ClickHouse/website/benchmark/hardware/index.html
2022-04-24 09:12:33 +03:00

104 lines
5.8 KiB
HTML
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

{% extends 'templates/base.html' %}
{% set title = 'Performance comparison of ClickHouse on various hardware' %}
{% set extra_js = [
'queries.js?' + rev_short,
'results.js?' + rev_short,
'../benchmark.js?' + rev_short]
%}
{% set url = 'https://clickhouse.com/benchmark/hardware/' %}
{% set no_footer = True %}
{% block content %}
<div class="container-fluid py-3">
<div class="row mb-3">
<div class="col d-flex align-items-center">
<a href="/" title="Main page" class="float-left mr-3">
<img src="/docs/images/logo.svg" alt="ClickHouse" />
</a>
<h1 class="h2 mb-0">{{ title }}</h1>
</div>
</div>
<div class="row mb-3">
<div id="selectors" class="col"></div>
</div>
<div class="row mb-3">
<div class="col">
<h2 class="h4 my-3">Relative query processing time <span class="text-muted">(lower is better)</span></h2>
<div id="diagram"></div>
</div>
</div>
<div class="row mb-3">
<div class="col">
<h2 class="h4 my-3">Full results</h2>
<div id="comparison_table" class="overflow-auto"></div>
</div>
</div>
<div class="row">
<div class="col">
<h2 class="h4 my-3">Comments</h2>
<p>Submit your own results: <a href="https://clickhouse.com/docs/en/operations/performance-test/">https://clickhouse.com/docs/en/operations/performance-test/</a></p>
<p>
Results for Lenovo B580 Laptop are from <b>Ragıp Ünal</b>. 16GB RAM 1600 GHz, 240GB SSD, Intel(R) Core(TM) i5-3210M CPU @ 2.50GHz (2 Core / 4 HT)<br/>
Results for Time4vps.eu are from <b>Ragıp Ünal</b>.<br/>
Results for Dell PowerEdge R640, R641 (in Hetzner) are from <b>Dmirty Titov</b>.<br/>
Results for Dell PowerEdge R730 are from <b>Amos Bird</b>.<br/>
Results for Dell R530 are from <b>Yuriy Zolkin</b>.<br/>
Results for Xeon 2176G are from <b>Sergey Golod</b>.<br/>
Results for Azure DS3v2 are from <b>Boris Granveaud</b>.<br/>
Results for AWS are from <b>Wolf Kreuzerkrieg</b>.<br/>
Results for Huawei Taishan are from <b>Peng Gao</b> in sina.com.<br/>
Results for Huawei Taishan (2) are from <b>Kurmaev Roman</b> at Huawei.<br/>
Results for Selectel and AMD EPYC 7402P are from <b>Andrey Dudin</b>.<br/>
Results for ProLiant are from <b>Denis Ustinov</b>.<br/>
Results for AMD EPYC 7502P 128GiB are from <b>Kostiantyn Velychkovskyi</b>.<br/>
Results for AMD EPYC 7502P 512GiB are from <b>Sergey Zakharov</b>.<br/>
Results for Pinebook Pro are from <b>Aleksey R. @kITerE</b>.<br/>
Results for AMD Ryzen are from <b>Alexey Milovidov</b>. Firefox was running in background.<br/>
Results for Azure E32s are from <b>Piotr Maśko</b>.<br/>
Results for MacBook Pro are from <b>Denis Glazachev</b>. MacOS Catalina Version 10.15.4 (19E266). For "drop caches", the "Free Up RAM" in CleanMyMac is used.<br/>
Results for AMD EPYC 7702 are from <b>Peng Gao</b> in sina.com.<br/>
Results for Intel NUC are from <b>Alexander Zaitsev</b>, Altinity.<br/>
