Fusion-io Flaunts New ioMemory Module
Thursday, August 19, 2010 - by Paul Lilly
Fusion-io has been blazing its own trail ever since the company was founded in 2006, tweaking technology to fit the kind of high speed products the company likes to release. Having seemingly hit a ceiling in PCI Express-based SSDs, Fusion-io has now gone and developed a new ioMemory module that will set the foundation for Fusion-io’s next crop of products.

The new ioMemory boasts support for the latest generation of MLC NAND Flash and 3X-nanometer (nm), which ultimately will allow Fusion-io to double up the capacity of its product lineup and squeeze up to 1.28TB of capacity onto a single PCI Express card.

The performance potential here is off the charts, with Fusion promising up to 285,000 sustained IOPS with under 25 microseconds commit latency. That’s crazy fast, but the new ioMemory module brings more than just speed to the table. Fusion says products built with the new chips will guarantee data integrity in the event of power loss, and the company’s self-healing Flashback protection offers Fusion’s proprietary, RAID-like chip-level redundancy.

Of course, this sort of thing is aimed at the enterprise crowd, both in performance and price - expect to pay four figures for the 1.28TB version.

HPC PROJECTS: OIL AND GAS
Super seismic

Data from seismic surveys must be processed rapidly in order to open up new oil and gas wells at the same rate as existing ones are depletedGPUs are established in both the visualisation and the processing of seismic data. Stephen Mounsey looks at how the technology is used, and at alternative hardware types available to geophysical analysts in the oil and gas industry

HPC Projects: June/July 2010

Drilling for oil is a tough business, and it’s only going to get tougher as global reserves begin to dwindle. Only a fraction of a per cent of the planet’s crust contains oil and, for any given reserve, only 10 to 60 per cent will be recoverable. It’s important for the corporate giants producing oil to open new wells to replace those that run dry, and so promising areas of land and sea are constantly being probed in order to locate new supplies.

Wherever the process occurs, be it on land or at sea, prospecting for oil and gas reserves begins with a seismic survey of a promising area. Low frequency sound waves are produced by an underwater spark or a compressed air gun, and as they travel through the Earth’s crust, they reflect off of areas in which the properties of the rocks change. Seismic surveyors use geophones or hydrophones to listen for these echoes, often collecting petabytes of raw acoustic data from a single seismic survey. In order for geophysical analysts to be able to pinpoint the locations of potential oil reserves, this data must be extensively processed into something they can visualise.

Laurent Billy, CEO of Visualisation Sciences Group (VSG), explains the steps necessary in producing a useful visualisation: ‘First the analysts start with the raw data from sensors – this needs to be interpreted in order to produce what is called the post-stack data – a seismic volume that can be analysed further. From this seismic volume, they identify horizons and geobodies (faults and cracks in the earth), and from here they can identify reservoirs in the areas where rocks are permeable enough to contain oil and gas.’ Compute-intensive reverse time migration algorithms are the technique of choice for turning seismic data into a seismic volume (3D map), and further calculations carried out on parameters within that volume are used to calculate the seismic attributes of the rock at various points in space. ‘The horizons in particular [areas where seismic properties change] are very important to visualise, and so they are modelled with a very large number of polygons,’ explains Billy. ‘This 3D data is therefore very heavy, and difficult to handle in the memory, and it can be very slow to display.’

VSG supplies computing components to facilitate the visualisation of these heavy datasets. ‘Most of our customers in the oil and gas industry are software vendors currently creating dedicated software for specialised tasks related to the exploration and production fields,’ says Billy. ‘Instead of creating and maintaining all of the visualisation capabilities of their software by themselves, our customers use our library components in order to deliver the best visualisation capabilities within their software. The end user – the geophysicist or whatever – is using our product indirectly, via the vendor who has embedded our technology within his own product.’

