Vmworld 2019 Content: Not for Publication Or Distribution

HCI3710BUS distribution or Optimize Infrastructure Modernization using VMware publication vSAN and Intel Optane for Technologies Not Content: Vivek Sarathy, Intel Nd Ngoka, VMware 2019

#vmworld # HCI3710BUS VMworld Disclaimer

This presentation may contain product features or functionality that are currently under development. distribution This overview of new technology represents no commitment from VMwareor to deliver these features in any generally available product. Features are subject to change, and must not be includedpublication in contracts, purchase orders, or sales agreements of any kind. for Technical feasibility and market demandNot will affect final delivery. Pricing and packaging for any new features/functionality/technology discussed or presented, have not beenContent: determined. 2019

The information in this presentation is for informational purposes only and may not be incorporated into any contract. There is no commitment or obligation toVMworld deliver any items presented herein.

©2019 VMware, Inc. 2 Intel Legal Disclaimers The products described in this document may contain design defects or errors known as errata which may cause the product to deviate from published specifications. Current characterized errata are available on request. Intel technologies’ features and benefits depend on system configuration and may require enabled hardware, software or service activation. Performance varies depending on system configuration. No computer system can be absolutely secure. Check with your system manufacturer or retailer or learn more atintel.com. Performance results are based on testing as of the date set forth in the configuration, and may not reflect all publicly availabdistributionle security updates. See configuration disclosure for details. No product or component can be absolutely secure. or 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 www.intel.com/benchmarks. publication Cost reduction scenarios described are intended as examples of how a given Intelfor-based product, in the specified circumstances and configurations, may affect future costs and provide cost savings. Circumstances will vary. Intel does not guarantee any costs or cost reduction. Intel does not control or audit third-party data. You should review thisNot content, consul other sources, and confirm whether referenced data are accurate. Intel, the Intel logo, and other Intel marks are trademarks of Intel Corporation in the U.S. and/or other countries. © Intel Corporation. Content: *Other names and brands may2019be claimed as the property of others. VMworld

publication for Not Content: 2019 VMworld Agenda

1 vSAN Adoption and Overview distribution or 2 vSAN and CPU 3 vSAN and Storage publication for 4 Performance of vSANNot with Optane Technologies 5 Summary Content: 2019 VMworld

#1 Market Share with distribution >20,000 Customers or

publication for Not Content: 1 2 3 4 5 6 20197 8 9 10 11 12 13 14 15 16 17 18 19 vSAN Customer Adoption VMworld Source: https://www.businesswire.com/news/home/20190625005943/en/

Customer-Managed Private Cloud Partner-Managed Cloud

vSAN distribution ReadyNodes or

Integrated publication Systems for Engineered Not Appliances

Content:VMware -Managed Private2019 Cloud Public Cloud VMware Cloud on Dell EMC VMware Cloud on AWS VMworld

distribution  Runs on any standard x86or server 1 2  Pools disks into a shared datastore publication vSphere vSAN for Delivers enterprise-grade scale and Not performance

Content:  Managed through per-VM storage policies vSAN Datastore 2019  Deeply integrated with the VMware stack VMworld

distribution or

vSphere vSAN publication for Not vSAN Datastore Content: 2019 VMworld

DDR4 DDR4  Efficiency of the CPU coredistribution and frequency UPI or

CPU CPU  NUMA optimization publication  Multiple Memory channels X16 PCIe for Not  Increase in storage bandwidth

Content:  Support of AES-NI offloading 2019 VMworld

Per Socket: distribution Integrated or Networking RDMA  Up to 28 cores (9200: 56 cores) DDR4 DDR4

UPI Intel® Intel® publication Up to 3.8GHz Xeon® Xeon® Processor Processor for VMD VMD Not  6 Memory channels X16 PCIe  48 PCIe Lanes

Content: ISA-L SPDK VROC Memory Drive  3 UPI busses 2019 CAS NVML VTUNE System Studio Parallel Studio XE VMworld

INTEL® XEON® INTEL® XEON® ® ® ® ® ® distribution® ® ® PLATINUM PLATINUM INTEL XEON INTEL XEON INTEL XEON INTEL XEON GOLD 6200 GOLD 5200 SILVERor 4200 BRONZE 3200 9200 8200 PROCESSORS PROCESSORS PROCESSORS PROCESSORS PROCESSORS PROCESSORS publication for Not

