Intel working on new chips to compete with NVIDIA GPUs

Intel & NVIDIA – Combat for Turf

The battle between Intel and NVIDIA seems to have been locked in new combat for turf which is the booming data centre market and the core of this conflict is the technology which is changing the world- Artificial Intelligence. – AI. The revenue of NVIDIA in the recent quarter ended April 30, had augmented by 48% touching $1.94 billion in comparison to earlier year.

They had received a big revenue crash from its Data centre business that had recorded $409 million revenue in the first quarter of this fiscal, 186% up, year-on-year, due to the exponential increase in the spike in demand for specific type of microprocessor known as Graphic Processing Unit – GPU, prepared by NVIDIA. Most of the huge technology companies such as Amazon, Google, Microsoft, Facebook, IBM as well as Alibaba have installed the best Tesla GPUs of NVIDIA to control their data centres to perform machine learning in analysing data collected from the cloud and gain awareness.

The MD of NVIDIA, Vishal Dhupar, had commented that they had seen the PC era that had been followed by the mobile era and now envisage the emergence of the AI era. Once viewed only as a gaming technology, GPU are now making inroads in data centres driving initiatives around machine learning- ML together with artificial intelligence – AI.

Data Centres – Storehouses of Huge Volumes of Data

Data centres are said to be storehouses of huge volumes of data that tends to connect the compute power of GPUs in order to make sense of huge repository of data. The arrival of GPUs have assisted in redefining the concept of systems designs for several of the high-performance application in the commercial and the non-commercial sectors of the market, according to Rajnish Arora, VP of enterprise Computing Research, IDC. Could it mean that Intel would be losing market share in the data centre space?

Arora had stated that they do not believe that Intel in losing share of the server CPU market to NVIDIA. The numbers evidently determine that Intelseemed to have consolidated its presence and market share in the server market. Presently Intel CPUs power data centres all across the globe and tends to revel in near control accounting for 95.7% of the entire server shipped in 2016.Moreover it is also betting big on AI and has financed greatly in start-ups that have assisted in building AI together with High-Performance Computing –HPC competences in its upcoming chips

Knights Mill/Knights Crest

Intel had bought Altera in 2015, for $16.7 billion, which was a company making programmable devices that could be optimised for AI and ML. It had gone ahead in August and had bought deep learning start-up Nervana Systems to strengthen AI specific solutions. The new chips called `Knights Mill’ will be sold later this year by Intel to address the growing Machine Learning – ML, market.

Moreover it is also working on `Knights Crest’ chips that will manage deep neural networks which is a component of deep learning – DL.A fellow at Intel Labs, Pradeep Dubey had stated that `competition has always been around and they are optimizing their chips for machine learning as well as deep learning applications since this is a big opportunity for them. There are also four new processors which will be released this year’. These products would be competing against NVIDIA GPU in the data centre market.

Intel unveils 18 core / 36 thread Core i9-7980XE

Intel was reeling under huge pressure from its competitors in past couple of months due to the launch of highly affordable range of processors. However Intel is making a comeback in a glorious fashion with the unveiling of new Core X series processors which will boast the 18 cores and 36 threads specification. Before this AMD was ruling the market with its Ryzen 9 1998X which had 16 cores and 32 threads along with the ThreadRipper processor.

This new offering from the Intel will be featured right alongside the other existing line up and it is expected to make use of the LGA 2066 pin socket that is usually found on the X299 motherboards. Intel has made this massive upgrade in the number of cores after a pretty long time that also at a better price point which has left a number of tech experts surprised.

Everything about the Intel Core X series

Before the official unveiling of the Core X series everyone was expecting it come with anything above 12 cores and no one was expecting the 16 or 18 cores offering at all. Intel has worked to keep the cores details under wrap and take the market by surprise as well subdue the consistent rise of the AMD varied SKU lineup with a remarkable product offering.

It is worth noting that the complete line up for the Skylake X and Kabylake X are designed to make use of the LGA 2066 socket and users will certainly get a wide count spread to choose from while powering their devices in future. The best thing about the new Intel platform is that they are highly scalable. Intel will be shipping the Core X series processors along with the latest version of the Intel Turbo Boost Max Technology 3.0 which will help in improving the single as well as dual core performances.

Secondly one should also note that the unveiling of the new X299 platform also brings the demise of the X99 platform. Intel has claimed that the new core X series will bring 10 percent faster multi-thread performance as well as 15 faster on the single threaded performance than the earlier generation processors.