For the end user, powerful graphics cards (GPUs) are a prerequisite of this visualisation process, and VSG has worked to allow these GPUs to be harnessed for use in other tasks that an analyst may wish to perform on his data set. ‘Our toolkit is designed to be able to handle very large models, and the management of these large models can be done on the visualisation side [in the GPU] as well as the computation side [in a cluster, for example]. This means that we have created an architecture that enables a software developer to perform very complex calculations on the GPU while he is displaying the data. The technology is more than just a viewer – it handles very large models that do not fit in the memory, and it allows developers to use this facility to run code on the visualisation models. This is very useful for calculating seismic attributes on the fly – information contained within the seismic data. We are not supplying high-performance computing tools, but rather we are supplying high-performance visualisation tools that take advantage of the computational capabilities of GPUs, and which also provide developers with a framework and toolset to develop HPC code linked to visualisation.’

Close ties
In practice, VSG works very closely with GPU developer Nvidia, as Billy explains: ‘We have a partnership with Nvidia with regards to their Cuda technology, but we have built what we call an abstraction layer into our product. The developer does not specify what he wants to use Cuda for; rather the library chooses the best way to perform the calculation, be that in real time or using more GPU or more CPU,’ he says. VSG is able to ensure that its products make full use of even the latest incarnations of Nvidia’s hardware and Cuda platform by virtue of the fact that it is a beta tester for the devices: ‘We get the latest generation before they hit the market, so that we can test them and report any suggestions and wishes we have back to Nvidia,’ says Billy.

As well as having been early adopters of the Cuda GPU programming environment, VSG is an early user of Nvidia’s CompleX application acceleration engine – a scalability tool that allows a large 3D volume to be split into a number of smaller parts to be distributed across however many GPU boards a user has attached to his or her system. CompleX is linked to Nvidia’s Quadro Plex products, which are high performance visualisation-oriented single workstations containing two, four, or eight GPU cards. ‘Thanks to the CompleX technology, the display can be four, eight, or 16 times faster,’ explains Billy. ‘It allows the calculations on a very large volume to be distributed [across several GPUs], essentially in real time. We are able to increase the frame-rate and the display speed by increasing the number of GPUs.’

Beyond display
A GPU (or several) is an essential piece of equipment for the visualisation side of the process, but the GPU may not always be a good choice when it comes to seismic processing. SRC Computing, based in Colorado Springs, USA, develops computing systems based on reconfigurable processors – accelerator boards consisting of arrays of FPGAs. ‘There’s been something of a love affair between the oil and gas industry and GPUs, and this is because for a number of years the holy grail of the industry has been visualising large chunks of the data – that need was delivered by graphics cards,’ says Mark Tellez, director of business development at SRC. Tellez believes that the move from visualisation applications to GPU-led processing was something of a natural progression, as many client companies have already invested in the necessary hardware. ‘In a lot of ways, GPUs were just a product looking for a solution that they can solve, but [GPU suppliers] were not building a compute solution, they were building a graphics card designed to solve a graphics problem,’ he says.

SRC supplies customers in the oil and gas industry with a way of speeding up the demanding reverse time migration algorithms used in processing seismic data. David Caliga, director of software applications at SRC, describes the company’s approach: ‘We provide what a lot of people might call accelerators, capable of speeding up compute-intensive portions of code. What we provide is a complete system, and not just a GPU- or FPGA-based accelerator card.’ The company’s products are based on highly parallel and reconfigurable FPGA accelerators, which it refers to as MAPs. Caliga states that the company optimises its systems to ensure that seismic data can be very rapidly moved into the MAPs. Several MAPs may be incorporated into a single system, working alongside the standard CPU microprocessor. ‘We treat the MAP processor as a peer to the microprocessor. The intent is to divide the compute intensive application across the two compute devices to get the most out of both of them. In one image processing application, for example, a system with five MAPs and one microprocessor replaced a cluster with 96 dual-core nodes.’

The performance of MAPs-based systems is described by the company as comparable to that of GPU-based systems while only requiring around a quarter of the power of an equivalent GPU system. ‘We can provide superior performance with a fraction of the power dissipation,’ states Caliga. ‘Fully loaded, the MAP consumes approximately 50W, compared to a GPU, which would be around 200W.’ A reduction in power consumption and heat dissipation also allows a reduction in the footprint of the system: ‘The potential reduction is from something that would consume multiple racks in a data centre down to something that would fill a small desk-side enclosure, or maybe a standard rack.’

