Top Banner
1 of 5 Competitive Analysis N-Series SoC Based Thin Clients User experience and performance analysis versus HP Thin Clients Figure 1: Windows Media Player 11 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Max Avg HP T5325 HP T410 N500 CPU Utilization: Windows Media Player Executive Summary Desktop virtualization using thin clients can be challenging, and choosing the right thin client that is cost effective, easy to use and manage, future proof and energy efficient is difficult. Here at NComputing we are trying to address this issue with solid test results and side-by-side comparison of NComputing thin clients with thin clients from vendors such as HP. The N500 (N-Series) was evaluated for user experience and performance against HP t410 and HP t5325 thin clients. The thin clients were evaluated for system efficiency, CPU consumption, memory consumption, performance, and Quality of Experience (QoE) for three types of workloads: Real-world, everyday desktop tasks using MS Office tools (Word, Excel and PowerPoint) Start-up, restart and shut-down tasks Streaming Flash video and MP4 playback of video clips stored locally and on network drive Tests were designed to conduct a fair, repeatable comparison between virtual desktop solutions, using Xen Desktop and Xen App environments. Overall, N-series had lower CPU and memory utilization in all the tests. The N-Series and HP solutions delivered comparable results on everyday desktop tasks. N-Series outperformed the HP t410 and HP t5325 in subjective playback and multimedia experience. Key Findings Lowest CPU Usage. In MS Office productivity tests, N500 had lower CPU utilization. For NComputing N500, the average CPU utilization was 15% and maximum CPU utilization of 25% versus HPt410, which had an average CPU utilization of 18% and maximum CPU utilization of 30% during MS Excel and MS PowerPoint benchmarking. Simple to manage. NComputing’s device management software, vSpace Management Server, a web-based application accessible over the web, is simpler to understand and easily maintains devices versus HP Device Manager, a thick software, which needs to be installed as an application, can only be accessed through RDP and is more complicated in terms of device maintenance. Best multimedia performance. In subjective video quality tests comparing the playback experience of the same video clip in three different formats − full 1080p, 720p, and standard format - the N500 outpaced the HP t410 by a recognizable margin with a rating of 3.5 to 1 on a scale where 5 is best and 1 is poor. In addition, the HP t5323 does not support 1080p video and its overall rating was also 1. Device boot-up vs. user login. N500 had longer cold boot cycle - from device start-up to login screen - than the HP t410 (60 sec vs. 35 sec), but shorter than the HP t5325 (95 sec). Once booted, the N500 took 10-15 sec to login into the Desktop VM session as opposed to both the HP t410 and HP t5325 (15 - 20 sec). N-Series SoC Based Thin Clients User experience and performance analysis versus HP Thin Clients
5

N-Series SoC Based Thin Clientscloudmty.com/.../04/SERIE-N-CITRIX-vs-THIN-CLIENTS-HP.pdf1 of 5 Competitive nalysis N-Series SoC Based Thin Clients User experience and performance analysis

Oct 05, 2020

Download

Documents

dariahiddleston
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: N-Series SoC Based Thin Clientscloudmty.com/.../04/SERIE-N-CITRIX-vs-THIN-CLIENTS-HP.pdf1 of 5 Competitive nalysis N-Series SoC Based Thin Clients User experience and performance analysis

1 of 5

Competitive Analysis

N-Series SoC Based Thin ClientsUser experience and performance analysis versus HP Thin Clients

Figure 1: Windows Media Player 11

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%Max

Avg

HP T5325HP T410N500

CPU Utilization: Windows Media Player

Executive SummaryDesktop virtualization using thin clients can be challenging, and choosing the right thin client that is cost effective, easy to use and manage, future proof and energy efficient is difficult. Here at NComputing we are trying to address this issue with solid test results and side-by-side comparison of NComputing thin clients with thin clients from vendors such as HP.

The N500 (N-Series) was evaluated for user experience and performance against HP t410 and HP t5325 thin clients. The thin clients were evaluated for system efficiency, CPU consumption, memory consumption, performance, and Quality of Experience (QoE) for three types of workloads:

• Real-world, everyday desktop tasks using MS Office tools (Word, Excel and PowerPoint)

• Start-up, restart and shut-down tasks

• Streaming Flash video and MP4 playback of video clips stored locally and on network drive

Tests were designed to conduct a fair, repeatable comparison between virtual desktop solutions, using Xen Desktop and Xen App environments.