More power comes with the Extreme Edition

Intel will also be bringing the Extreme edition CPUs which will take the performance levels to a whole new level for the computing activities. The new Core X series will allow the AVX-512 ratio offset along with the memory controller trim voltage control into the play. This functionality will help the users in getting a better stability at higher clocks while working on the robust projects.

This is specifically designed to handle the CPU heavy programs with ease with the existing programs without limiting the overall performance of the device. This will help in bringing fast video encode, audio production, image rendering as well as offering real time previews in the best resolution to the users.

Intel may be prepping new Core i5, i7 processors to attack AMD Ryzen

AMD will launch its new Ryzen in early March, and Intel is continuously making changes in order to give a good competition to AMD. Intel seems to be taking an intelligent move otherwise the chip that AMD is about to launch will turn out to be a potent threat to Intel. In 2003, Chipzilla was made to develop the Pentium 4 Extreme Edition so that they can survive in terms of the threat from Athlon 64, which is still known for wounding up so many winning games and several number of benchmarks for the consumers. Intel is concerned that they do not have to repeat the same situation, or they have to face the consequences if AMD would point change in prices that to in such a way that it might out run existing product stack of Intel.

Intel has also added Hyper-Threading to the family of Pentium, to avoid any risk of losing the market. We know that it is not common for a company to ramp up a faster version of core that to within a time period of just two months. This is generally heard to be done when a competitor launches their version of faster chip. But, since AMD is almost at the verge of presenting a vital architectural revision, so it is understandable on the part of Intel that why it is bent on propagating new launches of its own.

This can be viewed in a different way as it is all about manipulating all the competitors of the market by arranging them in a line so that they rage a war even if they do not have a great business plan in their mind. For more than a decade, Intel maintains a stable desktop plan. Pentium chips are known to be standing on modern Core architectures, but they move with a very low speed that too without Hyper-Threading. Whereas, Core i3 chips maintain a much faster speeds, along with Turbo frequencies, and Hyper Threading. Core i5 gives the same speed as Core i3 but are equipped with four cores that too without HT, whereas Core i7 are the highest ranks Cores of Intel that includes a HT back.

Intel have already lost the ground by switching on to Hyper-Threading in terms of its newly developed processors, but it would not be same as the difference that lies at the high speed. The difference between chips of Core i5 and Core i7 stands banal if Core i5 includes HT with it. Research have revealed that it does not really matter that how many Cores Intel go on increasing it will cost them the same price. and even if Intel is planning to make any changes, it is obvious that they do it on their own timeline instead of AMD’s.

Before reaching to any conclusion we need to keep in mind that this is not the time of Good Oi Days, when the frequency of CPU will flow almost like any liquid. Thus, necessary changes should be made in keeping this point in mind.

How can I tell if a PC Processor is any good?

Processors are the most important part of a PC, if you want to access your PC at its best condition including its speed then you need to make sure that the processor you are using is of the finest quality and provides you the optimum range of speed. A PC is far much better than a processor but it is also an undeniable fact that your PC cannot run without a processor. So for enabling all the benefits from your PC you need to choose a proper Processor for it.

It is said that every generation of core chips accolades through its own code name, Sandy Bridge, Haswell, and Skylake are of the same type. Kaby lake is said to be the latest version. Each generation accolades the users with several features leading to optimum amount of benefits. Some even make changes in the technology used for manufacturing, the shift is in terms of nanometers. The smaller is the chip the more benefits it bears. Core iX chip have undergone several changes in terms of its size, as it first decreased from 32nm to 22nm and today it has come down to 14 nm. This shrinking which is taking place enables the manufacturer to put in more features in every chip.

Changes started to take place with very first generation as the name started to come up with an addition of dash. For instance, i7-3770 core is a chip pertaining to third generation, while the seventh generation of chip that came up is identified as i7-7770 core, both belonged to the same version of CPU. There are many more chips belonging to the same processor and it is obvious that more higher is the number, the more features you are sure of availing for yourself.

The Intel processor also comprises of graphics co-processor, which is built including all the hi-tech technologies. This is also provided with a number as per the generation given. The seventh generation called Core i7 chip has more HD graphics in comparison to the sixth generation chip which has only 520 HD Graphics. So with the increase in number of generations the features also gets enhanced and the best graphics is sure to add an Iris Branding which will definitely be an assurance that it bears the best features in comparison to the other processors available in the market.

Everyone is now concerned about the performance of the computer in terms of game. The gamers always look for processors that enables super fast speed and they also look for processors that are small and does not consume more power. Chips that are suited to use less power and does not generate extra heat so that extra cooling is not required by them. So, we can find that it is the era of thinnest television and laptops that will have a strong battery that will allow the users to use them for a longer period. The best way to buy a CPU is to notice the name to understand its level of performance.

