Jump to content
Facebook Twitter Youtube

[Software] Intel manages to improve the performance of its processors by 40%


FNX Magokiler
 Share

Recommended Posts

Many users, when they buy a processor, look for it to have overlocking capacity to increase its performance in the future if necessary. It seems that Intel has found a way to increase performance without increasing frequencies. By optimizing Intel software, a performance improvement of up to 40% would have been achieved.

Currently, we are approaching the limit of silicon before certain processes take place that can alter the work that the processor does. This makes reducing the size of the transistors and increasing both the IPC and the frequency more and more difficult. This is why Intel has been starting to pay more attention to the software section.

Significant performance increase in Intel processors
Without much noise, this same year Intel has acquired the Israeli company Granulate. This company was founded in 2018 and already Intel in 2019 began to collaborate with this company. A few months ago they acquired it for an amount of 650 million dollars.

Granulate is a company specializing in software optimization that can reduce costs for customers. According to the company, under certain workloads there may be a performance improvement of up to x5 times.

As is logical, after the acquisition of the company, they have prioritized the development of solutions for their own needs. During the VisiON conference held a short time ago, Intel released some information about its plans regarding software optimization. The company has used many of Granulate's solutions, even releasing a free tool.

 

00q5bcd.png

 

Intel's gProfiler tool is intended to analyze and see where the bottlenecks are. This tool is designed for Data Centers and other massive and advanced computing facilities. The tool, after analyzing the structure, finds where there are performance problems and thus optimizations can be made.

Based on information from Intel, software optimization on Java or Ruby workloads, CPU usage can be reduced by up to 30% without changing code. We are talking about a notable reduction in the workload, which leads to a reduction in costs and energy consumption.

Intel indicates that in third generation Intel Xeon Scalable processors, with this tool, the use of the processor can be improved by up to 40%.

Will this reach the end user?
Obviously, we are all interested in having extra performance without the need to do anything. The truth is that it seems quite difficult, at least for now, to end up reflecting on the end user. We must be clear that this solution is designed for advanced workloads.

While this is positive in many respects, it gives the impression that the consumer user is being forgotten by Intel. We must remember that about a month ago Intel should have released drivers for its new graphics cards and at the moment, no trace and no news.

The launch of the Intel Arc 3 graphics has been a disaster and the Arc 5 and Arc 7 have been delayed without further data. The release of Intel XeSS, the upscaling technology that must compete with NVIDIA DLSS and AMD FSR, has also been delayed.

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
 Share

WHO WE ARE?

CsBlackDevil Community [www.csblackdevil.com], a virtual world from May 1, 2012, which continues to grow in the gaming world. CSBD has over 70k members in continuous expansion, coming from different parts of the world.

 

 

Important Links