.10

Jan ‘18

vhost-blkTarget

BlobFS

Integration

RocksDB

Ceph

Core

ApplicationFramework

GPT

PMEMVirtio scsi

(vhost) BDev QoSQEMU


SPDK Community

• SPDK community webpage: http://www.spdk.io/community/

• Github : https://github.com/spdk/spdk - Centralized place for the code management

• Trello : https://trello.com/spdk - Place to find SPDK’s ongoing or future work

• GerritHub : https://review.gerrithub.io/#/q/project:spdk/spdk+status:open -SPDK Code Reviews

• IRC: https://freenode.net/ we’re on #spdk – Post technical questions

• Mailing List: https://lists.01.org/mailman/listinfo/spdk - Available for usage

and development questions as well as design discussions

http://www.spdk.io/community/

https://github.com/spdk/spdk

https://trello.com/spdk

https://review.gerrithub.io/#/q/project:spdk/spdk+status:open

https://freenode.net/

https://lists.01.org/mailman/listinfo/spdk

29

БЫСТРЫЙ И НАДЕЖНЫЙ• Быстрые Intel® Optane™ DC ssds

• емкие intel® 3d nand DC SSds

• Больше PCIe ввода вывода

• Intel® Storage ПО• Intel® Intelligent Storage Acceleration Library (ISA-L)

• Intel® Storage Performance Development Kit (SPDK)

• Intel® Platform Storage расширения• Intel® Volume management device (intel® VMD)

• Intel® virtual raid on cPU (Intel® VROC)

6x Optane SSD + Intel® SPDK

Up to 5.2x higher IOPSUp to 3.3x lower latency

▪ Up to 60X better at 99% QoS1

▪ Up to 40X faster response time1

1 Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. For more complete information visit http://www.intel.com/performance Configurations: Tested by Intel as of Jun 2017; test versus out of box 4x NVMe SSD

Новая платформа для гибридных облачных STORAGE инфраструктур

http://www.intel.com/performance

32

ИТАК,Революция в использовании данных меняет ВСЕ индустрии

Новые датацентры должны быть мощнее, безопаснее и более конфигурируемы, чтобы справиться с требованием бизнеса

Новая платформа представляет мощность, безопасность и конфигурируемость (compute / storage / network)

Общий сокет и стабильность платформы (Skylake / Cascade Lake / Ice Lake = один сокет LGA 3647) дает возможности масшабирования


Purley with Cascade Lake OverviewCPU

Cascade Lake: Up to 28Core with Intel HT Technology(Drop in to Purley Platform @ 70W-205W)

New CapabilitiesFrequency and architecture improvements and

DDR-T/Apache Pass DIMM Support on select SKUs

Socket Socket P

Scalability 2S, 4S, & glueless 8S (>8S via xNC support)

Memory

6 channels DDR4 RDIMM and LRDIMM per CPU/ / 12 DIMMs per socket, 16Gb DDR4 based DIMMs supported, up to 2666 MT/s

2DPC, up to 2933 MT/s 1DPC and

Apache Pass DIMM Support on select SKUs

UPI(Ultra Path Interconnect)

Up to 3 links per CPU x20, speed: 9.6 and 10.4 GTS

PCIe PCIe Gen 3: 48 lanes per CPU (bifurcation support: x16, x8, x4)

Host FabricDiscrete Omni Path adapter (100Gb/s) [Integrated Fabric SKUs

available on Skylake only]

FPGA Integrated Intel Arria® 10 FPGA in multi-chip package

PCH – LewisburgIE, Intel QAT, eSPI, Integrated Intel Ethernet Connection: up to

4x10Gb/1Gb ports, Up to 14 SATA 3, Up to 14 USB 2.0, Up to 10 USB 3.0, Up to 20 ports PCIe* 3.0 (8 GT/s)

New capabilities/Changes relative to Skylake/Purley in Bolded Blue

Purley 4S Configuration

Cascade Lake

Cascade Lake

UPI

DMI DMI

DDR4 or AEP DIMM DDR4 DIMM

Cascade Lake

Cascade Lake

UPI

48 lanesPCIe

Purley 2S Configuration

Cascade Lake

Cascade Lake

UPI

DDR4/AEP DIMM DDR4/AEP DIMM

FPGAFPGA

LBG

DMI

LBG

LBG

DMI

LBG LBG

LBG

DDR4 or AEP DIMM DDR4 DIMM

48 lanesPCIe

48 lanesPCIe

48 lanesPCIe

48 lanesPCIe

48 lanesPCIe

DDR4/AEP DIMM

DDR4/AEP DIMM

DDR4/AEP DIMM

DDR4/AEP DIMM

33

Intel Confidential – For Use Under NDA OnlyData Center Group 34

Apache Pass Production System Requirements

CascadeLake CPU(AEP-enabled

stepping)