Xeon Gold 6230 server is using 4 x SAMSUNG datacenter class SSD in RAID-10.<br/>
Results for Yandex Managed ClickHouse for "cold cache" are biased and should not be compared, because cache was not flushed for every next query.<br/>
Results for AWS Lightsail is from <b>Vamsi Krishna B.</b><br/>
Results for Dell XPS laptop and Google Pixel phone is from <b>Alexander Kuzmenkov</b>.<br/>
Results for Android phones for "cold cache" are done without cache flushing, so they are not "cold" and cannot be compared.<br/>
Results for Digital Ocean are from <b>Zimin Aleksey</b>.<br/>
Results for 2x EPYC 7642 w/ 512 GB RAM (192 Cores) + 12X 1TB SSD (RAID6) are from <b>Yiğit Konur</b> and <b>Metehan Çetinkaya</b> of seo.do.<br/>
Results for Raspberry Pi and Digital Ocean CPU-optimized are from <b>Fritz Wijaya</b>.<br/>
Results for Digitalocean (Storage-intesinve VMs) + (CPU/GP) are from <b>Yiğit Konur</b> and <b>Metehan Çetinkaya</b> of seo.do.<br/>
Results for 2x AMD EPYC 7F72 3.2 Ghz (Total 96 Cores, IBM Cloud's Bare Metal Service) from <b>Yiğit Konur</b> and <b>Metehan Çetinkaya</b> of seo.do.<br/>
Results for 2x AMD EPYC 7742 (128 physical cores, 1 TB DDR4-3200 RAM) from <b>Yedige Davletgaliyev</b> and <b>Nikita Zhavoronkov</b> of blockchair.com.<br/>
Results for ASUS A15 (Ryzen laptop) are from <b>Kimmo Linna</b>.<br/>
Results for MacBook Air M1 are from <b>Denis Glazachev</b>.<br/>
Results for Xeon Gold 6140 are from <b>Shiv Iyer</b> (ChistaDATA Labs).<br/>
Comparison of EBS and EFS is from <b>Ramazan Polat</b>.<br/>
Results for Hetzner and Scaleway are from <b>Anthony Najjar Simon</b> (Panelbear).<br/>
Results for GCP are from <b>Vy Nguyen Tan</b>.<br/>
Results for ThinkPad P15 are from <b>Mikhail Shiryaev</b>.<br/>
Results for RockPi4 are from <b>Kirill Zholnay</b>.<br/>
Results for Xeon 6266C are from <b>David in Shanghai</b>.<br/>
Results for SSDNodes and Cavium are from <b>Lorenzo QXIP</b>.<br/>
Results for AMD EPYC 7662 64-Core Processor are from <b>Evgeniy Kuts</b>.<br/>
Results for scaleway GP1-S 8x x86 64bit 32GB ram 300gb NVMe are from <b>Dag Vilmar Tveit</b>.<br/>
Results for scaleway GP1-M 16x x86 64bit 64GB ram 600gb NVMe are from <b>Dag Vilmar Tveit</b>.<br/>
Results for Intel(R) Core(TM) i5-4440 CPU @ 3.10GHz are from <b>Peter, Chun-Sheng, Li</b>.<br/>
Results for MacBook Pro M1 are from <b>Filatenkov Arthur</b>.<br/>
Results for AWS instance type im4gn.4xlarge are from <b>Ananth Gundabattula</b> (Darwinium).<br/>
Results for AWS instance type im4gn.8xlarge are from <b>Ananth Gundabattula</b> (Darwinium).<br/>
Results for AWS instance type im4gn.16xlarge are from <b>Ananth Gundabattula</b> (Darwinium).<br/>
Results for AWS instance type i3.2xlarge are from <b>Ananth Gundabattula</b> (Darwinium).<br/>
Results for 2x EPYC 7702 on ZFS mirror NVME are from <b>Alibek A</b>.<br/>
Results for Intel 11th Gen Core i9-11900KF are from <b>Tim Xian</b>.<br/>
Results for AWS instance type m5a.4xlarge are from <b>Daniel Chimeno</b>.<br/>
</p>
</div>
</div>
{% endblock %}