Affordable Cloud Platforms

The AMD journey began with only a few dozen employees in 1969. It began like any modest start-up – shoe-string budget, a bunch of enthusiasts on the table and minds at work. The initial years of AMD were not without struggles; it isn’t devoid of struggles even now. But from those modest beginnings, Advanced Micro Devices grew to a company of 10,000+ employees and began manufacturing processors for low-end computers and dedicated servers. Fast forward to today, AMD develops high-end computing and graphics products to resolve some of the toughest and unprecedented challenges in the semiconductor industry. That is to say, AMD next gen processors are worth a lot, especially for the price at which they are being marketed.

However, AMD had been left to lie in the second fiddles by its competitor Intel, which began operations only a year before AMD. Headquartered in Santa Clara, California, Intel is currently the largest and highest valued chip manufacturer when it comes to revenue. The organization had, since it was founded, crushed competition because it had had technological edge over all its competitors. In fact, Intel could be rightfully said to be driver of revolution in IT industry, gradually improving over the years. It would not be inappropriate if I’d say that a large portion of Intel’s success came from its monopoly.

Intel – The big fish

Intel launched its first microprocessor in 1971 and it was not until PCs became so successful that Intel made microprocessor its mainstream business. A big percentage of Intel’s earnings come from microprocessors, although the company has a wide range of products ranging from ICs, graphics chips (GPUs), embedded processors, flash memory and a ton of other devices in the field of communications and computing.

Intel was founded in the year 1968, by the then pioneers R. Noyce and G Moore, who formulated Moore’s law. Until 1981, Intel was manufacturing DRAM and SRAM memory chips and the revenue out of these two products fed the entire organization. These two chips continued to remain the major revenue streams until 1981, 10 years after Intel started manufacturing microprocessors.

Over the years, Intel Corporation came out with a number of breakthrough technologies and patents, including the x86 series of processors, the most successful microprocessor Intel ever built.

That is to say, Intel was the whole of computer chips market until AMD next gen processors rolled out.

Intel invested heavily in the development of microprocessor designs during the 1990s.  Not only did Intel create a new market that was till then non-existent, it also became fostered the rapid growth of computer technology. It was during this period that INTC started predominantly selling microprocessors and also became known for aggressive, anti-competitive tactics to defend its position in the market.

AMD – a new player or an old fish on steroids?

Till now, AMD was seen as a budget-friendly solution for basic-use PCs. Though the manufacturer produced some high end processors too, they were deemed as incapable of competing with the performance of their Intel counterparts. AMD only occupied around 10% of market share until 2018, while Intel the remaining. Despite being extremely popular in low-end computers, AMD could not make the cut and continued accruing debts over the years. The organization was on the verge to get wiped up during 2007 and the year following thereafter. To cut short, AMD was nowhere on the chart until AMD next gen processors were rolled out.

Source – statista

Key Takeaways – The Rivalry

• Advanced Micro Devices Inc. (AMD) and Intel Corp (INTC) have been competing for the last 50 years.
• For much of these 50 years, AMD was on back-foot and INTC has dictated the market ever since.
• A number of other companies attempted to break into the microprocessor market, none of which could equal what Intel and AMD had achieved.

The change of winds – Ryzen Series, AMD Next Gen processors

In February, 2017, AMD launched its next-gen Ryzen microprocessor, built with 7-nanometer architecture, quickly surpassing its competitor in terms of sales. In response to AMD’s Ryzen, Intel launched its Ice Lake microprocessor in 2019. Ice Lake is the code name for INTC’s 10^th gen processors, which it deems to surpass the Ryzen in terms of performance and power savings. What’s interesting is the 10^th gen processors have been priced lower than their 9^th gen counterparts. This highlights what a tough time INTC is having competing against the Ryzen series of processors in the market.

Source – Tech compare While Advanced Micro Devices (AMD) is now pushing out new chips based on 7nm (nanometer) architecture, Intel is still relying on age-old 14 nm chips, while trying to max out its current technology as much as it can. From a technical standpoint, 7nm chips allow for more processing power in the same size. The smaller architecture chips are therefore lighter, faster and more power-efficient than other chips. However, Intel has not been able to figure out technology to mass-produce 7nm chips while AMD has almost nailed it. Intel only recently introduced 10nm architecture but that is yet to be made available, while AMD next gen processors built with 7nm process are available in bulk. 

Despite lagging behind AMD, INTC is trying to push what’s left of its hardware.