Lewisburg PCH

BIOS with Reference

Code

• Provides cache & memory control

Platform Requirements

Software Requirements2

Block Driver and Management Libraries

Persistent Memory Aware File System

• No board changes required1

Purley Platform

NEW NEW

NEWNEW NEWNEW

1. No board changes required assuming OEM system was designed to be compliant with PDG Guidelines2. Dependent on operating mode

Apache Pass

NEW


Cascade Lake VNNI Instruction Set

Vector Neural Network Instruction (VNNI) on Cascade Lake will accelerate Intel® Xeon® Scalable processor performance on Deep Learning and AI workloads

• VNNI : A new set of Intel® Advanced Vector Extension (Intel® AVX-512) instructions for Intel® Xeon® Platinum 8000 and Gold 6000 series processors

• 8-bit (int8) new instruction

• 16-bit (int16) new instruction

• No hardware changes are required to support VNNI on Purley Platform

• Minimal OS/VMM enabling if Intel® AVX-512F (foundation) state pre-exists

• SW development support will be enabled in MKL-DNN (Math Kernel Library – Deep Neural Networks)

35


*Future product that was formerly codenamed the “Crest family”All performance positioning claims are relative to other processor technologies in Intel’s AI datacenter portfolioAll products, computer systems, dates, and figures specified are preliminary based on current expectations, and are subject to change without notice.

Datacenter ai

Intel® Xeon® Processor Scalable FamilyIntel® Stratix® 10 FPGA

Flexible accelerationAccelerate the widest range of AI and other

workloads & configurations

Foundation for AIBegin your journey with

the AI you need on the chip you know

Deep learning by designAccelerate the most intensive

deep learning deployments with this custom-built processor

Intel® Nervana™ Neural Network Processor*


▪ Coherent interface benefits software developers▪ Superior performance for bandwidth & latency sensitive applications

▪ System flexibility with Intel Xeon CPU SKU options▪ Dedicated local memory▪ Can be slotted into 1U servers

Choose the Intel FPGA platform matched to your application needs

371UPI = Intel® Ultra Path Interconnect2HSSI = High Speed Serial Interface

Intel® FPGA Data Center Platform OptionsEnabled By The Acceleration Stack for Intel® Xeon® CPU with FPGAs

PCIe Acceleration Cards Server Platform Option with In-Package FPGA


Learn More, Get Started Today

38

Resource Location

Intel® Acceleration Solutions Site www.intel.com/accelerators

Intel® FPGA Acceleration Hub www.intel.com/fpgaaccelerationhub

Learn more about the Acceleration Stack www.altera.com/solutions/acceleration-hub/acceleration-stack.html

Start developing for Intel FPGAs using OPAE 01.org/OPAE

(Video) “Getting Started With Open Programmable Acceleration Engine”

www.brighttalk.com/webcast/10773/275799

(Whitepaper) “The Open Programmable Acceleration Engine (OPAE)”

01.org/sites/default/files/downloads/opae/open-programmable-acceleration-engine-paper.pdf

http://www.intel.com/accelerators

http://www.intel.com/fpgaaccelerationhub

http://www.altera.com/solutions/acceleration-hub/acceleration-stack.html

http://01.org/OPAE

http://www.brighttalk.com/webcast/10773/275799

https://01.org/sites/default/files/downloads/opae/open-programmable-acceleration-engine-paper.pdf

39

Xeon D

Broadwell (-DE, -DE NS)

Skylake (-D)

Intel ConfidentialIntel and the Intel logo are trademarks or registered trademarks of Intel Corporation or its subsidiaries in the United States and other countries. Other names and brands maybe claimed as the property of others. All products, dates, and figures are preliminary and are subject to change without any notice. Copyright © 2015, Intel Corporation. 40

