{"id":29703,"date":"2023-01-12T18:53:10","date_gmt":"2023-01-12T18:53:10","guid":{"rendered":"https:\/\/mailinvest.blog\/index.php\/2023\/01\/12\/amd-leads-for-aws-based-instances-in-cockroach-labs-2022-cloud-report\/"},"modified":"2023-01-12T18:53:10","modified_gmt":"2023-01-12T18:53:10","slug":"amd-leads-for-aws-based-instances-in-cockroach-labs-2022-cloud-report","status":"publish","type":"post","link":"https:\/\/mailinvest.blog\/index.php\/2023\/01\/12\/amd-leads-for-aws-based-instances-in-cockroach-labs-2022-cloud-report\/","title":{"rendered":"AMD leads for AWS-based instances in Cockroach Labs’ 2022 Cloud Report"},"content":{"rendered":"

<\/a>\n
<\/p>\n

\n

Benchmarking has long existed as an objective way to measure the speeds of CPUs, graphics chips and storage hardware.<\/p>\n

And in Cockroach Labs Inc.\u2019s \u201c2022 Cloud Report<\/a>,\u201d Advanced Micro Devices Inc. emerged with the best performance ratio tested on AWS instances: The AMD Milan chipsets set the pace for CPU benchmarks.<\/p>\n

\u201cWe look<\/span>\u00a0at micro-benchmarks around CPU, network, and storage,\u201d said\u00a0Keith McClellan<\/a> (pictured, right), director of partner solutions engineering at Cockroach Labs.<\/span>\u00a0\u201cAnd then our flagship benchmark is we use the database<\/span>\u00a0itself where we have the most expertise to create<\/span>\u00a0a real-world benchmark across all of these instances.<\/span>\u00a0This year I think we tested<\/span>\u00a0over 150 different discrete configurations.\u201d<\/span><\/p>\n

McClellan and\u00a0Jeff Bloom<\/a> (pictured, left), senior director of the Cloud Business Group at AMD, spoke with theCUBE analyst\u00a0Savannah Peterson<\/a>\u00a0during the recent\u00a0<\/span>AWS re:Invent<\/span><\/a>\u00a0conference, during an exclusive broadcast on theCUBE, SiliconANGLE Media\u2019s livestreaming studio. They discussed the report\u2019s findings both in relation to AMD chips and the end users themselves. <\/span>(* Disclosure below.)<\/span><\/i><\/p>\n

Fast but also cost-efficient<\/h3>\n

While it once lagged in the consumer hardware segment, AMD has\u00a0turned its fortunes<\/a> (and consumer perceptions) around. That confidence has bled from its well-known consumer products to enterprise high-performance computing use cases, such as data centers.<\/p>\n

\u201cWhat it means for us<\/span>\u00a0and for our customers is there\u2019s a good breath and depth<\/span>\u00a0of testing that has gone on from the lab,\u201d Bloom explained.<\/span>\u00a0\u201cAnd you look at this cloud report,<\/span>\u00a0and it helps them traverse this landscape of\u00a0<\/span>why to go with instance A, B, or C on certain workloads.<\/span>\u00a0And it really is very meaningful because<\/span>\u00a0they now have the real data<\/span>\u00a0across all those dimensional kinds of tests.\u201d<\/span><\/p>\n

That effort has resulted in the report citing<\/a>\u00a0that AMD Milan processors, based on the EPYC architecture\u2019s third generation, is setting the pace in performance and, more importantly, compute per dollar. In essence, customers get the highest performance at the lowest cost.<\/p>\n

\u201cWhen we\u2019re looking at instances,<\/span>\u00a0we\u2019re not just looking at how fast something is.<\/span>\u00a0We\u2019re also looking at how much it costs to get that level<\/span>\u00a0of performance, because CockroachDB as a distributed system<\/span>\u00a0has the opportunity to scale up and out,\u201d McClellan added.<\/span><\/p>\n

Here\u2019s the complete video interview, part of SiliconANGLE\u2019s and theCUBE\u2019s coverage of <\/span>AWS re:Invent<\/span><\/a>:<\/span>
<\/span><\/p>\n