supercomputing

Do you remember your very first computer? Pretty much everyone does; most of the people in their mid-30s, early-40s can still remember these noisy big white boxes with huge square screens and clicking dial-up modems that took ages to download a plain-text news article or even a basic e-mail with no attachments. Well, it took almost 40 years for the technology to get to that point, and there are still alive today witnesses to how it all started in the basements of the world famous universities and colleges.

Joyce Wheeler is someone who saw it all in those early days. She also can still remember her very first computer, and one of the reasons for that is because it was one of the first computers anyone used.

Dr. Joyce Wheeler was among the pioneers of programming

It was EDSAC (Electronic Delay Storage Automatic Calculator), a “proto-computer” that was assembled and served scientists at the University of Cambridge back in 1949. Joyce Wheeler was a member of the scientists group who were working on their PhD degree under the supervision of famous astronomer Fred Hoyle. They were researching the reactions inside stars, in particular the star lifecycle stages and their length.

In order to perform the research, Joyce needed some powerful calculating equipment, since the inner workings of the nuclear furnace that keep stars shining is a very complicated problem to solve with use of a human brain, pencil and a piece of paper. Mathematics capable of describing this level of nuclear energy processes is pretty formidable: Joyce remembers that she had to solve a nasty set of differential equations that describe their behavior and composition.

A copy of Edsac is being built at the National Museum of Computing

Completing these calculations manually would almost certainly result in errors, inaccurate data and ultimately could – and probably would – affect the research outcomes. And here is where she met EDSAC – a machine built by Professor Maurice Wilkes, a technical device the size of several average size bedrooms, that was there to do the kind of calculations that Ms. Wheeler needed to be done to complete her advanced degree.

The first challenge for young astrophysics student was to learn the sophisticated language that machine could understand. She was quite familiar with the machine itself, since it was showcased to her prior the start of her degree course in 1954. Being keen to get her research done accurately, Joyce sat down with an instructions booklet and worked her way through dozens of the programming exercises from that pioneering programming manual. That little book was called WWG (after the names of the authors: Maurice Wilkes, David Wheeler and Stanley Gill).

The foundations of programming were laid down by Edsac's creators — The foundations of programming were laid down by Edsac’s creators

While learning the programming, Joyce (whose family name was Blacker at that time) got talking to David Wheeler, since one of her programs was helping to ensure that EDSAC was working well. They eventually got to know each other, fell in love and married in 1957.

Joyce remembers that exciting time in detail: she could not stop wondering what the machine could do for her work. She was able to study the programming quite fast due to her strong mathematical background: she became very quickly able to master the syntax into which she had to translate the endless complex equations.

At certain point of time she realized that programming is very similar to Maths in the sense that one can’t do it for too long.

“I found I could not work at a certain programming job for more than a certain number of hours per day,” Joyce Wheeler remembers. “After that you would not make much progress.”

Research students like Joyce Wheeler had to use Edsac at night

Sometimes the solution to some programming problems that worried her from time to time would come into her mind while she was doing some other things outside of the computer lab: like doing the laundry or having lunch.

“Sometimes it’s better to leave something alone, to pause, and that’s very true of programming.”

When the programming bit was finally done, Joyce Wheeler was allocated a timeslot to run her programs on the EDSAC: it was Friday night. She remembers that this period was perfect for her: there were no lectures the next day she had to attend.

As an operator she was granted the right to run the EDSAC alone, but she had to make sure that everything she did was recorded. A quite common occurence for all the early computers (and EDSAC was no exception) was unexpected crashes. Joyce remembers that only occasionally she was lucky enough to keep machine running all night, and if it did crash, there was little she was allowed to do to try to fix it. Even the cleaners were not allowed to get near EDSAC.

Dr. Wheeler showed Joyce one procedure, that allowed the recalibration of the EDSAC’s two kilobyte memory, but if that did not help, Joyce had no other choice but to stop her work for the night. But despite the regular crashes, she made steady progress on finding out how long different stars would last before they collapsed.

“I got some estimates of a star’s age, how long it was going to last,” she said. “One of the nice things was that with programming you could repeat it. Iterate. You could not do that with a hand calculation. We could add in sample numbers on programs and it could easily check them. I could check my results on the machine very rapidly, which was very useful.”

Now, you should understand that “rapidly” back in the 1950’s meant “not more than 30 minutes”. This is the time that EDSAC required to run a program. After that the results were printed out for the researcher to analyze them. After that you had to re-program and wait another couple of days to run another round of complex calculations. Despite all these delays, Ms. Wheeler felt that she was a part of something that would change the world.

“We were doing work that could not done in any other way,” she said. And even though EDSAC was crude and painfully slow by modern standards, she saw that a revolution had begun.

We at FTI never fail to get inspired by pioneering scientists like Dr. Wheeler, with their single-minded dedication and commitment to innovating new solutions to existing problems, often in lonely circumstances and running against the tide of conventional thinking, driving them to expand the frontiers of discovery and learning in ways that eventually become implemented into normal life for the entire global population. Innovation, research and patience are some of the core values we cherish at FTI, and no-one demonstrated these better than Dr Joyce Wheeler.