Overall, N-series had lower CPU and memory utilization in all the tests. The N-Series and HP solutions delivered comparable results on everyday desktop tasks. N-Series outperformed the HP t410 and HP t5325 in subjective playback and multimedia experience.

Key Findings • Lowest CPU Usage. In MS Office

productivity tests, N500 had lower CPU utilization. For NComputing N500, the average CPU utilization was 15% and maximum CPU utilization of 25% versus HPt410, which had an average CPU utilization of 18% and maximum CPU utilization of 30% during MS Excel and MS PowerPoint benchmarking.

• Simple to manage. NComputing’s device management software, vSpace Management Server, a web-based application accessible over the web, is simpler to understand and easily maintains devices versus HP Device Manager, a thick software, which needs to be installed as an application, can only be accessed through RDP and is more complicated in terms of device maintenance.

• Best multimedia performance. In subjective video quality tests comparing the playback experience of the same video clip in three different formats − full 1080p, 720p, and standard format - the N500 outpaced the HP t410 by a recognizable margin with a rating of 3.5 to 1 on a scale where 5 is best and 1 is poor. In addition, the HP t5323 does not support 1080p video and its overall rating was also 1.

• Device boot-up vs. user login. N500 had longer cold boot cycle - from device start-up to login screen - than the HP t410 (60 sec vs. 35 sec), but shorter than the HP t5325 (95 sec). Once booted, the N500 took 10-15 sec to login into the Desktop VM session as opposed to both the HP t410 and HP t5325 (15 - 20 sec).

N-Series SoC Based Thin ClientsUser experience and performance analysis versus HP Thin Clients

Page 2: N-Series SoC Based Thin Clientscloudmty.com/.../04/SERIE-N-CITRIX-vs-THIN-CLIENTS-HP.pdf1 of 5 Competitive nalysis N-Series SoC Based Thin Clients User experience and performance analysis

2 of 5

Competitive Analysis

FeaTureS & CriTeria N500 HP T410 HP T5325

eaSe OF uSe

Installation kit included Yes No No

Auto discovery Yes Yes Yes

uSer exPerieNCe

System boot time 60 ~ 70 sec 35 ~ 40 sec 80 ~ 85 sec

Login time (warm boot) 10 ~ 15 sec 15 ~ 20 sec 15 ~ 20 sec

Session connect 10 ~ 15 sec 15 ~ 20 sec 15 ~ 20 sec

DeviCe CONFiguraTiON aND COmPaTiBiliTy

Processor Numo 3, ARM Cortex A9, 600MHz (Dual Core)

TI, ARM Cortex A8 1.2GHz (Single Core)

Marvell, ARM 1.2GHz (Single Core)

Memory 512 MB 1 GB 512 MB

Network 10/100/1000Base-T 10/100/1000Base-T 10/100Base-T

Power Consumption (working) (idle)

>5W >2W

~24W ~9W

~24W ~9W

Network Wi-Fi Yes1 No No

USB Web-Cams Yes No No

USB Head-set Yes Yes Yes

USB Printer Yes Yes Yes

DeviCe maNagemeNT

Management software vSpace Management Center

HP Device Management Suite

HP Device Management Suite

Device Administration Yes Yes Yes

Web-hosted Yes No No

Device Profile Management

Yes No2 No2

Backup & Restore Yes Yes Yes

Device Health Monitoring Yes Yes Yes

Asset Management Yes Yes Yes

Easy Device Upgrade (firmware / config)

Yes No3 No3

Device Groups Yes Yes Yes

Server Groups Yes No No

SeCuriTy

Secure tunnel (HTTPs) Yes Yes Yes

Hardware encryption Yes No No

Hardware security (Kingston security)

Yes Yes Yes

SuPPOrT

Premium Support Included No (purchase) No (purchase)

Future Upgrades With premium support Limited Limited

Hardware Support 3 Year Warranty 3 Year Warranty 3 Year Warranty

While selecting a thin client for a virtual desktop environment, IT administrators should consider criteria that provide a comprehensive view of the cost of ownership of the virtual computing deployments.

These criteria should include:

• Ease of use

• User experience

• Security

• Device Performance

• Device management

• Return on investment

Here at NComputing, we took a similar approach and evaluated NComputing’s HDX ready SoC based N500 versus HP HDX Ready t410 and t5325.