A NEW CLASS OF ADVANCED PERFORMANCE Content: 2019 UNINTERRUPTED GROUNDBREAKING EMBEDDED HARDWARE ENHANCED BUILT-IN LEADERSHIP WORKLOAD MEMORY AI ENHANCED AGILITY & VALUEVMworld PERFORMANCE INNOVATION ACCELERATION SECURITY UTILIZATION

distribution or

vSphere vSAN publication for Not vSAN Datastore Content: 2019 VMworld

vSAN uses the concept of disk groups to pool together SSD devices as single management constructs. distribution or Disk groups are composed of at least 1 flash device and 1-7 capacity devices. • Flash devices are use for cache/write buffer tier • Flash devices are also used for capacity tier publication • Disk groups cannot be created without a flash fordevice • Can have 5 disk groups per host = 35 capacityNot devices per host

Each host:Content: 5 disk groups max. Each disk group: 1 flash device + 1-7 capacity devices disk group disk group disk group disk group disk group 2019 VMworld

• Best Practices: distribution or

vSphere vSAN – Multiple disk groups create smaller 1 HBA publicationfailure domains for – Multiple disk groups provide better Not performance – Implement Balanced SSD SSD SSD SSD SSD configurations – e.g. do not Content: configure different hosts with different number of disks per hosts SSD SSD2019SSD SSD SSD and/or different number of disk groups Disk Group Disk Group Disk Group Disk Group Disk Group VMworld

• All-flash allocates the cache tier to 100% buffer writes • Writes are acknowledged back to VM when they hit the write bufferdistribution vSphere vSAN or 1 HBA • Data from committed writes accumulate Upper Layers in write buffer • Allpublication writes eventually destaged for Writes Reads Cache • Data written most often (hot) will say in Not write buffer Write Buffer • Data less frequently accessed (cold) Content:Capacity are de-staged to flash capacity layer 2019 • Reads come from the capacity tier, Capacity Tier once destaged • Hot Reads come from the write buffer if VMworldDisk Group not yet destaged

 vSAN all-flash dedicatesdistribution 100% of the cache device to writeor buffering.

 Endurance: Higher publication for Predictable I/O performance : Consistent Not  Low Latency : Faster

 Higher IOPs always helps Content: The class of SSD you choose has a direct impact on 2019 the performance of your overall system. VMworld

CPU CACHE distribution MEMORY 10s GB DRAM or HOT TIER <0.1usec GROWING 100s GB PERSISTENT MEMORY Capacity MEMORY CAPACITY Gap publication<1usec IMPROVING SSD Storage Performancefor 1s TB STORAGE PERFORMANCE NotGap <10usecs 3D NAND SSD WARM TIER 10s TB INTEL® QLC 3D NAND SSD <100usecs DELIVERING Content: EFFICIENT STORAGE2019 Cost Performance Gap HDD / TAPE 10s TB VMworld COLD TIER <10,000 usecs

distribution or TRANSISTOR-LESS DESIGN

Data is written at a bit level, Intel® Optane™ so each cell's state can be publication Memory Media changed from a 0 or 1 for independently of other cells Not Content: 2019 Intel® Optane™ Technology design is fundamentally different from NAND VMworld

Read and write in PAGES distribution ERASE only in BLOCKS Garbageor Erase BEFORE writing

NAND Collection publication S S S S for S S S SNot S S S S S S S S S Content:S S S S 2019 S S S S VMworld Permission to use garbage truck icon made by monkik from www.flaticon.com with reference

Intel® Optane™ DC SSD distribution Bit-level addressing x-pointor architecture Data writes without need for erase No garbage to collect! publication for Not Content: 2019 VMworld Permission to use garbage truck icon made by monkik from www.flaticon.com with reference

4K 70/30 RW Performance at Low Queue Depth1 (higher is better) distribution or 5-6x faster at low Queue publicationDepths1 for Not Vast majority of applications generate low QD storage Content: workloads Intel® SSD 2019DC P4600 Intel® Optane™ SSD DC P4800X