Intel® Xeon® Processor D Workload Positioning

Data center edge

workloads:High density at

low power resulting in good TCO

Enterprise Public Cloud HPC Big Data Storage Network

Data Mining Search CAE, Modeling Simulation

Real TimeAnalytics

Media Processing

BusinessProcessing

VirtualizedHosting

EDA Batch Analytics

Scale-out / distributeddatabase

Deep Packet Inspection

Collaboration,Email

Content Delivery

Digital Content Creation

No SQLdatabase

Enterprise SAN / NAS

Edge Routing

Application Development

Dynamic Web Serving

Financial Risk Analysis

Data caching Hot Storage Edge Security / Firewall

File & Print, Security, Sys Management

Dedicated Hosting

Gov’t / Defense / Academic

Scale up and distributed database

WarmStorage

Wireless Access

Dynamic Web Serving

Memory Caching

Geo Sciences; Siesmic

Large objectstore

Cold Storage Wireless Base Station

Scales to low TDP, high

performance, 10 GbE and PSE

Intel® Confidential – For Use Under NDA OnlyDatacenter Group 41

Next Gen*

Broadwell-DE-NSBroadwell-DE + 40G

QAT** + 2x10GbE

Intel® Xeon® Processor D Product Cadence

Density

14nm Process Technology

Broadwell SkylakeNEW Intel

Microarchitecture


HaswellNEW Intel

Microarchitecture

Broadwell-DE

Skylake-D100G QAT**,

4x10GbE

Q1 ’15,Q4 ‘15

Q3 ’17

Q1 ’18

Future Xeon®-D *


Future

NEW IntelMicroarchitecture

• *Product in Planning• ** Intel® QuickAssist Technology

Integration Acceleration

Intel ConfidentialIntel and the Intel logo are trademarks or registered trademarks of Intel Corporation or its subsidiaries in the United States and other countries. Other names and brands maybe claimed as the property of others. All products, dates, and figures are preliminary and are subject to change without any notice. Copyright © 2015, Intel Corporation.

Intel® Xeon® Processor D-1500 Platform Overview

CPU• Up to 16C BDW Xeon© SMT (14nm) • Targeted TDP ~20W - 65W • Intel® Xeon® Features

Other Features• Storage/Network Environment

(Reliability, Temp, Availability)• Validated with external Crypto

accelerator (Coleto Creek)• Non-Transparent Bridging (NTB),

Asynchronous DRAM self-refresh (ADR), Intel® QuickData Technology

Memory • 2 Memory Channels,

2 DIMMS/channel• DDR3L/DDR4 • 128GB Max Capacity

Integrated IOs• x24 PCIE 3.0, x8 PCIE2.0• x6 SATA3• x4 USB 3.0, x4 USB 2.0• x2 10 GbE Intel® Ethernet

BroadwellCores

Ad

van

ced

IO

s

Server Uncore

DDR4/3L128GB Max

x2 10GbEIntel® Ethernet

x24 PCIE3

x4 USB3

x8 PCIE2

x6 SATA 3

Intel®Switches

ECC UDIMM, SODIMM, RDIMM

FW, BIOS,Drivers

BIOS

EDK2

CS Drivers

LAN Drivers

OS/VMMEnabling/Testing

Intel® HW Intel® SW 3rd Party HW

SPS Firmware

x4 USB2

Intel® Xeon® Processor D-1500 Family

Legacy IOs

Intel® Communications

Chipset 89xx Series(Coleto Creek)

До 16 ядерДо 2.7GHz

Intel® Confidential – For Use Under NDA OnlyDatacenter Group

Intel®

Xeon™ Core

43

Xeon® D-2200 (Bakerville/Skylake-D) OverviewCPU Skylake-D

Socket & TDP FCBGA 45 mm x 52.5 mm; Up to 110W

Scalability 1S SoC

Cores Up to 18C

Memory

Up to 4 DDR4 channels* (up to 2DIMMs / channel)

RDIMM/LRDIMM, 1-2 DPC @2133-2666MT/s

Up to 512GB memory capacity

Ethernet/LANUp to 4 ports 10 GbE/1GbE Integrated Ethernet w/ iWARP

RDMA (Supports KR/KX/SFI)

Accelerators (QAT)