Tellez explains that even in a multi-billion dollar industry, power consumption is of key importance: ‘I know of a number of [seismic data processing] companies that are constrained by the amount of power they can pull from the grid at the location of their data centre. Additionally, even if they can pull enough power for their server farms, they’re typically working in a building that was never designed to have that much power generated on site (in terms of heat). They therefore have trouble in terms of cooling.’ He adds that many companies are now looking towards moving data processing operations nearer to the drilling location, on a ship, or even on the drilling platform itself, where power and space are even more constrained. ‘Typically, they put the data onto a number of hard drives and fly it by helicopter back to the mainland, where it is loaded into a computer. They compute whatever they are working on, and then they often have to send the results back in the same manner! If they could do all of the processing on site, they’d be saving a lot of time and money, not only in terms of just moving the data, but also in terms of avoiding idle time on the drilling platform while they wait for results.’ SRC believes therefore that its low-power processors offer the oil and gas industry an attractive alternative to both conventional clusters and GPU-based computation.

Tellez states that SRC has gone after customers in the industry of seismic data processing because importance of processing in the industry: ‘They are the low-hanging fruit at the moment, because the faster they can process a line of code, the faster they can either begin drilling or charge the drilling company for the information.’

Given the advantages of FPGA-based processing over cluster or GPU-based alternatives, why haven’t more oil and gas companies adopted the company’s solutions? Tellez believes that FPGAs (or reconfigurable processors) are seen as complicated and difficult to program. ‘The biggest challenge we have is to get people to realise what they can do with the current technology, and to realise that it’s not as scary as they may have heard.’ In a move analogous to Nvidia’s introduction of the Cuda programming environment for GPUs, SRC offers its CARTE platform for programming FPGAs. CARTE contains a set of standardised functions specific to the oil and gas industry, and the company hopes that this will enable more and more players in the industry to take advantage of the reconfigurable technology. ‘We’ve done a lot of work in order to simplify the move from the original calculation into the reconfigurable environment,’ says Tellez.

Before the introduction of GPU programming environments such as Cuda and OpenCL, GPUs were very difficult to program. As programming tools for reconfigurable environments become more established, programming the right tool for the job can only become easier.

GPU
Used primarily for 3-D applications, a graphics processing unit is a single-chip processor that creates lighting effects and transforms objects every time a 3D scene is redrawn. These are mathematically-intensive tasks, which otherwise, would put quite a strain on the CPU. Lifting this burden from the CPU frees up cycles that can be used for other jobs.

The first company to develop the GPU is NVIDIA Inc. Its GeForce 256 GPU is capable of billions of calculations per second, can process a minimum of 10 million polygons per second, and has over 22 million transistors, compared to the 9 million found on the Pentium III. Its workstation version called the Quadro, designed for CAD applications, can process over 200 billion operations a second and deliver up to 17 million triangles per second.

I’ve collected a few YouTube videos of our partner company Fusion IO.   The combination of FLASH Memory cards in your servers will dramatically improve your systems throughput.   Now couple that with even more Fusion IO cards installed in the NEXTIO expansion chassis…. Greater performance across more of your servers.

Be sure to read the White Paper about how the making of the Clash of the Titans 3D movie was reduced to 8 weeks rather than months with the use of the FLASH Memory cards.

For Oil & Gas Applications the advantage of FLASH memory installed in NEXTIO chassis will significantly improve your system performance by a factor of 10X.

Medical Imaging can also take advantage of this type of technology. For example, PACS systems to read images 10x faster means more scans, and more patient cases. NextIO can help harness the use of GPU and FLASH Memory in the same chassis resulting in 10X faster throughput.

Advantages of FLASH Memory for Movie Rendering

Dramatically Faster Video Rendering using Fusion IO

2000 Streaming Videos playing from ONE FLASH Memory card

Streaming Multiple Video sources

Case Study White Paper — Clash of the Titans 3D in 8 Weeks

Two of the world’s eight greenest supercomputers combine CPUs and GPUs to boost performance
Supercomputers that mix CPUs with graphics processors made their mark on the Green500 list of top energy-efficient supercomputers released on Wednesday.

