Banks and Outsourcing: Just Say ‘Latency’

By Derrick Harris

September 24, 2008

A common critique of external cloud computing services is that big-time IT users, like major corporations and financial institutions, are nowhere near getting on board. That might be true for the new breed of “cloud” services, but for the financial services sector, at least, outsourcing is far from a dirty word.

Especially in the world of electronic or algorithmic trading, latency becomes a huge issue that, for some firms, can only be solved by hosting trading platforms in the same datacenters as major exchanges. For those with less-constant, less real-time demands, renting cycles on a global grid architecture, or even renting your own dedicated grid, can guarantee CPUs whenever and wherever they are needed. However they choose to do it — and even if they opt otherwise — the financial world understands the benefits, pitfalls and nuances of outsourced IT.

Ted Chamberlain, research director in the Networking & Communications Services practice at Gartner, believes we’re actually experiencing a bit of an outsourcing renaissance. Traditionally, the financial sector does not like to let technology out of its grip because they view it as such a differentiator, but “I think we’re at the point now where so many of these financial houses either are running out of space or aren’t in the locations they want to be,” says Chamberlain. “Conversely, people like BT Radianz and Savvis actually have started to, I think, build their portfolios to lure [financial institutions] away from their own datacenters.” Especially in the past 12 months, he adds, the proliferation of exchanges going online has drawn a wide range of financial institutions — from investment banks to brokerage houses — to look into hosted datacenters for the sake of interconnecting with the various sources of market data. (Customers with space in one of Savvis’ 29 international datacenters, for example, can cross-connect with anyone in any of the provider’s other locations.)

“It’s definitely becoming a larger trend,” Chamberlain forecasts. “I don’t think we’re going to see a complete flip of all these financial service companies outsourcing to these exchanges in a cloud computing model, but we do see them diversifying and putting some level of their infrastructure in certain providers.”

Although Gartner does not forecast this market, Chamberlain says his gut feeling is that these services will grow somewhere between 30-35 percent year over year, with the EMEA and APAC markets experiencing a doubling in size.

Alex Tabb, a partner in the Crisis & Continuity Services division of the Tabb Group, also sees an increase in outsourcing interest, although not across the board. The big sell-side investment banks experimented with outsourcing several years ago but many have since shied away from the practice, he says. The reason is that potential cost savings were not worth the sacrifices they had to make in terms of control, management and integration. And with such large operations, outsourcing just became too much to handle. For medium-sized banks and smaller buy-side firms, though, Tabb says hosted solutions make a lot of sense because it is easier to pay someone with that expertise than to staff an entire IT department.

Latency: Public Enemy No. 1

One area where there is no contention is the importance of low latency: It is the No. 1 reason financial firms are moving to hosted solutions. Chamberlain calls latency the primary motivator for making the move, particularly when it can be provided without “gargantuantly scaled costs.” Granted, he notes, managed services often bear a premium over simply buying a point-to-point connection or buying traditional content delivery services, but the presence of guaranteed SLAs along with the low latency make it a premium with which they can live.

Tabb says latency is at its most critical in electronic trading scenarios, when trading applications require high-speed access to the exchanges. In particular, he says, “Latency becomes a killer if you’re running VLDBs (very large databases).” An example would be a firm like Bank of New York trading billions of dollars in bonds — for such an operation, Tabb stated, latency will bring down the system. Smaller houses doing algorithmic trading don’t have quite the latency demands as their bigger counterparts because the requirements probably are not as high with 50 people accessing the trading application versus thousands, he added.

“If you’re going to outsource it, it needs to be close — physically, the proximity needs to be close. Because latency, often times, has to do with location,” says Tabb. “If you’re talking about time-sensitive applications, that becomes a deal-breaker.”

Savvis, one of the service providers Chamberlain cites as an industry leader (along with BT Radianz), sees the need for low latency driving customer demand. According to Roji Oommen, director of business development for the financial services vertical at Savvis, latency is a big deal whether firms are doing arbitrage or trying to trying to hide basket trades via division multiplexing and long trade streams. “What they found is it’s a lot cheaper to move your infrastructure inside a datacenter rather than trying to figure out how to optimize your application or get faster hardware and so forth,” he says.