IBM Announces 7 nm Computer Chip Breakthrough

The latest breakthrough in technology seems to have overcome the main issue, which had posed one of the most "grand challenges" of the industry. IBM has created a test version of a semiconductor, which has the ability to shrink down the circulatry. This prototype is the first of its kind in the world. The microchip industry is always evolving and there are better products available in the market every new day. This ever evolving industry has to face immense difficulties due to the extent of their growth and due to various technological and physical limitations. But inspite of all the hindrances IBM has been successful in creating a chip with seven nanometers transistors.

Gradual advancement

The leader of this industry, Intel has been the recipient of several technical hurdles throughout. Technologists have even begun to question one of their foundation theories; they are questioning whether the long withstanding pace of chip improvement, the Moore's Law, could possibly continue beyond the present 14-nanometre generation of chips. Every generation of chip technology can be differentiated with the aid of the least size of fundamental component, which switches the current at nanosecond intervals. Now the advancement in technology has made the transition from the 14 nanometer to the manufacturing of 10-nanometer generation of chips.

Ever-evolving technology

This advanced technology created by IBM is basically an advanced version of the ultradense computer chips which has approximately four times the capacity of the other most powerful chips. As the advancements are brought forward in every generation of the chips, each of them brings forth an estimate of around 50% reduction in the region required by a given amount of circuitry. The new chips created by IBM are suggestive of a possibility that the semiconductor technology shall continue to shrink at least through 2018.

Individual identity

The advanced feature in these chips are the presence of the This advanced technology created by IBM is basically an advanced version of the ultradense computer chips which has approximately four times the capacity of the other most powerful chips. IBM has used silicon-germanium in their chips. In the other variants of chips, pure silicon is used. But IBM instead of using pure silicon has opted for silicon-germanium, which happens to be the key ingredient of the molecular-size switches. This also aids in faster transistor switching and in the lower power requirements.

Yet to be released in the market! 

However, it must be duly noted that most of these are mere speculations and experimentation. Nothing has been launched in the market as of yet. There are numerous complex procedures, which will test whether the product is viable for sale in the market. IBM shall take the final decision in collaboration with their development partners at the State University of New York's Polytechnic Institute in upstate New York. In the previous year, IBM had released a statement, which had declared their 3 billion dollars investment over a time period of five years to forward their chip technology and meet the demands of cloud computing, mobile products, data and other technologies.

Intel confirms the Haswell chip with 4.5W SDP

In the previous IDF San Francisco, Intel promised: Haswell will go down to 4.5 watts of SDP. While the first models announced earlier during the launch failed to reach such figures, but internal documents evoked such product have just been confirmed. Haswell (Core Processors fourth generation) will be of only a few watts! If the information seems to surprise some, it is not really surprising. Intel had indeed indicated that there is nearly a year that the chip goes beyond Ivy Bridge. The latter was to 7 watts of SDP for some particular models, and 10 watts to TDP. The Atom and his successors, they are limited to about 2 watts in terms of the products currently covered by the mark in terms of mobility. In June, we already learned that the purpose was reached already down to 6 watts for 2-in-1 and the Ultrabook, warning that he did not intend to stop there. And it's now official; there will also be versions of 4.5 watts only SDP that will emerge during the second half of 2013. And they will "build the first 2-in-1 passive based on Core architecture." It remains to know the price of this product and its performance. These will necessarily be limited, especially as the SDP wants a realistic indicator of power consumption under conditions of use. Hopefully accessible products may be right and it is not only for mini-luxury machines.