1 Source – Intel-tested: Average read latency measured at queue depth 1 during 4k random write workload. Measured using FIO 3.1. Common Configuration - Intel 2U Server System, OS CentOS 7.5, kernel 4.17.6-1.el7.x86_64, CPU 2 x Intel® Xeon® 6154 Gold @ 3.0GHz (18 cores), RAM 256GB DDR4 @ 2666MHz. Configuration – Intel® Optane™ SSD DC P4800X 375GB and Intel® SSD DC P4600 1.6TB. Latency – Average read latency measured at QD1 during 4K Random Write operations using FIO 3.1. Intel Microcode: 0x2000043; System BIOS: 00.01.0013; ME Firmware: 04.00.04.294; BMC Firmware: 1.43.91f76955; FRUSDR: 1.43. SSDs tested were commercially available at time of test. The benchmark results may need to be revised as additional testing is conducted. Performance results are based on testing as of July 24, 2018 and may not reflect all publicly available security updates. See configuration disclosure for details. No product can be absolutely secure. Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmarkand MobileMark, are measured using specificVMworld 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 www.intel.com/benchmarks.

Read QoS in Mixed Workload1 (lower is better) distribution or s) s) µ

up to 60x publicationup to 60x better at 99% QoS1 lower

Read Response Time ( Time ResponseRead for Not Ideal for critical applications Content: with aggressive latency requirements TimeTime (m (ms)s)

Intel® SSD DC P46002019Intel® Optane™ SSD DC P4800X

1 Source – Intel-tested: QoS: measures 99 percent QoS under 4K 70–30 workload at QD1 using FIO 3.1. . Common Configuration - Intel 2U Server System, OS CentOS 7.5, kernel 4.17.6-1.el7.x86_64, CPU 2 x Intel® Xeon® 6154 Gold @ 3.0GHz (18 cores), RAM 256GB DDR4 @ 2666MHz. Configuration – Intel® Optane™ SSD DC P4800X 375GB and Intel® SSD DC P4600 1.6TB. Latency – Average read latency measured at QD1 during 4K Random Write operations using FIO 3.1. Intel Microcode: 0x2000043; System BIOS: 00.01.0013; ME Firmware: 04.00.04.294; BMC Firmware: 1.43.91f76955; FRUSDR: 1.43. SSDs tested were commercially available at time of test. The benchmark results may need to be revised as additional testing is conducted. Performance results are based on testing as of July 24, 2018 and may not reflect all publicly available security updates. See configuration disclosure for details. No product can be absolutely secure. Software and workloads used inVMworld performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmarkand 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 www.intel.com/benchmarks.

Average Read Latency under Random Write Workload1 (lower is better) distribution or up to 40x faster up to publicationresponse time Write Pressure MB/s) MB/s) Pressure Write for under workload1 Not Consistently

Time (secondContent:s) amazing response Random Writes 2019Intel® SSD DC P4600 time under load Intel® Optane™ SSD DC P4800X

1 Source – Intel-tested: Response Time refers to average read latency measured at queue depth 1 during 4k random write workload. Measured using FIO 3.1. Common Configuration - Intel 2U Server System, OS CentOS 7.5, kernel 4.17.6-1.el7.x86_64, CPU 2 x Intel® Xeon® 6154 Gold @ 3.0GHz (18 cores), RAM 256GB DDR4 @ 2666MHz. Configuration – Intel® Optane™ SSD DC P4800X 375GB and Intel® SSD DC P4600 1.6TB. Latency – Average read latency measured at QD1 during 4K Random Write operations using FIO 3.1. Intel Microcode: 0x2000043; System BIOS: 00.01.0013; ME Firmware: 04.00.04.294; BMC Firmware: 1.43.91f76955; FRUSDR: 1.43. SSDs tested were commercially available at time of test. The benchmark results may need to be revised as additional testing is conducted. Performance results are based on testing as of July 24, 2018 and may not reflect all publicly available security updates. See configuration disclosure for details. No product can be absolutely secure. Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmarkand 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 thatVMworld product when combined with other products. For more complete information visit www.intel.com/benchmarks.

Drive Writes Per Day (DWPD) distribution Up to or Intel® Optane™ More data SSD DC P4800X 60.0 Up to 20X higher can be publicationDWPD endurance when written over for compared to typical the life of the Intel® SSDNot DC NAND medium drive P4600 (3D NAND) 3.0 endurance SSDs Content: DWPD 2019

Source – Intel: Endurance ratings available at https://www.intel.com/content/www/us/en/products/memory-storage/solid-state-drives/data-center-ssds/optane-dc-p4800x- series/p4800xVMworld-750gb-2-5-inch.html