With Savvis’ Proximity Hosting solution (its most popular), which places customers’ trading engines alongside the exchanges’ platforms inside Savvis’ strategically located datacenters, the latency difference compared to in-house solutions is huge, says Oommen. For example, he explains, 80 percent of the New York Stock Exchange’s equities volume occurs over BATS and Archipelago, both of which house their matching engines with Savvis. With Proximity Hosting, he says, latency goes from milliseconds to fractional microseconds. “It would be safe to say,” Oommen adds, “that more and more banks are looking at outsourced datacenter hosting, especially as they participate in markets all over the world, rather than building it in-house — particularly for automated trading.”

In Savvis’ Weehawken, N.J., datacenter, for example, customers share space with the American Stock Exchange, Philadelphia Stock Exchange, BATS Trading, FxAll/Accelor and the New York Stock Exchange.

Xasax, a Naples, Fla.-based company with space at key datacenters across the country (including Savvis Weehawken, Equinix Secaucus, Equinix Cernak and NYSE Metrotech, among others), has created an environment and solution set optimized for hosting financial software that requires access to real-time market data. According to Noah Lieske, Xasax CEO, the customers of the company’s xsProximity solution have the option of housing a trading platform 30 microseconds from NASDAQ. If customers opt to collocate in Equinix Secaucus, they are 1 millisecond from NASDAQ. Essentially, he says, Xasax’s customers want to localize trading logic next to the exchanges, and “they couldn’t do it faster because we built it out the fastest possible route — and if there’s a faster way to do it, we’ll do it.”

“For those who truly understand the low-latency game,” he adds, “it doesn’t take much for them to understand the value of putting their machines at the closest proximity to the exchange as possible.”

Other Driving Forces

Latency, cost and ease of management are not alone in driving financial services customers to outsource. Savvis’ Oommen, for example, credits a deluge of market data with bringing customers into Savvis’ fold. He says the output of options traffic in North America, for example, has increased from 20 MBps to 600 MBps. Therefore, an options trading firm could be staring down a 20x increase (from about $5,000 a month to $100,000 a month) just in raw network connectivity to handle this data load.

From Xasax’s point of view, datacenter power and space constraints also play a role in growing the business. While growth in high-frequency trading is exponential, says Lieske, the fields where Xasax’s customers and the financial institutions collocate are pretty much out of power and space. “As fast as these facilities can be built, they’re being filled up,” he says. By having space in these coveted locations, Lieske sees his company as having “a bit of a monopoly.” However, he acknowledges that limited space sometimes requires Xasax to send customers to secondary locations — a problem that is mitigated by physical cross-connects between the various datacenters. Customers can have a secondary location as their hub but still maintain the lowest possible latency between other datacenters.

In terms of cost, Lieske says a comparably equipped in-house infrastructure could cost a couple of hundred thousand dollars per month versus $20,000-$30,000 with Xasax.

For IBM, who claims a large number of financial customers for its Computing on Demand solutions, the real draw is the ability to handle peak loads without overprovisioning hardware. Christina Cunningham, a project executive on the Computing on Demand division, says IBM’s customers tend to have extreme workloads requiring lots of capacity for short periods of time. These types of jobs include risk management calculations and Monte Carlo simulations. “Why purchase something that you need to have … in your datacenter taking up space 24×7 when you might only need that in a cyclical way?” she asks rhetorically. By renting time on IBM’s global grid, customers can get resources on a dedicated, variable or dynamic basis depending on their needs. Cunningham also cites space and power constraints as a driver, noting that IBM helps customers solve these issues by offering grid resources in three strategic, minimal-latency locations: New York, Japan and London.

But Financial Firms Hate Letting Go of Their Stuff …

Stating that “[i]t is not a one-size-fits-all solution,” Tabb Group’s Tabb says there are technical, management and oversight challenges that outsourcing brings into play. Other key concerns include the effect on the bottom line; continuous capabilities — whether service providers can sustain operations in the event of a power failure or weather event; legal implications of doing computing overseas; and how much control firms require over their data. Tabb says buy-side firms often have very complex algorithms, which they hold dear. “They look at that as the mother lode; that is their firm,” he explains. “They don’t want anyone else to see it, much less touch it.” In terms of continuity, the Tabb Group often tells smaller clients that outsourced solutions offer better results than they can achieve in-house.