We, at Go4hosting, decided to stack both the processors side by side and measure which of these two outperformed the other and under what circumstances. Before we do so, let understand a few things to help you have a better understanding of how different microprocessors architecture can get. 

Below, you will get an insight into the key things why AMD next gen processors stand out amongst its peer.

Core –

Core is the processing unit within a processor.

A few decades back CPUs comprised of a single core. The only way to expedite processing was increasing the clock speed (discussed later). Once the clock speeds hit plateau, engineers were set out on search of ways to accelerate performance of their processors. Initially, they thought of using multiple processors in the computers but doing so drastically increased the cost. It was then that incorporating multiple cores in a single processor was thought of as a viable option, which engineers achieved a few years down the line.

So, if a processor comprises 4 cores, it means it can process information at 4 times the rate of a single core CPU (provided other parameters are scaled likewise).

Threads –

Think of a thread as a conveyor belt. So a CPU with 4 core/8 threads has 2 threads assigned to each core. A thread divides a physical core into multiple virtual cores for faster operations

This could be better understood through smartphones. When you open an application a circle is seen spinning for a brief period. In the meantime the application and all the concerned files are processed and loaded in the background. This is achieved using special purpose thread created to serve this purpose only. While one of the threads displays a circle and processes data as if it were spinning, the other thread loads files and overrides the screen when the info has been processed.

Clock Speed –

Clock speed is an important factor in determining CPU performance. It is a measure how many cycles per second a processor can execute. Ideally, more the clock speed, more are the pulses and bits moved per second. Though clock speed can significantly increase performance of the computer, the number of instructions executed per cycle is also accounted for while calculating the overall performance.

AMD next gen processor not only rips higher clock speed but also higher core count. This has put Intel at a disadvantage against the ‘underdog’.

Readers should not confuse between base clock – and max clock speed.

1. The base clock speed is the number of cycles that a computer ordinarily executes per second.
2. The max clock speed is the number of cycles achievable using turbo boost.

When Giants Compete

Right after the development of its revolutionizing x86 chipset and IPO (1978), Intel became the dictator of the microprocessor industry. As of February 2020, Intel’s market capitalization was estimated at $288 billion, almost 5 times when compared to AMD’s comparatively-meagre market cap of $59.9 billion.

This is in fact understandable. AMD has been the underdog in the microprocessor market and focused on producing middle to low- range chipsets to lure those who could not afford Intel in the first place. This trend had more or less continued until the latest Ryzen CPU was rolled out – the only reason we all came here in the first place. Though, all this time, sophisticated users who like tinkering with their computers, did so using AMD’s unlocked CPUs.

Unlocked CPUs can be overclocked.

Even until 2016, AMD only possessed one-quarter of the entire CPU market, whereas Intel dominated almost the entire of the remaining three quarters. AMD turned tables of Intel, however, by introducing Ryzen series of its novel 7nm architecture. The Zen 2 architecture is all about leveraging the performance of tightly packed transistors in AMD next gen processors.

Intel vs AMD – the bout

Having understood the essential terms relating to microprocessors, we now proceed towards comparing products based on use-case.

1.Intel’s take on the matter

Image source – Intel

The above comparison was released by Intel itself but as computer experts we found the results were rather hard to grasp. First, the configuration of the workstation configured with Ryzen was not specified by Intel. Typically, tests should be performed by keeping all the other parameters comparable, if not equal. A high-end computer with slow processor could outrun low-end computer with advanced processor. The data could have become more credible had Intel released the system environment where these tests were performed.

2.The duel of Cores

Both Intel and AMD have reserved their high-core count for their enterprise line products, reserved especially for virtual machines and data center servers. Below we have compared enterprise-level CPUs from both the manufacturers.

Test results on Corona 1.3 benchmark.

Test results on Blender 2.7

Image source (Anand tech)

Intel’s Reply to AMD

In August last year (2019), Intel released its 10^th gen processor in response to AMD’s Ryzen. The new architecture was codenamed Ice Lake and Intel claimed an overall performance improvement of well over 18% over its predecessor. Intel still had the silver lining during the dark times as its high-end market continued to grow. High-end CPUs not only have a better profit-margin but also contribute more than half of the company’s overall sales. The way I see it, Intel’s chips are far behind AMD next gen processors.