Crypto: Up to ~100 Gb/s IPSec/SSL, Compression: Up to ~100Gb/s,

Public Key: Up to RSA2K 100K Ops

PCIe x32 PCIe Gen3 + up to 20 PCIe Gen3 lanes via Flexible IO

Flexible IO

20 High-Speed Flexible Lanes configured as PCIe or SATA or USB3

Up to 20 lanes of PCIe 3.0 (Bifurcation support: x4, x2) or

Up to 14 lanes of SATA Gen 3

Up to four ports of USB Gen 3

Other IO Up to 4 USB 2.0, eSPI/LPC, SMBus, Integrated Clocking

Manageability Intel® Manageability Engine (Intel® ME), Innovation Engine (IE)

Block Diagram

New capabilities relative to BDW-DE in Bolded Blue*Varies with SKUs

Ethernet(4x1, 4x10)

Intel® ME

Innovation Engine

LegacyIO

SATA3 USB3.0

Intel®

Xeon™ Core1MB

L2

QAT

20 Flexible HSIO Lanes

x20 x14 x4

x20

PCIe Gen32x16

PCIe Gen 3

DDR42133-2666

Intel®

Xeon™ Core

1M

B L

2

DDR42133-2666

DDR42133-2666

DDR42133-2666

USB2.0x4

Intel®

Xeon™ Core1M

B L

2

LAUNCHED Q1’18 !!!

Intel® Confidential – For Use Under NDA OnlyDatacenter Group

SpecGrangeville with

Broadwell-DE SoCGrangeville NS with

Broadwell-DE-NS SoCBakerville with Skylake-D SoC

SOC TDP 20-65W ~20-65W ~60-110W

Cores Up to 16C with Intel® HT Technology Up to 8C with Intel® HT Technology Up to 18C with Intel® HT Technology

Package Size 37.5mm x 37.5mm 37.5mm x 37.5mm 45mm x 52.5mm

Key Target Network/Storage/Compute Network Storage/Compute/Network

AVX AVX-256 AVX-256 AVX-512 New Instructions

CacheLLC: 1.5MB/core LLC: 1.5MB/core LLC: 1.375MB/core

MLC: 256k/core MLC: 256k/core MLC: 1MB/core

Memory2 channels DDR4 per SOC

RDIMM, LRDIMM, SODIMM, UDIMMUp to 128G memory capacity

2 channels DDR4 per SOCRDIMM, LRDIMM, SODIMM, UDIMM

Up to 128G memory capacity

4 channels DDR4 per SOCRDIMM, LRDIMM, SODIMM*

Up to 512GB memory capacity

Ethernet Up to 2 lanes of 10GbE/1GbE Up to 4 lanes of 10GbE/1GbE Up to 4 lanes of 10GbE/1GbE with RDMA

PCIe

PCIe 3.0 (2.5, 5.0, 8.0 GT/s) PCIe 3.0 (2.5, 5.0, 8.0 GT/s) PCIe 3.0 (2.5, 5.0, 8.0 GT/s)

24xGen3 + 8xGen2 lanes 24xGen3 + 8xGen2 lanes 32xGen3 lanes + Up to 20xGen3 over High Speed IO

SATA 6xSATA ports 6xSATA ports Up to: 14xSATA

Crypto/Compression

Via External ControllerColeto Creek/Cave Creek or similar

Integrated Intel® QuickAssist Technology: Up to 40G Crypto/20G Compression, 40kOps PKE 2K

Integrated Intel® QuickAssist Technology:Up to 100G Crypto/Compression + 100kOps PKE 2K

Gen Over Gen Comparison

44*Supported, not validated.See Memory POR Section.

New microarchitecture, memory subsystem and integration in Skylake generation

Intel® Confidential – For Use Under NDA OnlyDatacenter Group 45

Details of Intel® QuickAssist FunctionsLook Aside HW Accelerator for Security &

Compression / Decompression Co-processing services

Symmetric Cryptographic Functions:

• Ciphers:

• AES, 3DES, (A)RC, Kasumi, SNOW3G, ZUC

• Hash functions:

• MD5, SHA-1/2/3

• Authentication:

• HMAC, AES-XCBC-MAC

• Kasumi, SNOW3G, ZUC

• Cipher-Hash/Authenticate Combined Operation

• Key Derivation / Pseudo Random Functions

Public Key Functions:• RSA, DH, DSA

• ECDSA, ECDH

• Intel® Key Protection Technology

Compression/Decompression:

• Deflate Lossless LZ77 & Huffman coding

PCIe 3.0

x8

DMI Gen3

x4

Integrated Accelerator & PCH

PCIeGen3Root

Intel® QuickAssist Technology

PCIe 3.0

x16

4x10 GbE/1 GbE

with RDMA

10 GbE/1 GbEPorts

USB3

SATA3

GPIO

x1 GbE

SPI & eSPI

SMBus

InnovationEngine

(IE)

Crypto

Compression

46

Пример использования

47

Atom C -uServerистория

Atom C2000

Atom C3000

48

До 8 ядерДо 2.4GHz

Atom C2000

49

Примеры использования

Network appliance Cold storage

Intel Confidential – For Use Under NDA Only

Intel ® Data center SoC Roadmap

51

2015 2016

Intel® Atom® Processor C2000 Product Family

Intel® Atom® Processor C3000

Product Family(Denverton / Denverton NS)

Intel® Xeon® D-1500 Product Family

(Broadwell-DE)

2017+

Next Gen

Xeon® SoCs

Leadership 64-bit datacenter SoC roadmap targeting lower power, high density design points


Estimated Denverton PerformanceGeneral Purpose Server Performance: SPECint*_rate_base2006

Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using

specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance

tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. Results have been simulated and are provided for

informational purposes only. Results were derived using simulations run on an architecture simulator or model. Any difference in system hardware or software design or configuration may affect actual

performance. All products, computer systems, dates and figures specified are preliminary based on current expectations, and are subject to change without notice. *Other names and brands may be

claimed as the property of others.Copyright © 2016 Intel Corporation. All rights reserved

52


LowPower

IOIntegratedNIC

Atom®Cores

(Up to 16C)

Denverton/Denverton NS

ME

DDR 4

4X10/2.5/1GbE, Gen3 IO, Legacy IO

Core Ingredients

OS/VMM Enabling/Testing

MC

TP

(S

DV

)

I/O & LAN Drivers Intel SW/FW

Open Source SW

FW, BIOS, Drivers

IE

Harrisonville Platform overview

53

MemCtrl.

Intel HW

legend

SSD Recommendations• Intel SSD Data Center Family for SATA

• SATA 2.5in x 7mm• M.2 Boot (PCIe* and SATA)

QAT

MESPS FW

uEFI BIOS

Innovation Engine

3rd party FWДо 16 ядерДо 2.2GHz

Atom C3000


Intel® Xeon® processor D-1500

(Broadwell-DE)

Intel® Atom® processor C3000

(Denverton, Denverton-NS)

Production

AvailabilityMar 2015, Nov. 2015* July-August 2017

CPU core Broadwell Xeon core Goldmont Atom® core

TDP 20-65W 8.5-32W

Cores # Up to 16 cores Up to 16 cores

Cache 1.5MB L3 cache per core ≤8C L2 is 2MB/Core; >8C is 2MB/Core Pair

Performance** Up to 3.7x Atom® C2750 Up to 2.6x Atom® C2750

Memory TypeECC SODIMM, UDIMM, RDIMM ECC SODIMM, UDIMM, RDIMM

2CH DDR4, DDR3L 2CH DDR4

Max Mem. Cap. 128GB 128GB

Integrated Ethernet 2x 10 GbE Intel® Ethernet 4x10 GbE Intel® Ethernet

Flexible High Speed

I/ONo Yes

I/O x24 PCIE gen 3, x8 PCIE gen 2

x6 SATA3x4 USB 3.0, x4 USB 2.0

20 lanes of configurable HSIO 16 lanes PCIe gen 3, x16 SATA3,x4 USB3

+ fixed eMMC

Quick Assist (QAT) Discrete - Coleto Creek*** Integrated QAT****

Innovation Engine No Yes

Xeon® Processor D-1500 vs. Atom® Processor C3000

*Depend on SKUs** Comparing top bin on-roadmap SKUs Based on SPECint*_rate_base2006 projection*** Coleto Creek: DH89xxCL includes Chipset 8900 Series SKUs where 8925 ≤ SKU ≤ 8955**** QAT on select Denverton-NS SKUs only

54


Ну и совсем немного про MEHLOW (coffee lake xeon)

55


Branding update - Product Numbering Conventions

Intel® Xeon® W- 2 1 # #

Intel® Xeon® E- 2 1 # #

Skylake / Basin Falls

Mehlow (Coffee Lake CPU)