We conducted multiple tests from an enterprise user view, using applications such as:

• MS Office Word, Excel and PowerPoint

• Adobe PDF Reader 11

• Windows Media Player 12 with K-Lite Codec pack

Additionally, each platform was verified for expansion options as well as support for dual display, USB web cams, legacy connectivity options and video quality over webcast.

All systems running on these thin clients were running Window 7 Professional SP1 in 1080P capable, multi-monitor configurations with playback scenarios of high-definition video in MP4 and WMV formats.

Background

1 WiFi is available on the N500w

2 Seperate software is requried

3 Available from FTP Server

Page 3: N-Series SoC Based Thin Clientscloudmty.com/.../04/SERIE-N-CITRIX-vs-THIN-CLIENTS-HP.pdf1 of 5 Competitive nalysis N-Series SoC Based Thin Clients User experience and performance analysis

3 of 5

Competitive Analysis

Software Performance

PowerPoint®

This test simulates a user presenting high resolution imagery embedded in a slide using PowerPoint 2010 automated slide show option. This test starts with opening a PowerPoint slide show that contains photos, diagrams, animations and backgrounds with color gradients. The slide show displays each slide for 3 seconds.

Excel®

This test was performed in a dual-monitor environment and simulates a user working with Excel 2010 and performing various actions, such as opening an Excel sheet, resizing the Excel window to full screen, selecting all cells, scrolling down and up using the arrow keys and mouse wheel and finally closing the Excel sheet.

Test Results:

N500 had the average CPU utilization (13%) and maximum CPU utilization (25%) while HP t410 had average CPU utilization (18%) and maximum CPU utilization (29%) and HP t5325’s average CPU utilization (21%) and maximum CPU utilization (30%).

Test Results:

N500 had the average CPU utilization (7.6%) and maximum CPU utilization (15%) while HP t410 had average CPU utilization (12%) and maximum CPU utilization (21%) and HP t5325’s average CPU utilization (15%) and maximum CPU utilization (23%).

Microsoft Office System Performance: Microsoft Office productivity testing was conducted to determine the relative performance of each thin client solution. Workloads were designed to stress screen redraw with complex images, gradients and textures. To demonstrate the performance differences between solutions evaluated, we used scrolling through slide shows and spreadsheets as the typical office worker use cases.

CPU Utilization: Microsoft Excel

0%

5%

10%

15%

20%

25%

30%

35%

40%Max

Avg

Min

HP T5325HP T410N500

CPU Utilization: Microsoft PowerPoint

0%

5%

10%

15%

20%

25%

30%

35%

40%Max

Avg

Min

HP T5325HP T410N500

Microsoft, Excel, PowerPoint and Windows Media Player are trademarks of the Microsoft group of companies.

Windows Media Player® 11

This test was run with dual-monitor configurations for both N500 and HP t410 (Note: HP t5325 does not support dual monitors). Video playback was done in various formats (WMV and MP4) over Citrix HDX protocol.

On average, N500 provided lower resource utilization across all media formats with an average CPU utilization of 18% and maximum of 25%, versus HP t410, which was scaling at average CPU utilization at 70% and Max at 90; we had the same results against t5325 where average CPU utilization was 70% and Max was 95% (Refer to Figure 1 above).

In other tests to verify the quality of video or EUVE (End User Video Experience), N500 delivered a clean and smooth video where audio and video were in sync and no choppiness was observed versus HP t410 where the video was choppy and out of sync with audio. Video was often jumping frames and similar experience with t5325.

The utilized video file was recorded at 720P and 30 fps from an animated movie. Duration of video was 2 minutes and 37 sec.

Figure 1: Windows Media Player 11

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%Max

Avg

HP T5325HP T410N500

CPU Utilization: Windows Media Player

Page 4: N-Series SoC Based Thin Clientscloudmty.com/.../04/SERIE-N-CITRIX-vs-THIN-CLIENTS-HP.pdf1 of 5 Competitive nalysis N-Series SoC Based Thin Clients User experience and performance analysis

4 of 5

Competitive Analysis

Performance vs PriceWe found that the N500 thin client provides very comparable performance for productivity tools to HP t410 and t5325 but outperforms HP thin clients in multimedia tasks. The price point for N500 is 20% less than HP t410 and it ships with mounting accessories (VESA Kit) whereas a separate accessory must be purchased with HP t410 for mounting purposes.

In addition to performance at lower price, N500 provides a broad set of features and functions (see table on page 2).