Eight of the world’s greenest supercomputers combined specialized accelerators like GPUs with CPUs to boost performance and make supercomputers more power efficient, according to the Green500 list, which is released twice a year. The list was released by the same group that compiles the Top500 list.

Supercomputers with accelerators are three times more energy efficient than their non-accelerated counterparts on the list, according to Wu Feng, associate professor of electrical and computer engineering at Virginia Polytechnic Institute and State University’s college of engineering.

Two of the top eight green supercomputers are new entrants from China, and combine graphics processors from Nvidia with Intel’s CPUs. In the previous list issued in November, only one supercomputer combined CPUs with GPUs from Advanced Micro Devices, but that machine has now dropped to the 11th spot.

The Green500 list is compiled to “ensure that supercomputers only simulate climate change and not create it,” according to the Green500 Web site.

The list rates the greenest supercomputers by measuring performance in relation to power consumed. The calculation takes the megaflops-per-second performance (MFLOP/s) of a supercomputer and divides it by per watt of energy consumed. Supercomputers with accelerators averaged 554 MFLOP/s per watt, while other measured supercomputers without accelerators produced 181 MFLOP/s per watt.

The supercomputers combining GPUs with CPUs include the Dawning Nebulae supercomputer, which was in the fourth spot, and the Mole-8.5 supercomputer, which took the eighth spot. The supercomputers are in China and combine Nvidia’s Tesla C2050 graphics processors with Intel’s Xeon X5650 quad-core processor, which runs at 2.66GHz. The Nebulae supercomputer achieved efficiency of around 492.64 MFLOP/s per watt, while the Mole-8.5 achieved efficiency of 431.88 MFLOP/s per watt.

The top three green supercomputers were IBM supercomputers, all in Germany. The supercomputers include PowerXCell 8i processors from IBM, with custom field-programmable gate array accelerators to boost application performance. The supercomputers were also the top three in the previous list issued in November.

Overall, IBM chips were used in six of the top eight green supercomputers.

There is growing interest in building supercomputers that use graphics processors along with CPUs. GPUs are typically faster than traditional CPUs at executing certain tasks, such as those used in scientific and computing applications. Some institutions like the Tokyo Institute of Technology have announced plans to deploy more GPUs in an effort to squeeze more performance out of servers.

GPU ReAssignment from NextIO

It’s difficult to say if or when Fusion-io’s newest development will directly affect the average consumer, but as with many things in the technology field, what starts at the highest levels of enterprise eventually filters down to the consumer once kinks have been worked out, prices have adjusted downward and more partner companies have had time to adopt the new process. We are guessing that’s exactly the path that ioMemory will take, which is a new flash-optimized OS subsystem.

ioMemory VSL (Virtual Storage Layer) is definitely an enterprise application for now. It has little direct connection to the parts within your existing notebook or desktop, but the ideas presented here could definitely affect your notebooks and desktops of the future. For starters, not too many consumers even own a Fusion-io product. The company mostly sells high-end, extraordinarily expensive flash-based SSDs to enterprise and corporate customers running servers and workstations that simply cannot wait for sluggish, oftentimes unreliable hard drives to spin up information. But we have already seen that Fusion-io is testing the consumer waters, and as NAND prices continue to settle and SSDs continue to become more mainstream, they’ll most likely continue to venture out into these uncharted territories.

According to the company, ioMemory VSL is the “first and only OS subsystem that combines the benefits of the traditional I/O sub-system (block-level reading and writing) with the benefits of the virtual memory subsystem – virtualizing ioMemory devices and offering a ‘fusion’ of both memory and storage.” Fusion-io makes no bones about the use here, though, stating that ioMemory VSL is unique to the enterprise flash industry. Basically it “approaches flash as an extension of the memory hierarchy and as a new building block for computer hardware and software architecture, rather than confining it only to traditional storage paradigms.” That’s pretty deep on the “technobabble” scale, but the gist of it is this: the new fused architecture can provide near-linear performance scaling with very little software/hardware overhead , and it virtualizes Fusion’s ioMemory technology, presenting it not just as traditional block storage, but also as a virtualized storage/memory hybrid with a much richer set of interfaces.