Intel® Xeon® processor D- 1 5 # # Intel® Xeon® D- 2 2 # #

Intel® Xeon Phi™ 7 2 # # a

processor

processor

processor

Change processor

Skylake-DE

Knights Landing

Broadwell-DE

Intel® Xeon® processor E7 - 8 8 # # V4a


Broadwell-EX

Broadwell-EP


Broadwell-EP 1S workstation



New

NewIntel® Xeon® processor E3 - 1 2 # # V6a

Greenlow (Kaby Lake CPU)

a

# of Cores G = integrated graphics

Intel® Xeon® Platinum 8 1 # # a

Intel® Xeon® Gold 5 1 # # a

Intel® Xeon® Silver 4 1 # # a

Intel® Xeon® Bronze 3 1 # # a

Intel® Xeon Processor Scalable Family (Skylake/Purley)

processor

processor

processor

processor

New

a

Intel® Xeon® Gold 6 1 # # a processor

= SKU # (E.g. 20)Level Generation= Integrations/Optimizations (if applicable, E.g. F = Fabric)


Mehlow platform will result in up to 30-50% improvement on multi-threaded workloads*

Mehlow Platform with Intel® Xeon® E processor

57

2016 2017 2018 2019

Xeon E3

Entry

Server &

WS

Greenlow: LGA SKL PCH-H, 2ch 4 DIMM

DDR4/3

Mehlow: LGA, CNL PCH,

2ch 4 DIMM

Coffee Lake-E: ≤6cSkylake –E ≤4c Kaby lake -E ≤4c

Source: Intel. Configuration: Windows* 10,dual-channel DDR4-2400, 7200RPM HDD, 4 phase VR (~$1 BOM adder)Note: Values are based on estimates.


Storage

Workstation

Internet of Things

Small Business

Creativity from anywhere –Optimized HD graphics for reliable,

stable performance with your professional workloads

Energy efficient; Faster I/O bandwidth with graphics for Entry

Enterprise. Delivering value to software defined storage (SDS) and

SAN/NAS.

Best Perf/Watt delivery for IoTapplications in the public sector,

medical, and industrial production

Equip your SMB with a server, allowing for increased security and reliability with small incremental investment.

Intel® Xeon® E Delivering Value to Multiple Segments

CoffeeLake-S


Intel® Xeon® E processor SKU Stack

All SKUs can be used in either server or workstation platforms

Note: Subject to change without notice.

SKU DecoderBrand String, #Cores/Threads, Frequency, Core+Graphics, Hyper threading, Test TDP

* GT0 and GT2 graphics can be used for pedestal workstation, server, and AMT-server. Note: iGFX not supported in intel® Server Platforms Services (server)

4 Core 6 Core

pGFX

No pGFX

E-2174G 4C/8T

3.8/4.7 4+2 HT 71W

E-2144G 4C/8T

3.6/4.5 4+2 HT 71W

E-2134 4C/8T

3.5/4.5 4+0 HT 71W

E-2124G 4C/4T

3.4/4.5 4+2 HT 71W

E-2124 4C/4T

3.3/4.3 4+0 No HT 71W

E-2104G 4C/4T

3.2 4+2 No HT 65W

E-2186G 6C/12T

3.8/4.7 6+2 HT 95W

E-2176G 6C/12T

3.7/4.7 6+2 HT 80W

E-2146G 6C/12T

3.5/4.5 6+2 HT 80W

E-2126G 6C/6T

3.3/4.5 6+2 No HT 80W

E-2136 6C/12T

3.3/4.5 6+0 HT 80W

Please contact your local customer QNR rep for reliability differences between Xeon® E, Xeon® D, and Xeon® SP.


Note: Non-Xeon CPUs do not support Intel® vPro™ or Intel® AMTPlease reference the client Dashboard for CFL-S sample/prod schedules, CDI# 570312Intel® Core/Pentium/Celeron SKU stack – Desktop Platform 5Q Roadmap link

Coffee lake Core i3, Pentium, and Celeron CPU Support

60

Desktop CPUIntel® Core i3-8350K

Intel® Core i3-8300

Intel® Core i3-8100

Intel® Core i3-8300T

Intel® Core i3-8100T

Intel® Pentium G5600



Intel® Celeron G4920

Intel® Celeron G4900

Intel® Pentium G5500T

Intel® Pentium G5400T

Intel® Celeron G4900T

All SKUs and features are PRELIMINARY and

can change without notice

https://iref.intel.com/GetDoc.aspx?RefLibObjectID=0902007c8004311a&Latest=True


Intel® Xeon® E LGA1151 Package & Socket

▪ CFL-S will use the same LGA1151 package & socket as KBL-S.