“It does take a certain amount of learning curve and risk tolerance to be able to take their trading systems out of their control, so to speak, and into a hosted environment,” says Xasax’s Lieske. Customers have to trust their provider is providing a high-quality environment so the customer can focus on their core competencies rather than the nuts and bolts of running a trading system, he adds. Xasax is in a good position to offer 100 percent uptime because it requires its approximately 40 partners to source things four times (it encourages six), a level of redundancy that allows for multiple failures before a customer might experience ill effects.

Oommen says Savvis has been pretty lucky in terms of having to overcome obstacles. For one, he says, the company can site its hosting of the New York Stock Exchange as a security proof point, as well as many other exchanges and every “brand-name” bank and hedge fund. For basic collocation, Oommen says customer concerns usually center around “is it staffed, is it secure [and] is it outside a nuclear blast radius of New York?”

IBM, says Cunningham, tackles security concerns by offering very secure point-to-point connections, with one datacenter featuring 26 incoming carriers for maximum redundancy. Big Blue also offers a diskless model where CPU and performance are carried out on the grid but no data is left there. However, its biggest security claim probably is IBM’s willingness to work with customers. “Most of the companies know who the manufacturers are that are really secure in the standards they want to abide by in the industry, and so we work with those types of providers to make sure we have the equipment,” Cunningham says.

Additionally, IBM has ethically hacked its infrastructure twice in the past three years “to make sure there’s no way anybody can get in”; complies with ITAR (International Traffic in Arms Regulations) for government customers; and offers “top secret and above” clearance, Cunningham says.

Gartner’s Chamberlain doesn’t see too many issues around security due to the heavy investments made by service providers and the tendency for financial firms to “kick the tires” before adopting any new technology. However, he believes too much growth in the hosting market could actually be an issue. “I don’t think we’re going to see too many instances where information was stolen [or] IP networks were sniffed. I think we’re OK there,” Chamberlain says. “I think the only future potential issue is that with the low-latency requirements, you can only do so much until the speed of light trips you up.”

If scaling gets to the point where providers cannot meet ideal latency levels, Chamberlain wouldn’t be surprised to see a customer like the Chicago Mercantile Exchange bring its systems back in-house and build out its own fiber network. However, he acknowledges, there is a lot of untapped network capacity and dark fiber in this country, so it would take a major growth spike to reach that point.

Virtualization is the Future

Nearly everyone interviewed for this story cites virtualization, in one form or another, as being a key to future growth. Oommen says Savvis has seen a big uptake in virtualization, and the company offers customers the option of running applications in a multi-tenant cloud. He cites asset management firms as a big user here, as they have many end-users who want to see their portfolios in real time — a prime job for Web servers running within the cloud. Of course, he says, there are still concerns around both security and performance along this front, and Savvis doesn’t force anyone to run in a shared environment. In terms of performance, though, he says Web services generally run smoothly on virtualized platforms.

Xasax also is growing through automation and virtualization, from storage to provisioning. One use case the company really likes is the ability to sandbox new clientele in evaluation environments, where potential customers can sign up for a VM, bring it into production, and “get the same connectivity as Credit Suisse or Bank of America,” Lieske says. And while high-frequency traders won’t mess around with a virtual OS, Lieske says they seem to have no problems with virtual storage. Virtualization also allows Xasax to offer a variety of other services, he adds, like production-quality execution management systems, FIX (Financial Information eXchange) gateways for brokers, etc. Internally, virtualization has allowed Xasax to grow with less overhead and management than would have been required in a strictly physical environment.

IBM Computing on Demand’s Cunningham says her division also is seeing an interest in virtualization from clients, particularly around virtual desktops. However, she admits, there are some hurdles to be overcome by customers before they are ready to deploy virtual desktops in real-time scenarios. “When they look at cloud, they’re really focused on server capacity that’s taking up their datacenters,” she explains. “They’re a little bit less concerned about the number of servers taking up their Exchange; what they really want is the high-performance computing, and they want to be able to send it out.” While IBM does have a virtual desktop offering for financial services, it is seeing more questions than takers at this point.