Standard SAS / SATA/ NVMe SSDs Intel® Optane™ SSD DC P4800X distribution or

publication for Not Content: * Up to 87%2019reduction in caching tier SSD size when using Intel® Optane™ SSD https://blogs.vmware.com/virtualblocks/2018/08/23/extending-all-flash-vsan-cache-tier-sizing-requirement-for- differentVMworld-endurance -level-flash-device/

* Source©2019 : https://blogs.vmware.com/virtualblocks/2018/08/23/extending VMware, Inc. -all-flash-vsan-cache-tier-sizing-requirement-for-different-endurance-level-flash-device/ - compare Figure 2 600 GB – 2 Disk Group and Figure 3 4.8TB – 2 Disk Group distribution or

Performance of vSAN with Intel®publication Optane™ SSDs for Not Content: 2019 VMworld

100% Sequential 64KB Writes—Throughput 3,000 distribution or 2,500 Dell EMC VxRail* with Intel® Optane™ SSDs 2,000

1,500 publication Throughput drops 39% during SAS cache aggressive write Dell EMC VxRail* with NAND SSDs 1,000 buffer destage. for Throughput (MB/Sec) Throughput Not 500

0 1 Content: 251 501 Time (sec)

2019 Source: Enterprise Strategy Group

Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmarkand 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 www.intel.com/benchmarks. Performance results are based on testing as of August, 2019 and may not reflect all publicly available security updates. See configuration disclosure for details. No product or component can be absolutely secure. 1 Source: Enterprise Strategy Group. Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Configuration details available from Enterprise Strategy Group: https://www.esg-global.com/validation/esg-technical- validation-dellVMworld-emc-vxrail-with-intel-xeon-scalable -processors-and-intel-optane-ssds. Intel does not control or audit third-party benchmark data or the web sites referenced in this document. You should visit the referenced web site and confirm whether referenced data are accurate. * Other names and brands may be claimed as the property of others

OLTP Workload – HammerDB* distribution 12,000,000 + 61% 10,190,789or 10,000,000

8,000,000 6,342,419 6,000,000 publication for 4,000,000 + 61% Not 2,215,489 2,000,000 1,378,798

0 Content:1 2 Series1 Series2

2019 Source: Enterprise Strategy Group

Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmarkand 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 www.intel.com/benchmarks. Performance results are based on testing as of August, 2019 and may not reflect all publicly available security updates. See configuration disclosure for details. No product or component can be absolutely secure. 1 Source: Enterprise Strategy Group. Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Configuration details available from Enterprise Strategy Group : https://www.esg-global.com/validation/esg-technical- validation-dellVMworld-emc-vxrail-with-intel-xeon-scalable -processors-and-intel-optane-ssds. Intel does not control or audit third-party benchmark data or the web sites referenced in this document. You should visit the referenced web site and confirm whether referenced data are accurate. * Other names and brands may be claimed as the property of others

SLOB* - workload I/O generator for Oracle “like” databases Throughput (MB/s)distribution IOPS higher is better higher is better 3,000 or 300,000 Increase throughput by 34% Boost IOPS by 35% 2,500 250,000 2,000 200,000 1,500 publication 150,000

100,000 for1,000 50,000 Not 500 0 0 1 2 1 2 Content: Source: Principled Technologies Intel® Optane™ SSD based vSAN shows the ability to process more IOPS while still supporting high throughput indicates2019 a solution’s ability to support periods of heavy user activity.

Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmarkand 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 www.intel.com/benchmarks. Performance results are based on testing as of August, 2019 and may not reflect all publicly available security updates. See configuration disclosure for details. No product or component can be absolutely secure. 1 Source: PrincipledVMworld Technologies . Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Configuration details available from Principled Technologies : https://www.principledtechnologies.com/hpe/Intel- Optane-HPE-ProLiant-VMware-vSAN-Oracle-workload-testing-0519.pdf. Intel does not control or audit third-party benchmark data or the web sites referenced in this document. You should visit the referenced web site and confirm whether referenced data are accurate. * Other names and brands may be claimed as the property of others

