Explorar >   IBM PartnerWorld
 |  |  |  |  | 
by Transposh - translation plugin for wordpress

A translation widget is provided for your convenience to facilitate translation of the English language version of this blog into several languages. If you choose to utilize this automated translation facility, please understand there may be deviations between the automated translation and the original English version. IBM não é responsável por qualquer desses desvios de tradução automática e oferece a versão traduzida "COMO É" sem garantias de qualquer tipo.

EUROPE Infra-estrutura Partner Experience Em todo o mundo

IBM e M Computers: Mudando o futuro com tecnologias cognitivas

Eu (Andrey) recently caught myself thinking about why I’m not easily excited by new sci-fi stories. Everything I see in movies like Anon ou o Westworld series has become so common. Visual image recognition—so what? Artificial intelligence becoming more like human intelligence—not surprising at all.

There are serious reasons why I feel this way—I see such “miracles” in my daily life. It’s not unusual to see our consultants in IBM Systems Services Lab deploying PowerAI vision use cases, teaching neural networks and sharing what they can do. Por exemplo, recently I created a visual recognition prototype that detects people on roads and highways whether or not they’re wearing a hi-res vest. Cognitive technologies are changing the future, and they’ve become a big part of my everyday work and life.

In IBM Systems Lab Services, we’re helping clients and IBM Business Partners by making cognitive discovery more possible. Our most agile IBM Business Partners, with our assistance, are exploring and prototyping new opportunities around artificial intelligence, visual recognition and other cognitive solutions.

The M Computers story: Leveraging PowerAI technology

Let me introduce Petr Plodík, Sales Director of M Computers. IBM has worked closely with M Computers since 2016, and below Petr shares more about how they are changing the future.

Petr Plodík: M Computers is the leading company in the high-performance computing industry in the Czech Republic. Our customer base comes mainly from the public sector: universities, several departments at the Czech Academy of Science or other R&D institutions. M Computers, as a technology company, looks for the latest innovations to test, and wants to understand recent trends and bring the latest technology to customers.

The typical AI applications M Computers tested on IBM Power Systems™ are image/video recognition applications based on deep neural networks that process a large set of pictures in training mode and provide online processing for a runtime client. Our customers are also interested in IBM Power® hardware performance to see how their AI workload could benefit from IBM POWER9™ processors with their fast NVLink connection to Nvidia GPUs. A crucial part of image/video processing is the data preparation stage; in some cases, data management and optimization is more demanding than the neural network processing. This is very important for getting the best view of the source data, not only to put the original image in the computing process. Our tests confirmed Power and Nvidia innovations that can process input data in better resolution or process more data much faster. And we received positive feedback on the POWER9 processor performance; some of the AI applications we tested use relational databases or NoSQL databases to store processed data and results for various security reasons. The data I/O storage and processing is very fast on POWER9 processors, which also contributes to the overall faster response time of the AI processing.

For those who read Czech, here’s a comprehensive article about M Computers solutions in AI and machine learning.

Power Systems and IBM Spectrum Scale technology exploration

From the beginning, M Computers has been interested in accelerated IBM Power Systems technologies such as IBM POWER8® and POWER9 with Nvidia GPU and NVLink, even FPGA accelerators used for various workloads like I/O acceleration and especially the network stack.

Nós decided to invest in IBM Power technologies, so we started with a deep-dive education of our technical staff so we could provide end-to-end support for our different customers’ demandsfrom small HPC/IBM Spectrum Scale™ tests to larger cognitive workload POCs. During these sessions, we also developed a deep relationship with technology experts from IBM Systems Lab Services to get access to the latest evaluated technologies and have stable support for difficult architectural use cases. To be able to satisfy each customer’s needs to test IBM POWER hardware and software technologies, M Computers collaborated with IBM Czech Republic and IBM Systems Lab Services to get access to the Czech server resources (accelerated AC922 with 4GPUs) for several tests and POCs, with lab teams’ technical support. They are able to run there even the most complicated AI and deep learning workloads developed on a complex deep learning stack and the x86 platform.

It’s important to mention that M Computers is also the leader in IBM Spectrum Scale implementations within the Czech Republic and offers deep knowledge of the products as well as the large, petabyte-scale, HPC Spectrum Scale references.

Collaboration with IBM Systems Lab Services

M Computers already has a number of customers lining up to test their AI solutions on the accelerated IBM Power Systems servers.

Finally, IBM Systems Lab Services and M Computers are actively participating in EuroHPC initiative activities within the Czech Republic. This initiative gives IBM an opportunity to demonstrate the live pre-exascale HPC cluster solution the EU is looking for, as well as to contribute innovative technologies to the EuroHPC R&D program. Products like IBM Spectrum Scale and accelerated IBM Power Systems HPC hardware could be highly suitable solutions for this purpose.

The future starts now

Como você pode ver, we’re living in the future now, and IBM cognitive technologies are helping our next-generation ecosystem to explore this new world and grow their businesses.

And please reach out to IBM Systems Lab Services today if we can partner with you on a cognitive technology solution.

Andrey Safronov
Parceiro de Negócios Líder Técnico Enablement, IBM Systems Lab Serviços Europa
Twitter: https://twitter.com/Safron_on_Twit
LinkedIn: https://www.linkedin.com/in/safronovandrei/

Petr Plodík
Sales Director, M Computers s.r.o.
LinkedIn: https://www.linkedin.com/in/plodik/

Watch an interview from the recent IBM Systems Technical University in Prague, in which John Ferris, IBM Systems Hardware Services & Training Business Unit Executive, sits down with Petr Plodík:


Andrey Safronov é o Partner Enablement Líder Técnico de Negócios para Serviços de Laboratório de Sistemas da IBM na Europa. Andrey gastou mais 20 anos na indústria de TI e serviço, com 10 anos em posições de gestão e liderança. Ele levou equipes regionais, assegurando elevados resultados de desempenho operacional, bem como estabelecer novas equipes e crescente suas ofertas de serviços em quantidade e qualidade. Desde 2016, Andrey tem-se centrado no reforço da rede de canais da IBM para aumentar o valor da IBM Systems para Parceiros de Negócios IBM. Você pode seguir Andrey no Twitter: https://twitter.com/Safron_on_Twit e no LinkedIn: https://www.linkedin.com/in/safronovandrei/

Petr Plodík is the Sales Director for IBM Business Partner M Computers s.r.o. Petr has spent more than 15 anos na indústria de TI, with almost 10 anos em posições de gestão e liderança. He led the IBM System x® team, Distributor IBM division, assegurando elevados resultados de desempenho operacional, and established new ways of doing business. In all his roles, Petr has focused on growth and business results, and for the past two years at M Computers has concentrated on HPC business and AI technologies. Petr also teaches at a high school in Prague and is a drummer in his band. You can follow Petr on LinkedIn.

Related Articles


Seu endereço de e-mail não será publicado. Campos obrigatórios são marcados *

Nome *

Email *

Web site