At a recent IT conference on becoming more strategic, the topic of machine learning and artificial intelligence kept creeping into the conversation. Whether IT professionals are using AI and ML to improve their security posture or speed business decision-making, it’s becoming apparent that virtually every senior IT professional at least understands that potential that modern data science brings to IT and the business.
What surprised me is the disconnect between the general understanding of these principles and the actual adoption of them in any kind of a broad sense. Although IT leaders are looking for products that can help automate and orchestrate IT and business processes alike, there is not a deep understanding of how to modernize their own existing code base to take advantage of the latest technology has to offer.
Part of the disconnect is the belief that existing IT infrastructure doesn’t have the horsepower needed for these advanced technologies, but even run-of-the-mill Intel servers in most enterprise deployments already offer a multi-core, multi-CPU platform that can deliver screaming performance – if you know how to make code ready for it.
One toolset that should be in every developer’s arsenal is the Intel® Parallel Studio XE. This combination of libraries, compilers, optimization and test tools provides everything IT pros, developers, data scientists and DevOps teams need to take advantage of every bit of the horsepower inside their compute arsenal. What kind of improvement? Intel has documented over 100x (that’s 10,000%) performance* improvement for code by using vectorization, multithreading, and multi-node parallelization and memory optimization, whether your code is C, C++, Python, or even FORTRAN (Yes, that’s STILL a big thing for data science).
Intel Parallel Studio XE lets developers build code faster by simplifying the process of creating parallelized code, thus supporting modernization that not only delivers performance and scalability, but improves portability by supporting both CPUs and FPUs, including Intel Xeon and others.
Why is this important? Because CHANGE IS HARD. Especially for seasoned IT pros. I know, I’ve been around the block a few times myself. So, having a toolset that enables IT to get more from existing code – especially as new workloads are constantly being added – is going to be adopted faster than having to start from scratch, infrastructure-wise.
Another benefit for Intel Parallel Studio XE is the ability to develop and deploy on a wide range of cloud platforms, which lets dev teams see how well performance scales across large clusters without having to purchase a ton of new gear, and further simplify workload migration and portability as business needs demand workloads shift between on-premises and cloud providers.
*Source: See Vectorize & Thread or Performance Dies Data. For more complete information about performance and benchmark results, visit www.intel.com/benchmarks.