Technology has come a long way. We see constant development and growth in terms of computing hardware, software and storage. Although some technologists say, those gains are stalling, perhaps limited in by the physical boundary of raw materials that are used in central processing units.

Supercomputers are developing even faster than consumer-oriented computing technologies

Recently Microsoft thinks it may have found the solution namely field programmable gate arrays (FPGAs). A FPGA is an integrated circuit designed to be configured by a customer or a designer after manufacturing – thus it is said to be “field-programmable”. To define the behavior of the FPGA, the user provides:

a hardware description language (HDL)
or a schematic design.

The HDL form is more optimal for large structures as you can specify them numerically rather than having to draw every piece by hand. However, schematic entry can allow for easier visualization of a design. FPGA allows you flexibility in your designs and is a way to change how parts of a system work without introducing a large amount of cost and risk of delays into the design schedule. Many designers have the false impression that building a system with a modern FPGA means you have to deal with millions of logic gates and a massive amounts of connections just to do something useful. But if that were the case, FPGA use wouldn’t be growing: Instead, there would only be about a half dozen FPGA users left. It turns out FPGA designers have done much of the heavy lifting of adding commonly needed components so all that you have to concentrate on is customizing those functions that are specific to your application. Examples of components produced by designers comprise: clock generators, dynamic random access memory (DRAM) controllers and even whole multicore microprocessors.

This type of computer chip that can be reprogrammed for specific tasks after they leave the factory floor, are adding firepower to Microsoft’s network of on-demand computing power.

What does this mean in real terms?

Using all of the power of Microsoft’s data centers worldwide, the company could translate all 5 million articles on the English language Wikipedia in less than a tenth of a second.

In the past 2 years Microsoft has quietly been installing FPGAs on the new servers; Microsoft added to its global fleet of data centers. They’re present usage includes ranking results in the Bing search engine and speed the performance of Microsoft’s Azure cloud-computing network. Microsoft is alone among major cloud-computing players in widely deploying FPGA technology.

There are also implications for high performance computing and data storage such as solutions for Network Attached Storage (NAS), Storage Area Network (SAN), servers, and storage appliances.

Project Catapult is the technology behind Microsoft’s hyperscale acceleration fabric. The supercomputing substrate is built with the aim to accelerate the efforts in networking, security, cloud services and artificial intelligence.

Project Catapult combines an FPGA integrated into nearly every new Microsoft datacenter server. By exploiting the reconfigurable nature of FPGAs, at the server, the Catapult architecture delivers the efficiency and performance of custom hardware without the cost, complexity and risk of deploying fully customized ASICs into the datacenter. Moreover, the performance gain compared with CPUs is monumental and with less than 30% cost increase, and no more than 10% power increase.

Other FPGA Applications

Due to their programmable nature, FPGAs are an ideal for numerous markets. As the industry leader, Xilinx provides comprehensive solutions consisting of FPGA devices, advanced software, and configurable, ready-to-use IP cores for markets and applications such as:

Aerospace & Defense – Radiation-tolerant FPGAs along with intellectual property for image processing, waveform generation, and partial reconfiguration for SDRs.
Industrial – Xilinx FPGAs and targeted design platforms for Industrial, Scientific and Medical (ISM) enable higher degrees of flexibility, faster time-to-market, and lower overall non-recurring engineering costs (NRE) for a wide range of applications such as industrial imaging and surveillance, industrial automation, and medical imaging equipment.
Medical – For diagnostic, monitoring, and therapy applications, the Virtex FPGA and Spartan® FPGA families can be used to meet a range of processing, display, and I/O interface requirements.
Security – Xilinx offers solutions that meet the evolving needs of security applications, from access control to surveillance and safety systems.
Wireless Communications – RF, base band, connectivity, transport and networking solutions for wireless equipment, addressing standards such as WCDMA, HSDPA, WiMAX and others.

The advancement in computing power and storage capability combined with substantial savings and efficiency introduced through FPGA technology mean the world of supercomputing is more accessible then ever.

At the same time, the advancement of supercomputing technologies automatically means more data storage requirements. Every year world produces more and more data and therefore more and more information needs to be stored. Some information needs to be stored for more extended periods of time like 10-20-30 or even 100 years. There is no better long-term data storage mediums than archival discs.

Falcon Technologies International LLC has developed Century Archival product line, which is a cutting-edge technology with built in Gold or Platinum layers to ensure maximum security, longevity and protection for stored data.

Century Archival products are designed to secure data for hundreds of years – FTI’s Century Archival DVD product has been demonstrated in testing to be capable of storing data for up to 200 years, whereas the Century Archival CD product has demonstrated a longevity in excess of 400 years, making the Century Archival line the most durable and secure archival digital media product available in the market today.

Tag: supercomputing

At the Dawn of the Computer Age: Memories of the “Informational Revolution” Pioneers.

Powerful computing and storage implications