The best news is that existing software such as file systems, volume managers, and applications are able to access ioMemory without modification, but for those seeking more advanced use, the tech can be tweaked by the end-user in order to bring out an advanced set of enhanced programmatic interfaces that seek to “further exploit ioMemory to improve throughput, response times, and reliability features.”

Today, the company is simply introducing the technology, and judging by the quotes from select users down in the full release below, at least a couple of larger customers have found it quite useful. That said, the potential of this new architecture remains untapped. The good news here, though, is that it shouldn’t remain untapped for long. The new software (and possibly firmware; details on the upgrade are scant) will work on all existing and future Fusion-io products, giving each one the ability to treat flash-based storage as a new memory tier instead of a super-fast hard drive. Imagine all of the things you need RAM for, then think of how sweet it would be if you could slap 128GB of the stuff into a spare PCIe slot. We aren’t positive that this will end up working out precisely like that when it hits the mainstream, but given just how speedy all of Fusion-io’s products are, we wouldn’t put it past them to re-write the story of system memory in one fell swoop.

Fusion-io Continues Its Innovation Leadership by Introducing First Flash-Optimized Subsystem

The Company’s New ioMemory VSL (Virtual Storage Layer) Subsystem Offers a Unique Fusion of Today’s I/O and Virtual Memory Subsystems, Further Distancing The Industry Leader from Competition

SALT LAKE CITY– Fusion-io, pioneer of a new memory tier of flash-based solid-state (ioMemory) technology, today announced the release of a new flash-optimized OS subsystem called the ioMemory VSL (Virtual Storage Layer). The ioMemory VSL is the first and only OS subsystem that combines the benefits of the traditional I/O sub-system (block-level reading and writing) with the benefits of the virtual memory subsystem – virtualizing ioMemory devices and offering a “fusion” of both memory and storage.

The ioMemory VSL is unique to the enterprise flash industry. It approaches flash as an extension of the memory hierarchy and as a new building block for computer hardware and software architecture, rather than confining it only to traditional storage paradigms. The result is an elegant cut-through architecture that provides near-linear performance scaling with very little software/hardware overhead, unprecedented flash reliability and endurance, customer flexibility in formatting, software development opportunities, and future-proof field upgradeability.

The ioMemory VSL virtualizes Fusion’s ioMemory technology, presenting it not just as traditional block storage, but also as a virtualized storage/memory hybrid with a much richer set of interfaces. Existing software such as file systems, volume managers, and applications are able to access ioMemory without modification. But, with an advanced set of enhanced programmatic interfaces, applications can be adapted to further exploit ioMemory to improve throughput, response times, and reliability features. The ioMemory VSL is a springboard to an entirely new flash-optimized software ecosystem that has already begun to emerge with Fusion’s OEM partners, third-party solution developers, and research groups.

“Faster hardware features are often cited as the primary way that improvements in application performance are achieved. But, as Fusion-io shows, it certainly is not the only way,” said Jerome Wendt, DCIG Lead Analyst and President. “The ioMemory VSL enables organizations to achieve significant improvements in performance, reliability, and endurance. With the tens of thousands of devices that Fusion-io has in production in the marketplace today, customers can realize this benefit at no additional cost to Fusion-io customers.”

“Fusion‘s ioMemory VSL allowed us to develop a flash optimized file system called DFS that outperforms ext3 by 20 percent for direct access and over 149 percent for buffered access. It delivered up to 250 percent higher application performance in benchmarks while reducing CPU overhead,” said William Josephson, a PhD student at Princeton University working under Kai Li, founder of Data Domain. “Additionally, we used the advanced features of the ioMemory VSL to handle difficult operations such as block allocations, de-allocations and crash recovery safety. This allowed us to develop in just six months what typically takes three years.”

“With the ioMemory VSL, information processing systems can improve performance, reliability and features set by taking full advantage of the unique capabilities flash has to offer,” said Neil Carson, CTO of Fusion-io. “Our approach to flash integration creates a new tier of virtualized memory, not just a legacy block storage device. This approach results in a more efficient, balanced infrastructure that is often capable of ten times the workload–enabling customers to do more with less.”