Future Proof N-Series thin clients are equipped with one firmware methodology, which means the Citrix receiver is packaged with the firmware of N-Series, making it easily upgradable and robust for any future releases of ICA receiver integrated into the N-Series firmware file. For HP thin clients, administrator must download Citrix receiver separately along with the thin client OS or image, making it less flexible.

Energy Savings Since N500 only consumes 5W of power, it can result in significant power savings for the organization. We calculated the energy savings by using (Time x Power x Money) formula and standard electricity rates in USA ($0.13 per unit); the results are:

N500 HP T410 HP T5325

eNergy & iNSTallaTiON

Power consumption >5W 24W 35W

Electricity cost (per unit) $0.13 $0.13 $0.13

Hours utilized 8 8 8

Idle hourse 16 16 16

1 day expense $0.93 $4.30 $5.53

3 years energy cost $101.83 $470.85 $605.53

VESA mounting kit Included $20-35 $35

Comparison Criteria Testing was conducted on functionally-equivalent NComputing and HP thin clients side-by-side. We determined the system performance as a subjective comparison of these systems, user experience, video quality and manageability. Video comparison was rated on the scale of 1 - 5 with 1 being poor and 5 being best. The scores are:

FOrmaT N500 HP T410 HP T5325

WMV 3.5 1 1

MP4 3.5 1 1

Flash 3 3 3

Adobe Reader®

This test simulates a user working with Adobe Reader 11 and taking actions such as open a PDF file (with rich contents), scroll down/up using the arrow keys and mouse wheel, observe any choppiness v/s a smooth scroll and finally, save and close the file and shut down Adobe Reader.

Test Results:

N500 had the average CPU utilization (8%) and maximum CPU utilization (25%) while HP t410 had average CPU utilization (10%) and maximum CPU utilization (35%) and HP t5325’s average CPU utilization (11%) and maximum CPU utilization (35%).

CPU Utilization: Adobe Reader

0%

5%

10%

15%

20%

25%

30%

35%

40%Max

Avg

Min

HP T5325HP T410N500

Software Performance (continued)

Adobe Reader is a trademark of Adobe Systems Incorporated.

Return on Investment

Page 5: N-Series SoC Based Thin Clientscloudmty.com/.../04/SERIE-N-CITRIX-vs-THIN-CLIENTS-HP.pdf1 of 5 Competitive nalysis N-Series SoC Based Thin Clients User experience and performance analysis

5 of 5

Competitive Analysis

868976_V4

3979 Freedom Circle, Suite 600, Santa Clara, CA 95054 | Phone: 1-408-380-8400 | Fax: 1-408-380-8401 | ncomputing.com

© Copyright 2012. NComputing, Inc. All rights reserved. NComputing is the property of NComputing. Other trademarks and trade names are the property of their respective owners. Preliminary beta specifications subject to change without notice. Performance may vary, depending on the configuration of the shared computer

Testing Setup and Methodology

In order to create a customer environment, we used a Gigabyte Ethernet Switch attached to a domain where Active Directory was used to authenticate users.

The server infrastructure was running Xen Server 6.0.2 over Dell PowerEdge R610, equipped with:

• Intel Dual Core Xeon E5530 @ 2.4 GHz

• 10 GB RAM

• 1 GB Network Interface Card (Broadcom NetXtream II 4 Port)

Xen-Desktop 5.6 was used as main VM desktop provider, and Xen APP 6.5 was used to publish the application to end users.

All the platforms were running Windows 7 Professional SP1 as their base image connected to two 22” HD 1920x1200 displays, on which all the test where performed.

Five Reasons for Choosing N-series SoC Thin Clients Compared to HP1. Best value: Compared to other venders in the market and with an average price between $142

- $205 (depending on model), NComputing offers one of the lowest cost HDX ready thin clients without compromising the performance.

2. Performance wins: Because of the offloading and intelligent use of the NUMO3, SoC components, performance is dramatically improved. All the HDX features are seamlessly delivered leaving processor headroom for more multitasking.

3. The N-Series is future proof: Since the N-series is SoC based, standard codecs such as H.264, MPEG-4, and VC1 along with hardware encryption are processed on the Numo chip, reducing processing time and saving server resources. The Citrix Receiver is prepackaged in the device firmware, making receiver upgrades fast and easily manageable.

4. lower energy costs: With power consumption at >5Watt, N-series can offer drastic savings to energy bill.

5. Smaller form factor and mounting kit: Each N-Series device is shipped with VESA mounting kit that attaches to the back of the monitor, leaving a clean and clutter free work desk for end users.