Alex Tabb also sees desktop virtualization taking off, especially among smaller firms. Markets are running 24 hours a day, he says, and traders need to access their terminals no matter where they are in the world, from any computer.  “In today’s economic environment, things change on a dime: the world’s coming to an end, and all of a sudden, in 10 minutes, we’re in the big rally,” he quipped. “Having that flexibility is really important.”

Whatever Works Best

Although the credit crisis of last week was not a failure of IT, Tabb says, the turmoil does underscore the importance of finding the right trading solution, be it in-house or outsourced. And with IT spending among financial service organizations in limbo — a 180-degree turn from “definitely on the rise” six months ago — outsourcing could look even more appealing.

“A strong information technology department and strong capabilities with your human resources and your people can make all the difference in the world,” Tabb says. “Having the ability to react quickly to market changes — and here is where latency becomes a huge issue — can have a significant impact on your trades.”

Subscribe to HPCwire's Weekly Update!

Be the most informed person in the room! Stay ahead of the tech trends with industy updates delivered to you every week!

Advancing Modular Supercomputing with DEEP and DEEP-ER Architectures

February 24, 2017

Knowing that the jump to exascale will require novel architectural approaches capable of delivering dramatic efficiency and performance gains, researchers around the world are hard at work on next-generation HPC systems. Read more…

By Sean Thielen

Weekly Twitter Roundup (Feb. 23, 2017)

February 23, 2017

Here at HPCwire, we aim to keep the HPC community apprised of the most relevant and interesting news items that get tweeted throughout the week. Read more…

By Thomas Ayres

HPE Server Shows Low Latency on STAC-N1 Test

February 22, 2017

The performance of trade and match servers can be a critical differentiator for financial trading houses. Read more…

By John Russell

HPC Financial Update (Feb. 2017)

February 22, 2017

In this recurring feature, we’ll provide you with financial highlights from companies in the HPC industry. Check back in regularly for an updated list with the most pertinent fiscal information. Read more…

By Thomas Ayres

HPE Extreme Performance Solutions

O&G Companies Create Value with High Performance Remote Visualization

Today’s oil and gas (O&G) companies are striving to process datasets that have become not only tremendously large, but extremely complex. And the larger that data becomes, the harder it is to move and analyze it – particularly with a workforce that could be distributed between drilling sites, offshore rigs, and remote offices. Read more…

Rethinking HPC Platforms for ‘Second Gen’ Applications

February 22, 2017

Just what constitutes HPC and how best to support it is a keen topic currently. Read more…

By John Russell

HPC Technique Propels Deep Learning at Scale

February 21, 2017

Researchers from Baidu’s Silicon Valley AI Lab (SVAIL) have adapted a well-known HPC communication technique to boost the speed and scale of their neural network training and now they are sharing their implementation with the larger deep learning community. Read more…

By Tiffany Trader

IDC: Will the Real Exascale Race Please Stand Up?

February 21, 2017

So the exascale race is on. And lots of organizations are in the pack. Government announcements from the US, China, India, Japan, and the EU indicate that they are working hard to make it happen – some sooner, some later. Read more…

By Bob Sorensen, IDC

ExxonMobil, NCSA, Cray Scale Reservoir Simulation to 700,000+ Processors

February 17, 2017

In a scaling breakthrough for oil and gas discovery, ExxonMobil geoscientists report they have harnessed the power of 717,000 processors – the equivalent of 22,000 32-processor computers – to run complex oil and gas reservoir simulation models. Read more…

By Doug Black

Advancing Modular Supercomputing with DEEP and DEEP-ER Architectures

February 24, 2017

Knowing that the jump to exascale will require novel architectural approaches capable of delivering dramatic efficiency and performance gains, researchers around the world are hard at work on next-generation HPC systems. Read more…

By Sean Thielen

HPC Technique Propels Deep Learning at Scale

February 21, 2017

Researchers from Baidu’s Silicon Valley AI Lab (SVAIL) have adapted a well-known HPC communication technique to boost the speed and scale of their neural network training and now they are sharing their implementation with the larger deep learning community. Read more…

By Tiffany Trader

IDC: Will the Real Exascale Race Please Stand Up?