WAS IS up to CPU distribution % orIncreased Intel® 3D Intel® 1 CACHE NAND Optane™ VM density DC SSD SSD 60 Intel® SSD Intel® Optane™ DC P4600 SSD DC P4800X publication 1.6TB 375GB for up to Not Intel®Intel® 3D 3D Intel®Intel® 3D 3D Potential 3-yr NAND™Intel®Intel® 3D 3D NAND™Intel®Intel® 3D 3D NAND™NAND™ NAND™NAND™ $ CAPACITY DCNAND SSD DCNAND SSD DCDC SSD SSD DCDC SSD SSD 1 DC SSD DC SSD cost savings Intel® SSD Content:Intel® SSD 25% DC P4500 DC P4500 2019 Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Performance tests, such as SYSmarkand 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 www.intel.com/benchmarks. Performance results are based on testing as of October, 2018 and may not reflect all publicly available security updates. See configuration disclosure for details. No product or component can be absolutely secure. 1 Tests by The Evaluator Group. Software and workloads used in performance tests may have been optimized for performance only on Intel microprocessors. Configuration details available from The Evaluator Group at https://www.evaluatorgroup.com/document/lab-insight-latest-intel-technologies-power- new-performance-levels-vmware-vsan-2018-update/. See Appendix A for server cost estimate details and assumptions. * Other names VMworldand brands may be claimed as the property of others

NEW distribution INTEL® 3D NAND Memory/ DRAM INTEL® OPTANE™ or SSDs Storage Type DC SSDS & SATA

Higher Cost Lower Cost Characteristics Smaller Capacity Data persistence Larger Capacity Faster Performance publicationSlower Performance for Processing Intel® Xeon® 2ndNot Gen Scalable Processor (cascade lake) Content: 2019FAST MEMORY, PERSISTENCE OF STORAGE. FLEXIBLE AND SCALABLE TO ACCELERATE YOUR DATA INSIGHTS. VMworld

“APP DIRECT” MEMORY MODE MODE PLATFORM/OS/APP ACCESS TO APP/WORKLOAD DIRECT ACCESS TO High SPEED, HIGHdistribution CAPACITY MEMORY HIGH SPEED, HIGH CAPACITY STORAGE or

High capacity Persistent and much Targeting >1.2X More higher data capacity publication VMs* for Affordable capacity High availability/ Not 128GB, 256GB and less downtime 512GB Modules Ease of adoption Significantly faster Content: No code changes storage 2019 required VMworldBUILT -IN FLEXIBILITY TO USE BOTH MODES SIMULTANEOUSLY

The current VMmark 3.1* benchmark is CPU intensive and was not designed to benchmark larger memory configurations, so Intel and VMware created a new test methodology (modified VMmark 3.1) for incremental memory for VMmark. This methodology was used to build and optimize the Intel Select Solutions for VMware vSAN configurations.

Intel Select Solution: Intel Select Solution: Base Configuration (4 nodes) Plus Configuration (4 nodes)distribution 192GB DDR4or DRAM + 512GB System Memory 384GB DDR4 DRAM Intel® Optane™ DC persistent memory

Intel 3D Intel 3D Storage Media + publication+ NAND SSDs for NAND SSDs Up to VMs per Node Higher VM Not Density/per (higher is better) 95 152 60% Server Estimated Cost per Improved VM 1 1 Performance $321 USDContent: PER VM $305 USD PER VM TCO (lower is better) 5%

2019Intel® Xeon® Gold 6230 Intel® Xeon® Gold 6252 Processor Processor

1Plus system: CPU: 2x Intel® Xeon® Gold 6252 Processor MEMORY: 512 GB Intel® Optane™ DC Persistent Memory in Memory Mode and 192GB 2666MHz DDR4 DRAM Memory NETWORK: 2x 25GbE Intel® X710 STORAGE: 2x Intel® Optane P4800X + 6x Intel® P4510 vs Base system: CPU: 2x Intel® Xeon® Gold 6230 Processor MEMORY: 384GBVMworld DDR4 2666MHz DRAM Memory NETWORK: 4x 10GbE Intel® X722 STORAGE: 2x Intel® Optane P4800X + 6x Intel® P4510 . Intel® Optane™ DC persistent memory pricing & DRAM pricing referenced in TCO calculations is provided for guidance and planning purposes only and does not constitute a final offer. Pricing guidance is subject to change and may revise up or down based on market dynamics. Please contact your OEM/distributor for actual pricing. For more complete information about performance and benchmark results, visit www.intel.com/benchmarks. VMware VMmark® is a product of VMware, Inc. Pricing Guidance as of June 20, 2019. Valid until September 30, 2019. ©2019 VMware, Inc. 35 distribution or

Jointly committed to deliver best publicationin-class vSAN HCI for solutions built on cutting-edgeNot Intel® Optane™ Technology Content: 2019 VMworld

publication for Not Content: 2019 VMworld