Skip to main content

Intel, North Carolina State researchers develop chip that boosts performance

transistor design chaos theory intel xeon processor e7 v4
Image used with permission by copyright holder
As much as multithreaded applications can take advantage of much more powerful processors and many more cores than they could a few years ago, there are still a number of constraints that prevent the full power of multiple cores from being utilized. One of those is the software transfer process of data between the cores, but that’s about to change thanks to Intel and North Carolina State University researchers, who have developed a new hardware solution to bridge the cores.

The development is known as the core-to-core communication acceleration framework, or CAF for short. In an interview with TechSpot, Yan Solihin, co-author of the paper that explains its benefits, said CAF can improve the performance in communication between cores by up to 12 times, which can ultimately lead to calculations being performed around twice as fast as before.

This sort of effect is likely to be most pronounced in systems with hyperthreading (or the AMD equivalent) and with lots of cores, as well as with applications that can support it.

The reason such gains are said to be possible is because of the new chip design, which incorporates a queue management device (QMD) — essentially a hardware alternative to the software method of sending data between cores. This fast track handles all of the communication between the cores at a much higher pace than can be achieved with conventional processor designs.

Better yet, when not required for heavy multitasking, the queue management device can handle basic computational tasks on its own, potentially giving us a look at how Intel may be able to lower power usage in future chip designs — by letting the QMD handle basic operations and only leveraging the main power of the processor when required.

Of course, much of this is in the very early stages of development, and even though the proof of concept is there, we wouldn’t expect this to show up in retail hardware for years. However, when it does, the university may have additional features in mind, as Solihin and his team have already conceptualized a number of additional processor add-ons that could further improve communication between cores.

Editors' Recommendations

Jon Martindale
Jon Martindale is the Evergreen Coordinator for Computing, overseeing a team of writers addressing all the latest how to…
AMD’s 12-core Ryzen 3900XT claims a 47% performance boost over Intel’s best
AMD Ryzen 9 3900x

After announcing its new Ryzen 3000XT desktop processors, AMD has made some ambitious performance claims about its chips. Specifically, the company has shown how 3000XT processors can outperform competing CPUs from Intel, including the best ninth-generation CPU, the Core i9-9900K.

Based on AMD's optimized 7nm node and utilizing the Zen 2 processor architecture, the new Ryzen 3000XT series starts at $249 with the Ryzen 5 3600XT. The family includes upgraded configurations like the $399 Ryzen 7 3800XT and the premium $499 Ryzen 9 3900XT. Unlike the standard non-XT Ryzen 3000 processors, the XT chips gain up to a 200 MHz boost in clock speeds thanks to refinements to the 7nm node to squeeze out more performance, AMD claims.

Read more
Intel Rocket Lake-S to use hybrid chiplet architecture for 25% performance boost
Stock photo of Intel 9th gen core processor.

Intel's new Xe graphics architecture may soon arrive as an integrated desktop graphics solution for its upcoming Rocket Lake-S processor. With Rocket Lake-S rumored to launch later this year, barring any development or production delays caused by current uncertainties surrounding the global COVID-19 pandemic, we could see Intel's integrated Gen 12 graphics, which are based on Xe, paired with this new processor.

While Intel has not confirmed Rocket Lake-S, the processor series was previously mentioned in a leaked processor road map. It's believed that Intel will utilize a hybrid chiplets approach, combining a 14nm process for the processor with a 10nm process for its Gen12 graphics based on the Xe architecture.

Read more
Intel develops a cryogenic control chip expected to streamline quantum computing
Photo of Intel's Horse Ridge chip

Intel’s latest control chip is expected to help streamline the development of “commercially viable” quantum computers.

Intel announced the creation of a new control chip to be used with quantum computers via a press release posted on its Newsroom webpage on Monday, December 9. Code-named Horse Ridge, the new chip is notable because it’s expected to be able to control “multiple quantum bits” (or qubits) without having to rely on wires to transmit data to and from them.

Read more