February 21, 2017

So the exascale race is on. And lots of organizations are in the pack. Government announcements from the US, China, India, Japan, and the EU indicate that they are working hard to make it happen – some sooner, some later. Read more…

By Bob Sorensen, IDC

TSUBAME3.0 Points to Future HPE Pascal-NVLink-OPA Server

February 17, 2017

Since our initial coverage of the TSUBAME3.0 supercomputer yesterday, more details have come to light on this innovative project. Of particular interest is a new board design for NVLink-equipped Pascal P100 GPUs that will create another entrant to the space currently occupied by Nvidia's DGX-1 system, IBM's "Minsky" platform and the Supermicro SuperServer (1028GQ-TXR). Read more…

By Tiffany Trader

Tokyo Tech’s TSUBAME3.0 Will Be First HPE-SGI Super

February 16, 2017

In a press event Friday afternoon local time in Japan, Tokyo Institute of Technology (Tokyo Tech) announced its plans for the TSUBAME3.0 supercomputer, which will be Japan’s “fastest AI supercomputer,” Read more…

By Tiffany Trader

Drug Developers Use Google Cloud HPC in the Fight Against ALS

February 16, 2017

Within the haystack of a lethal disease such as ALS (amyotrophic lateral sclerosis / Lou Gehrig’s Disease) there exists, somewhere, the needle that will pierce this therapy-resistant affliction. Read more…

By Doug Black

Azure Edges AWS in Linpack Benchmark Study

February 15, 2017

The “when will clouds be ready for HPC” question has ebbed and flowed for years. Read more…

By John Russell

Is Liquid Cooling Ready to Go Mainstream?

February 13, 2017

Lost in the frenzy of SC16 was a substantial rise in the number of vendors showing server oriented liquid cooling technologies. Three decades ago liquid cooling was pretty much the exclusive realm of the Cray-2 and IBM mainframe class products. That’s changing. We are now seeing an emergence of x86 class server products with exotic plumbing technology ranging from Direct-to-Chip to servers and storage completely immersed in a dielectric fluid. Read more…

By Steve Campbell

For IBM/OpenPOWER: Success in 2017 = (Volume) Sales

January 11, 2017

To a large degree IBM and the OpenPOWER Foundation have done what they said they would – assembling a substantial and growing ecosystem and bringing Power-based products to market, all in about three years. Read more…

By John Russell

US, China Vie for Supercomputing Supremacy

November 14, 2016

The 48th edition of the TOP500 list is fresh off the presses and while there is no new number one system, as previously teased by China, there are a number of notable entrants from the US and around the world and significant trends to report on. Read more…

By Tiffany Trader

Lighting up Aurora: Behind the Scenes at the Creation of the DOE’s Upcoming 200 Petaflops Supercomputer

December 1, 2016

In April 2015, U.S. Department of Energy Undersecretary Franklin Orr announced that Intel would be the prime contractor for Aurora: Read more…

By Jan Rowell

D-Wave SC16 Update: What’s Bo Ewald Saying These Days

November 18, 2016

Tucked in a back section of the SC16 exhibit hall, quantum computing pioneer D-Wave has been talking up its new 2000-qubit processor announced in September. Forget for a moment the criticism sometimes aimed at D-Wave. This small Canadian company has sold several machines including, for example, ones to Lockheed and NASA, and has worked with Google on mapping machine learning problems to quantum computing. In July Los Alamos National Laboratory took possession of a 1000-quibit D-Wave 2X system that LANL ordered a year ago around the time of SC15. Read more…

By John Russell

Enlisting Deep Learning in the War on Cancer

December 7, 2016

Sometime in Q2 2017 the first ‘results’ of the Joint Design of Advanced Computing Solutions for Cancer (JDACS4C) will become publicly available according to Rick Stevens. He leads one of three JDACS4C pilot projects pressing deep learning (DL) into service in the War on Cancer. Read more…