Source – CPU talk

No more 10nm for Intel Now

Although Intel hurriedly launched its 10nm Ice Lake, there are no 10nm CPUs in its arsenal for the year. The demand vs supply curve for Intel has been somewhat unprecedented, with supply falling short all over the globe. In all desperation, Intel is trying to prove that it still holds the throne but AMD next gen processors are miles ahead of Intel currently.. 

The two giants will continue dueling in the future until one comes out victorious over the other, which, despite all the likelihood, is almost unlikely to happen.

Selecting the top data center services in India is not an easy task because you will find many data center providers in the country. However, it is necessary to understand your own data center requirement before signing up with any provider. This is because a data center service is an ideal solution for smaller businesses which find it hard to maintain a private data center; for larger corporations, maintaining on-site data centers is not an issue. So, while choosing data center services in India, you should carefully consider many important factors:

• To start with, location of the data center provider is of utmost importance when choosing a colocation provider. For instance, the data center services in India that you choose should not be located at too great a distance from your business. When the distance is too great; you will end up paying huge transportation costs for sending manpower to the site to carry out repairs and upgrades.
   
• Moreover, the data center should not be located at a place which is known to be prone to natural calamities such as floods, earthquakes, tornadoes etc. These threats may cripple the infrastructure and functions of the data center making it inoperable.
   
• Since the key to data center functioning is communications, it is important that the company has excellent fiber connectivity. So, the data center location must have multiple carriers to guarantee redundant connections that will ensure uninterrupted power supplies.
   
• Data center electricity expenses have escalated drastically in recent years and this means that you must choose data center services in India that can provide high-density power.
   
• The data centers must also be located in sites which are easily accessible. So, IT suppliers and hardware manufacturers should be able to supply men and materials to the facilities without delays. Air connectivity is desirable to ensure that support staffs can be sent faster to deal with outages.
   
• The data center services in India that you have shortlisted should be able to offer you with scalable hosting solutions. In other words, they should be able to offer you additional space, power and cooling facilities if required. Different providers will offer different degrees of flexibility and scalability; while some can offer innovative solutions, others will only provide the standard solutions. It is therefore necessary to select a provider which lets you scale up and down your resources depending on your demands.
   
• When choosing data center services in India, you need to consider how reliable they are. Reliability is evaluated in terms of their uptime guarantee. You need to review their Service Level Agreements to be sure of this guarantee; you must also check for customer feedbacks, staff certifications and on-site support systems.
   
• When choosing a provider, it is also necessary to select one which can deploy the resources quickly. Businesses are always in a hurry to set up the new space and networking equipments; so, deployment efficiency is a rather significant criterion when choosing a data center provider.
   
• Finally, you will not want to choose data center India services that are new in the industry and may shut down their operations after a few years. You should ideally check for the company’s financial stability by reading through financial reports and press releases to understand how viable it is.

The multi-pronged security effects and rising maintenance prices related to data centers and IT infrastructure has brought up a reverse trend within the IT market. Due to tight budgets and lesser IT resources being available, the majority of organizations are moving towards data center consolidation. This ever-changing culture is working towards reduction of the infrastructure for having smarter and energy economical IT infrastructures. As a result, we can see a growing range of facility mergers, and virtualization services of organizations.

The data center consolidation trend is certainly effecting the IT market. Together with the wide scope of the IoT technology, information Infrastructure Management particularly has long term effect on data center. The data networks are consolidated into compact areas to shrink the operational cost involved.

The design of infrastructure can help in building data centers that are easier to manage, removing the burden on the working of IT. However, this shift will come with lot of challenges. Based on the quality of legacy systems, typically with one infrastructure leg on website and the alternative in external network territories, unexpected complications rise up during consolidation.

Go4hosting makes sure that Data Center Migration consolidation is positively changing the way we store data on the internet, and aims to offer advanced service quality and features. Go4hosting provides an advanced and high network protection experience for your data.

A More Efficient Web Of Data

Cost reduction is usually a powerful force behind deep infrastructure changes. It pulls huge amounts of servers into a smaller structure by permitting the organizations to lower their operational prices and also measure their IT issues. Having a compact network and application design leads the business to have more cost effective hardware, servers and routers. They can accomplish deploying servers effectively by adding linked technologies, server and storage virtualization by implementing the cloud computing, and higher capability designs for future scaling up.

By reducing the number of servers, computer code standardization is done effectively, which implies that fewer dynamic applications are required to run the servers effectively. The implementation of cloud-based Software-as-a-Service (SaaS) facilitates additional reduction in the working prices. By use of lesser hardware, there is reduced need of power which, besides shrinking energy bills, helps in approaching business expansion properly.