– Maximizes design reuse and socket cost structure.

▪ No support for forward or backward compatibility with legacy platforms.

LGA1151

37.5mm x 37.5mm

15.5mm

Not drawn to scale.Do not use for design purposes.


Integrated Intel® Wireless-AC (Jefferson Peak)

▪ CFL delivers integrated Wi-Fi & Bluetooth* 5.

– Wi-Fi/BT MAC integrated into CNL PCH.

– Separate companion PHY+RF (CRF) 802.11ac R2 + Bluetooth* 5 module with 2x2 (160 MHz Channels) and 1x1 options.

▪ Standard M.2, hybrid Key E pinout allows support for integrated (CNVi) and discrete.

– Routing CNVi alone drives I/O savings.

▪ Soldered-down option also available.

▪ Driver & BIOS auto-detect CNVi or discrete.

▪ .

Integrated Intel® Wireless-AC deliversa flexible, compelling and competitive

WiFi/BT solution.

Integrated

(CNVi)

Discrete

Jefferson Peak

Thunder Peak

CRF

Module

M.2 2230

Cannon Lake PCH

Pulsar

MACPCH Resources

(Power, I/O,

Chassis)Wi-Fi/BT

PHY+RF

Ke

y E

CNVio (Wi-Fi)

GPIOs (BT), Clock

Platform Pwr

PCIe (Wi-Fi)

USB (BT)

Platform Pwr

Discrete

Module

M.2 2230

Cannon Lake PCH

Pulsar

MACPCH Resources

(Power, I/O,

Chassis) Wi-Fi/BT

Ke

y E

CNVio (Wi-Fi)

GPIOs (BT), Clock

Platform Pwr

PCIe (Wi-Fi)

USB (BT)

Platform Pwr

Освещена только часть линейки продукции для встраиваемых решений

https://www.intel.com/content/www/us/en/partner/solutions-alliance/program-overview.html

https://solutionsdirectory.intel.com/

2017 2018

https://solutionsdirectory.intel.com/

Internet of Things Group Intel Confidential – NDA Platform Roadmap. All dates and plans are subject to change without notice. 64

Expanded Capacity Range

Multiple form factors

Flexible feature sets

Extended supply life cycle

Embedded SSD Product Family Dedicated to IoT!

Data center

EmbeddedClient + Form/Fit/Function

Continuity

3yr*

Low Power

PLI

Intel® Remote Secure Erase

M.2 80 & 110mm

2.5”

As low as 64 GB

Extended Temperature

*NSG product availability for Embedded products is three years from when the product(s) is first available on the Intel SSD 5Q roadmap to when the product(s) last ships.


Intel Technology Innovations Fill the Memory and Storage Gap Performance and Capacity for Every Need

72

КОНКУРЕНТЫ

Datacenter Marketing Group

Industry Adoption Growing

Comms Service ProvidersEnterprise & Gov’tCloud Service Providers

Intel® Xeon® Scalable Momentum

73

Datacenter Marketing Group 74

System Pricing Comparison

Supermicro 6029U-TRT• 12x 16GB DDR4-2666• Quad 1GbE• 1 SATA HDDhttp://www.thinkmate.com/system/ultra-superserver-6029u-trt

Supermicro 2023US-TR4• 16x 16GB DDR4-2666• Quad 1GbE• 1 SATA HDDhttp://www.thinkmate.com/system/ultra-superserver-6029u-trt

Processor List Price 2S Intel Xeon-SP 81602 x $4702/processor

2S AMD EPYC 76012 x $4200/processor

System Price $16148 $16280

• Both systems configured with 1 DIMM per channel to get full memory bandwidth.

Intel does not control or audit the design or implementation of third party benchmark data or Web sites referenced in this document. Intel encourages all of its customers to visit the referenced Web sites or others where similar performance

benchmark data are reported and confirm whether the referenced benchmark data are accurate and reflect performance of systems available for purchase.

