Intel’s Comet Lake Brings Performance at a Budget
Posted on June 24, 2020
The last couple of years have been quite interesting for the microprocessor industry. AMD shook up the market with the announcement of their groundbreaking Ryzen processors, ending half a decade of unchallenged dominance by Intel.
With its combination of power and affordability, the Ryzen series upended the competition, and suddenly, everyone was swearing by AMD. While Intel’s Skylake processor did perform better than the Ryzen on single-core performance, there was little else it had going for it, and sales tanked.
After watching its customer base slipping away steadily, Intel has finally risen to the challenge. And its answer? The Comet Lake.
With 32 processors announced across the price spectrum, Intel is throwing its all into this launch. But is the Comet Lake really up to the challenge? Let’s find out.
Improvements Brought by the Intel Comet Lake
Usually, new microprocessor generations are built on a whole new lithography, but the Comet Lake uses the same 14nm process that was used to manufacture the Skylake chips. As a result, there is no massive leap in its capabilities from its predecessor.
The differences lie in small, incremental changes, which individually do not amount to much, but taken together create a significant improvement.
You may have heard of terms like Duo-core and Quad-core before. What do they mean?
Microprocessor chips are complex things. Containing a precise architecture of logic gates and semi-conductor pathways, these microscopic wafers are responsible for carrying out all computational processes of the computer.
The only drawback? They cannot really multitask. Whenever the CPU receives a series of instructions from the computer, it must process them one at a time. This slows down the chip’s speed considerably.
To overcome this shortcoming, companies these days embed multiple processor cores on a single chip. Each additional core is another site for processing, allowing the CPU to deal with multiple instructions simultaneously.
While earlier Intel processors were duo, quad, and occasionally octa-core, the Comet Lake line up has increased the physical cores on every product up to ten cores.
Perhaps the most important feature in this new line of processors is the universal support for hyperthreading.
Hyperthreading is Intel’s proprietary technology that allows even a single-core processor to function as if it had two cores. For example, a quad-core processor can normally take care of four instructions simultaneously. With hyperthreading, it can process eight at a time.
It should be noted that hyperthreading does not actually increase the processing power of the chip, but only allows more efficient use of its existing capacity. As most instructions rarely tax a processor to its fullest, hyperthreading makes use of the idle power to process multiple simple instructions at once.
Earlier, hyperthreading was reserved for the most expensive chips. But with the release of the Comet Lake series, Intel has introduced hyperthreading on each and every chip. Not only does this greatly boost performance, but it also marks a big advantage over AMD processors lacking any such technology.
Thermal Velocity Boost
In recent years, overclocking the CPU has become popular as a way to boost performance. Thanks to the high yields of good quality silicon in the manufacturing process, many low-cost chips can actually perform far above their weight class, making overclocking a very beneficial prospect.
To help users get the most out of their processors, Intel had introduced the Thermal Velocity Boost (TVB) technology. With TVG, all processor cores can boost their clock speed to the tune of 100 MHz as long as the temperature is maintained below a specified minimum.
This is in addition to the gains made through Turbo Boost, another Intel technology that enables processors to outperform their official limits.
Is it the right processor for embedded PCs?
Industrial applications are more sensitive to performance than casual home users. Even gains of just 100 MHz on each chip can greatly boost the efficiency of a large-scale server rack, to say nothing of the gains made through in an increased number of cores.
Furthermore, while software aimed at home users are seldom designed to take advantage of multi-threading, commercially deployed apps rely heavily on parallel processing capabilities.
This means that embedded computers can take advantage of Hyperthreading to achieve better response times, carrying out a large number of smaller tasks far quicker.
Comet Lake U
The U series is the low power guzzling version of the Comet lake processor, meant for building compact systems like fanless computers with a smaller energy footprint.
The Comet Lake U is also suitable for AI heavy applications, thanks to the large number of physical cores that complement the highly parallel nature of AI and graphics processing quite well.
With the discontinuation of the Haswell chipsets that were the mainstay for many industries, the Comet Lake U series of chips are poised to become more widespread in the coming future.
Contact your Global American Sales Rep today to find out about COMING SOON Embedded Comet Lake boards and systems.