Samsung unveils Exynos 5 Octa 5420 processor

As expected, Samsung has unveiled its Exynos 5 Octa 5420. More than just a rise in frequency, it makes more changes in it. A new ARM Mali GPU which is a more efficient memory controller is loaded. Samsung has just announced its second Exynos 5 Octa, based again on the ARM architecture big. It latter allows for reminder to have in a single chip with two quad-core CPUs that are used depending on the needs: Cortex A7 one side to the low power consumption and a much more powerful Cortex A15. The current Exynos 5 Octa 5410, found in some versions of the Galaxy S4, is clocked at 1.2 GHz for the CPU Cortex-A7 Cortex-A15 while running up to 1.8 GHz. For the new version, the 5420, only the first 100 MHz wins. The Korean giant opportunity to change the dual channel memory controller which now supports the LPDDR3 @ 933 MHz, a theoretical bandwidth of 14.9 Gb/s. Another change which is not only limited with its GPU. In fact, Samsung comes with the ARM Mali T628, while the 5410 has a PowerVR SGX544MP3 solution. Anyway, this GPU is advertised as 10x more powerful in comparison with ARM Mali 400MP found on Exynos 4212 or 4412, but more than 2x the PowerVR chip used on the Exynos 5 Octa 5410. However, Samsung has, apparently, removed a pair of processing unit shaders compared to the reference design for a total of six instead of eight. It will be interesting to see what gives the final outcome. Supported the definition is identical to the previous generation, or 2560 x 1600 pixels. Added by the manufacturer against a MIC (Multiple Image Compression), supposed to offload some tasks normally intended for the GPU. Regarding the wireless display, it is supported up to a maximum resolution of 1920 x 1080 pixels.

 Samsung says it demonstrates the performance of the beast at SIGGRAPH being held in Anaheim, while mass production will start from next month. There remains a question. Samsung; will it use this Exynos 5 Octa 5420 in his Note 3 should be announced at the upcoming IFA in Berlin in early September, and the brand will it do that again the Galaxy S4 and its many variants performance very different? With volume production so late, it's a safe bet that this should be the case again. Unless of course the manufacturer decided to rely on Qualcomm for their next Smartphone, it would be really amazing to have.

AMD announces eight-core FX-9590

This week was not only a stage for numerous gaming innovations, including fresh hardware seen in Los Angeles for the first time the light of the world. AMD, for example, presents the five-gigahertz monster FX 9590. The most popular chip maker AMD is particularly proud of to their new FX-9590. Because the processor based on the Piledriver architecture is the first of up to five gigahertz clock frequency. This is the new record reached the eight-core but only in Turbo mode, the default clock is 4.7 GHz. Disadvantage of the immense computing power, the priority is to inspire enthusiasts is the up to 220 watts power consumption. The Intel competitor takes less than half of the power, and also their own predecessors are much more modest. Their second chip rookie is named FX-9370. It is clocked at 4.7 GHz, but this is also consume up to 220 watts. The extreme power consumption places less demands on motherboards. Home users are likely but hardly the itch, because the new processors from AMD targets exclusively to OEM partners like Acer and Hewlett

Haswell, Intel’s Fourth-Gen Core Architecture Core i7-4770K

Intel has introduced the fourth generation of the processor Core i-series "Haswell". Is it have everything is new? Not at all! Tested the Core i7-4770K "Haswell" has as its predecessor 3770K from the "Ivy Bridge" series four cores on board, working with exactly the same clock: 3.5 GHz. It can be utilize Turbo Boost technology, a single core with 3.9 GHz clock - such as a grueling task as you shrink a video for Smartphones. Haswell Intel processors manufactured with advanced 22-nanometer process. Advantage: In a small area, many circuits (transistors) can be accommodated. Haswell has a whopping 1.4 billion transistors on a very small area of only 177 square millimeters. The operating system pretends the 4770K as its predecessor by Hyper-Threading to eight cores.

Windows can allocate parallel programs running on multiple (virtual) cores. The cache (L3) is very generous with eight megabytes. A genuine innovation: Intel has incorporated the voltage converter into the Haswell CPU. In the previous models sat still on the motherboard. Advantage: for example, even in energy losses it can be contained and the processor is operating efficiently. Intel launches Haswell recently with a new processor generation. That's for normal desktop PCs, but also Laptops, Ultrabooks and tablet PCs Convertibles. Anyone who wants to upgrade his PC with a Haswell CPU needs a new main board: The 4770K has now 1150 pins at the bottom. The work rate of 4770K is only slightly higher than that of 3770K. For this, the Haswell CPU is playing 27 percent more nimble than the previous model. The testers were certainly not surprised that the Haswell Office and internet tasks completed only minimally faster than the older 3770K.

 The performance advantage laid with the Haswell processor only six percent. Instead of the lame HD4000 graphics chip, the newly developed HD4600 is used. Although it is good for games but still demanding games like "Battlefield 3", but enough for a smooth playback of titles such as "The Sims 3". In the test, it delivered almost 12 frames per second - about 27 percent more than the 3770K. New generation of processors, Is it less power consumption? No! The Haswell even approved in the test a little more juice than its predecessor 3770K. The cause: If the stronger graphics chip called the current processor, it needs more energy. In pure mode, the Office Haswell CPU is significantly more economical than the Ivy Bridge processor: instead of 38.30 watts, there were only 32.8 watts - nearly 20 percent less!