By John Russell

IBM Wants to be “Red Hat” of Deep Learning

January 26, 2017

IBM today announced the addition of TensorFlow and Chainer deep learning frameworks to its PowerAI suite of deep learning tools, which already includes popular offerings such as Caffe, Theano, and Torch. Read more…

By John Russell

HPC Startup Advances Auto-Parallelization’s Promise

January 23, 2017

The shift from single core to multicore hardware has made finding parallelism in codes more important than ever, but that hasn’t made the task of parallel programming any easier. Read more…

By Tiffany Trader

Tokyo Tech’s TSUBAME3.0 Will Be First HPE-SGI Super

February 16, 2017

In a press event Friday afternoon local time in Japan, Tokyo Institute of Technology (Tokyo Tech) announced its plans for the TSUBAME3.0 supercomputer, which will be Japan’s “fastest AI supercomputer,” Read more…

By Tiffany Trader

Leading Solution Providers

CPU Benchmarking: Haswell Versus POWER8

June 2, 2015

With OpenPOWER activity ramping up and IBM’s prominent role in the upcoming DOE machines Summit and Sierra, it’s a good time to look at how the IBM POWER CPU stacks up against the x86 Xeon Haswell CPU from Intel. Read more…

By Tiffany Trader

Nvidia Sees Bright Future for AI Supercomputing

November 23, 2016

Graphics chipmaker Nvidia made a strong showing at SC16 in Salt Lake City last week. Read more…

By Tiffany Trader

BioTeam’s Berman Charts 2017 HPC Trends in Life Sciences

January 4, 2017

Twenty years ago high performance computing was nearly absent from life sciences. Today it’s used throughout life sciences and biomedical research. Genomics and the data deluge from modern lab instruments are the main drivers, but so is the longer-term desire to perform predictive simulation in support of Precision Medicine (PM). There’s even a specialized life sciences supercomputer, ‘Anton’ from D.E. Shaw Research, and the Pittsburgh Supercomputing Center is standing up its second Anton 2 and actively soliciting project proposals. There’s a lot going on. Read more…

By John Russell

TSUBAME3.0 Points to Future HPE Pascal-NVLink-OPA Server

February 17, 2017

Since our initial coverage of the TSUBAME3.0 supercomputer yesterday, more details have come to light on this innovative project. Of particular interest is a new board design for NVLink-equipped Pascal P100 GPUs that will create another entrant to the space currently occupied by Nvidia's DGX-1 system, IBM's "Minsky" platform and the Supermicro SuperServer (1028GQ-TXR). Read more…

By Tiffany Trader

IDG to Be Bought by Chinese Investors; IDC to Spin Out HPC Group

January 19, 2017

US-based publishing and investment firm International Data Group, Inc. (IDG) will be acquired by a pair of Chinese investors, China Oceanwide Holdings Group Co., Ltd. Read more…

By Tiffany Trader

Dell Knights Landing Machine Sets New STAC Records

November 2, 2016

The Securities Technology Analysis Center, commonly known as STAC, has released a new report characterizing the performance of the Knight Landing-based Dell PowerEdge C6320p server on the STAC-A2 benchmarking suite, widely used by the financial services industry to test and evaluate computing platforms. The Dell machine has set new records for both the baseline Greeks benchmark and the large Greeks benchmark. Read more…

By Tiffany Trader

Is Liquid Cooling Ready to Go Mainstream?

February 13, 2017

Lost in the frenzy of SC16 was a substantial rise in the number of vendors showing server oriented liquid cooling technologies. Three decades ago liquid cooling was pretty much the exclusive realm of the Cray-2 and IBM mainframe class products. That’s changing. We are now seeing an emergence of x86 class server products with exotic plumbing technology ranging from Direct-to-Chip to servers and storage completely immersed in a dielectric fluid. Read more…

By Steve Campbell

What Knights Landing Is Not

June 18, 2016

As we get ready to launch the newest member of the Intel Xeon Phi family, code named Knights Landing, it is natural that there be some questions and potentially some confusion. Read more…

By James Reinders, Intel

  • arrow
  • Click Here for More Headlines
  • arrow
Share This