Once enforced, data center consolidation will facilitate alleviation of the tasks that create burdens on IT teams and operations staff. Fewer channels in efficient designs makes an optimized data transport system. It also gives maximized data path utilization, and gives clear data traffic pattern routes. All this amounts to higher security and higher compliance management.

New Horizons

The Data center reconstruction can result in price reduction; it also provides increased security up and reduction of power consumption. This helps in building up an extremely great foundation from where the interactions with partner businesses and clients can be done effectively.

Cloud migration will continue to be among the biggest data center industry growth trends in 2017. More data providers focus on having added interconnectivity in order to expand the public cloud. Expanding the data for working of cloud effectively makes more sense than using complex sites for building data centers.

Ultimately, information consolidation has showed a cut back on prices, increasing the business of IT firms. That is the way companies overcome the challenge of their stretched IT teams and tight budgets to successfully handle the transition to data centers.

It pays to understand the value of data center tiers in the context of enterprise IT infrastructure. Knowledge about data center tiers can enhance the overall understanding of what to expect from a data center resource.

Overview of DC Tiers

A data center tier is simply a nomenclature used to rank different data centers by categorizing their uptime performance. There are four tiers of data centers as established by Uptime Institute so that IT professionals are able to know the standards of availability and uptime.

Tier I Data Center– Tier I Data Center provides the fundamental and essential support to your IT infrastructure by shifting the same from your in-house facility to a dedicated building called as a data center. 

The attributes of data centers belonging to the Tier I category are basic yet far more superior to the in-house IT facilities of majority of enterprises. Your IT infrastructure including servers can be positioned in specially assigned racks inside a dedicated building of data center. It is supplied with seamless power with UPS backup facilities for enhanced availability and uptime. 

Climate controlled environment is provided to maintain server and other networking equipment in good running condition. The cooling equipment is also backed by incessant power supply. The overall arrangement in the Tier I data center is aimed at greater efficiency and performance of websites with enhanced availability in spite of traffic spikes or unexpected surge in resource requirements. In the event of unexpected power failure, the data centers in Tier I category are equipped with power generators for assured availability of electrical supply to keep your website up and running.

Redundancy in Tier II Data Centers

In addition to the facilities of Tier I data centers, a Tier II data center offers extra redundancy in terms of cooling and power resources. Although these systems are not absolutely redundant, these can certainly surpass the provisions of Tier I data centers. There is a greater assurance of uninterrupted cooling and power arrangements in Tier II Data Centers.  

Tier III DC- Concurrent maintenance and superior connectivity 

Needless to mention, a Tier III Data Center is equipped to provide all features of Tier I as well as Tier II Data Centers. It is designed to facilitate concurrent maintenance of equipment without causing any disruptions as far as the availability and connectivity is concerned. 

There is a dual power availability to keep all IT components running without any hassles and the vital systems such as networking and cooling are supplied with dual UPS systems to facilitate maintenance of one of the UPS units. However, since the Tier III Data Centers lack independent feeds of power and cooling systems from external sources, these are not entirely fault tolerant. 

Fault tolerant Tier IV Data Centers

Combine all the attributes of Tier I to Tier III data centers and add 2N fault tolerance capacity to get a Tier IV Data Center for your mission critical web ventures. The 2N Fault Tolerance implies that each and every IT component receives dual power and cooling supplies. Independent cooling and data paths ensure that the IT processing cannot be affected even if any component of cooling or power supply fails to perform.  

Significance of uptime values in DC Tiers

Although, all Data Centers belonging to different tiers assure above 99 percent uptime annually, there can be considerable variance in terms of total annual downtime as you move up from Tier I to Tier IV data centers. 

This is mainly due to the surplus redundancy provided by top Tiered data centers. This means a Tier I data center can experience a total annual downtime in excess of 29 hours while a Tier IV data center can be guaranteed to perform for 99.995 percent of the time. 

There is a strong association between the nature of your business and selection of a data center based on different Tier categories. So a smaller enterprise can be content with a Tier I data center while large and complex ecommerce or government websites should be provided with facilities of Tier IV Data Centers. 

If you are only a beginner and have limited resources to spare for procurement of a Data Center resource, then you need not go for top tiered data centers. In contrast, if your website is expected to drive business critical applications that need to be incessantly available, then you must choose either a Tier III or a Tier IV data center environment. 