NextIO vSTOR™ Application Acceleration Appliance and Fusion’s ioDrive Duos Set a New Flash-based Storage Appliance Performance Record

AUSTIN, Texas–(BUSINESS WIRE)–NextIO, the premier provider of next-generation I/O consolidation solutions, and Fusion-io, pioneer of a new memory tier of flash-based solid-state (ioMemory) technology, today announced that they have broken the record for IOPS performance in a single rack-mount flash array.

“Fusion-io is excited to work with NextIO in the creation of their new vSTOR platform, which is setting new records for performance and adaptability”
A single NextIO vSTOR™ Application Acceleration Appliance populated with six ioDrive Duo ioMemory cards and two SuperMicro® servers equipped with Six-Core AMD Opteron™ processors reached performance levels of 2.2M read input/output operations per second (IOPS), easily besting the previous single-box flash performance record of 1.7M IOPs.

vSTOR’s innovative implementation makes multiple high-performance PCI-Express connections available to servers, each of which provides significantly higher host connection bandwidth than flash arrays using Fibre Channel or Serial Attached SCSI (SAS) host interfaces. Its ability to use widely available, industry-standard flash media ensures best-of-breed performance and features, while eliminating the proprietary media lock-in that characterizes other competing flash arrays.

To help NextIO meet this goal, they have entered into an Original Equipment Manufacturer (OEM) agreement with Fusion-io. The agreement supports a new Fusion-io initiative to make the incredible performance capabilities associated with its ioMemory technology available to a wide variety of vertical markets. By adding Fusion-io to the NextIO platform, NextIO will be able to create a solution that provides customers with seven terabytes of high-performance flash in a 3U/300W appliance, and does so at a lower price per gigabyte (GB) and watts per GB than any other similarly performing flash array on the market today, with improvements that will continue to lead the industry.

“Fusion-io is excited to work with NextIO in the creation of their new vSTOR platform, which is setting new records for performance and adaptability,” said Jim Dawson, Senior Vice President of Worldwide Sales for Fusion-io. “By incorporating Fusion’s ioMemory technology into the vSTOR platform, NextIO will be able to offer customers the ability to extend their capacity and performance beyond the limitations of their server.”

NextIO’s vSTOR Adaptive Application Accelerator represents a paradigm shift in the functionality and design of solid-state arrays, at a lower price/performance point. By utilizing NextIO’s vConnect™ I/O virtualization switching technology, vSTOR can reallocate flash on the fly between servers, eliminating the need to overprovision Tier 0 storage. This capability also reduces the overall cost of the flash by allowing the resource to be shared by multiple servers and applications, something existing solid-state arrays or dedicated flash resources cannot duplicate. The technology promises to speed up a variety of applications such as Microsoft Exchange, relational databases, Massive Multiplayer Game hosting platforms, video editing, and content delivery applications while simultaneously being more cost effective.

“The OEM relationship with Fusion-io gives our NextIO vSTOR platform industry-leading performance density, and represents a significant expansion in the number of media providers for our product,” said Mike Heumann, VP of Marketing, NextIO. “Fusion-io also brings a significant market leadership and customer base to this relationship, which NextIO hopes to leverage with our vSTOR product.”

About NextIO

NextIO, Inc. is the leader in next-generation network consolidation solutions for today’s dynamic data center in a variety of industries including enterprise, oil and gas, high performance computing, digital media and financial services. With its innovative vConnect™ platforms, NextIO offers the unique ability to virtualize I/O technology on any server, operating system, hypervisor and storage architecture. Leveraging PCI Express, NextIO offers true I/O consolidation for any end-point technology. vConnect delivers unprecedented rack-level scalability, with I/O and server resources that can be scaled independently for 50-70% savings in capital, power, and cooling. NextIO’s any-to-any I/O connectivity boosts performance and reliability while streamlining IT deployment, simplifying administration and reducing costs. For more information, visit www.nextio.com.

Supermicro is a trademark or registered trademark of Super Micro Computer, Inc., or its subsidiaries in the U.S. and other countries.

AMD, the AMD Arrow logo, AMD Opteron, and combinations thereof are trademarks of Advanced Micro Devices, Inc. HyperTransport is a licensed trademark of the HyperTransport Technology Consortium. Other names are for informational purposes only and may be trademarks of their respective owners.