Systems priced online at www.Thinkmate.com on 10/12/2017.

http://www.thinkmate.com/system/ultra-superserver-6029u-trt

http://www.thinkmate.com/system/ultra-superserver-6029u-trt

http://www.thinkmate.com/


1,0

0

1,0

5

1,0

5 1,1

8

1,2

2

1,2

7

1,2

9

1,6

4

1,3

7

1,2

7

1,1

5

1,2

2

1,1

8

1,4

5

1,3

1

1,6

8

1,6

4 1,7

5

0

0,2

0,4

0,6

0,8

1

1,2

1,4

1,6

1,8

2

2S AMD EPYC7601

(32C, 2.2GHz)

Server Side Java Node.jsJavascript

PHP Runtime DPDK L3 Fwding HammerDBDatabase

NoSQLDatabase

OnlineTransactionProcessing

MemoryCaching

ServerVirtualization

2S AMD EPYC 7601(32C, 2.2GHz)

2S Intel Xeon Platinum 8160(24C, 2.1Ghz)

2S Intel Xeon Platinum 8180(28C, 2.5Ghz)

Internal Measured Performance: Enterprise and Cloud Workloads

• EPYC perf variability on DPDK L3 Forwarding, HammerDB, and Memory Caching – see slides 13, 15

Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. Refer to Configuration: Enterprise and Cloud Workloads, slides 25-28

Relative Performance

Higher is better

75

Storage bound

workload

Network bound

workload

Network bound

workload


Performance per Core

Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. Intel does not control or audit the design or implementation of third party benchmark data or Web sites referenced in this document. Intel encourages all of its customers to visit the referenced Web sites or others where similar performance benchmark data are reported and confirm whether the referenced benchmark data are accurate and reflect performance of systems available for purchase. Refer to Configuration: Performance per Core, slide 29

1,38

1,53

0

0,2

0,4

0,6

0,8

1

1,2

1,4

1,6

1,8

2S AMD EPYC 7601 (32C, 2.2GHz)

2S Intel Xeon Platinum 8160 (24C, 2.1GHz)


This is based on public SPECint*_rate_base2006 Performance / Number of CoresRelative Performance

Higher is better

76


Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmark and MobileMark, are measured using specific computer systems, components, software, operations and functions. Any change to any of those factors may cause the results to vary. You should consult other information and performance tests to assist you in fully evaluating your contemplated purchases, including the performance of that product when combined with other products. Intel does not control or audit the design or implementation of third party benchmark data or Web sites referenced in this document. Intel encourages all of its customers to visit the referenced Web sites or others where similar performance benchmark data are reported and confirm whether the referenced benchmark data are accurate and reflect performance of systems available for purchase. Refer to Configuration: HPC Workloads, slides 30-34

Relative Performance

Higher is better

77

0,7

3

0,8

1

0,8

5

1,0

0 1,2

3 1,4

6

1,4

7

1,5

2

2,6

9

0,7

5

0,8

6

0,8

8 1,1

0 1,3

5 1,6

3

1,5

9

1,6

5

3,1

6

0

0,5

1

1,5

2

2,5

3

3,5

2S AMD EPYC7601

(32C, 2.2GHz)

STREAM-OMP Triad

OpenFOAM v4.1

MILC v7

NAMD v2.12

GROMACS v2016.2

LAMMPS v17-Aug-17

Black-Scholes Kernel

Monte Carlo Kernel

Double PrecisionMatrix

MultiplicationDGEMM2S AMD EPYC 7601

(32C, 2.2GHz)2S Intel Xeon Gold 6148 (20C, 2.4GHz)


Internal Measured Performance: HPC Workloads - Single Node

• All measurements above are single node. Expect Intel performance to improve on cluster measurements - more typical of HPC deployments

• EPYC tested with GCC, AOCC/LLVM and Intel compiler. The best AMD result was used for each workload

• Intel® Xeon Phi™ processors are a good option for customers who have memory-bandwidth-bound workloads

Memory sensitive apps

Compute sensitive apps


Экосистема не готова – Open Source проекты не оптимизированы

Нужно по крайней мере 3-5 лет для оптимизации (2020+)

Microsoft НЕ МОЖЕТ эмулировать x86 код на ARM архитектуре

78


99% серверов с GPGPU – на Xeon

Интел в процессе разработки продуктов для рынка

Nervana и MobileEye продукты – ASIC или AP+ASIC. FPGA проекты

79

15 лет доступности - karma-group.ru · † Intel may support this extended...

Documents

Transcript of 15 лет доступности - karma-group.ru · † Intel may support this extended...