For large enterprises, the option of developing an onsite infrastructure of a Tier IV data center may sound attractive. However, there are a large number of financial, geographical, and technical factors to be considered before choosing between an outsourced and in-house data center. The Data Center standards set by Uptime Institute help enterprise clients make the right choice by assessing their website requirements. 

To some, the thought of visiting data center can be intimidating at first. One such company, Go4hosting, owned by India’s fastest emerging IT firm ‘Cyfuture’, has one such data center here at Noida.

Now, we had heard of how secure these ‘data fortresses’ are, but never had the chance to visit any of it until now.

The First Look –

The facility is situated at the heart of Noida Special Economic Zone (NSEZ) [refer map above]. It is from this place the company delivers its cloud hosting service and almost every other service, even though it has some more data centers around India. Entry to the zone itself is restricted, with walls ten to twelve feet high running along the entire perimeter.

NSEZ Entrance

Right at the entry, the customs flagged us because our car did not have the gate pass. After a round of questions that lasted couple minutes (about our purpose at NSEZ), we were given the green signal. The camera at the façade had registered our license number. Had things went bad that day in NSEZ, I would not have been surprised to see Police banging on my door.

We skipped a roundabout and took right from the next one. Hardly have we travelled 150 meters when another high walls, this time covered with barbed wire, came into view. We had reached the data center.

There was no other way into the facility except for the main gate, which was so heavily guarded it seemed like military installation than anything else. Had we not had the authority to enter, we would rather had rummaged the car into the wall than come upfront with the guards at the gate.

It was not a cakewalk getting into such high security building.

Threats Encumbering Data Centers In India

Data center market in the nation is expected to reach 1.5 billion USD by the end of 2022. Apparently, tech-giants like Google, Amazon and Microsoft also realized this and have been investing large sums every quarter. But it is not only the investment that has witnessed a rise.

Reports indicate that of all the countries, data theft has hit India the most.

Of all the industries affected, healthcare has had the maximum attacks and education has had the least.

Is Physical Security Sufficient Alone?

Physical security alone is not sufficient to check data thefts. A data center is exposed to both physical and network threats and from practical standpoint, a network attack is more likely because: a) it is comparatively easier than actually stealing servers from data centers, and, b) it can be pulled off more covertly.

Go4hosting estimates, of all attacks 96% are carried online and result in more loss than physical theft.

Tiered-system of Data Centers

Depending upon how sophisticated the systems are, data centers are classified into tiers [I, II, III, and IV] – tier IV being the most advanced. However most data centers operating in India are still tier III (even go4hosting’s) and the system is almost comparable to its elder brother (Tier IV) except that the latter has few more upgrades.

Learn more about data center tiers in the video.

Application Security

A holistic approach is needed to bring down the risk of data loss.

“The premise is monitored 24/7, each corner is under CCTV surveillance. The footage is not touched for months until disposed. Clients have the rights to claim access to footings – this ensures nobody is tempering with their data, not even us”, one of the insiders said.

 When asked about how go4hosting’s data centers are so secure, he did not divulge much details though.

Firewall alone is not sufficient to safekeep data packed inside the servers. A data center will not only have firewall, but also antivirus and anti DDoS services running concurrently.

Environment Control

Data centers operating in optimal conditions can function almost 2-3% more efficiently than those operating in adverse climate.

Temperature	68 to 71 degree Fahrenheit
Humidity	Dew Point Range – 41.9 to 59 degree Fahrenheit.
Relative humidity	Not more than 60%
Static Electricity Monitoring	Round the clock
PUE (Power Usage Efficiency)	Most ideal (theoretically) – 1Anything above 1 is consider inefficient
Configuration	Hot aisle/cold aisle cabinets

Server racks are not placed at random, at least not in a modern data center – aim being maintaining a uniform temperature throughout.

The back of a server is where the turbo fan is situated, the way it is in a refrigerator. Air close to the rear is therefore hotter when compared to the front. This difference in temperature creates an internal air current and subsequently hot air diffuses to the cooler regions.

If hot air could be contained in its space and stunted from moving towards the cooler regions, power used in running chillers would reduce and, in turn, help save cost. (Refer figure above)

Challenges in Running Data Center in NCR

India is one of the most polluted countries in the world – notice India has only Red and no green flags in the WHO air-quality map.

Particulate matter, Sulphur and nitrogen in the air impose a big problem to data centers in India. The air quality in Delhi is even worse. The city happens to be the most polluted capital in the world. The amount of toxins present in air make hardware corrode faster and fail before its expected life.