Mish will they pull that tokyo trick and say they cant manage all the volume and have to go back to the old HUMAN trader systems? hehe
DJ Archipelago In Pact To Make Fincl Chief Available To NYSE
.
WASHINGTON (Dow Jones)--Archipelago Holdings Inc. (AX) said Wednesday it has signed an agreement with the New York Stock Exchange under which Archipelago Chief Financial Officer Nelson Chai will be available to provide services to the NYSE.
Archipelago and the NYSE are set to merge some time next month in a deal that will turn the exchange into a publicly traded company.
Under the agreement, Chai will provide services "commensurate with the duties of the chief financial officer of NYSE," according to a Securities and Exchange Commission filing. During the term of Chai's service to the NYSE, he will continue to be an Archipelago employee, and Archipelago can recall Chai at any time, according to the filing.
The agreement will end once the two companies merge or if any party ends the agreement.
Also, as required under the agreement, Chai resigned all officerships and directorships he holds with Archipelago and its subsidiaries, according to the filing.
Archipelago has named Stephane Biehler to serve as interim principal financial officer and interim principal accounting officer. Biehler joined Archipelago as a managing director and corporate controller in March 2004, according to the filing.
investor.archipelago.com
Based on the new Sun Grid and Archipelago's electronic matching technology, the companies plan to introduce a new electronic trading environment that will allow customers to bid on CPU usage cycles. Being able to dynamically bid for open compute cycles will provide companies across the globe with unprecedented flexibility in planning for the purchase and use of compute power. This is a new paradigm in computing where companies can access an unlimited number of CPUs as they need them.
"Archipelago leads the way when it comes to electronic trading technology," said Robert Youngjohns, executive vice president of strategic development and Sun financing at Sun Microsystems, Inc. "With Sun Grid, and Archipelago's matching technology, we expect companies will be able to access an unlimited number of CPUs as they need them -- and have access to technology that is reliable, simple to use, powerful, and sophisticated -- at a single point of contact."
"We believe the technological underpinnings of the Archipelago Exchange could be customized to trade nearly anything, and as the demand for computing power increases, we see great potential in building an exchange for trading CPU usage cycles," said Steve Rubinow, CTO of the Archipelago Exchange. "We're excited about partnering with Sun and developing this opportunity."
The compute exchange announcement comes on the back of Sun's new utility offerings, which include the Sun Grid compute utility, a $1 (USD) per CPU per hour pay-per-use offering ($1/cpu-hr), and the Sun Grid storage utility, a $1 (USD) per gigabyte per month offering ($1/GB-mo). In the coming months, Sun will also roll out additional Sun Grid offerings for the desktop and developer communities.
iee.org
Utility computing is one response to the increasingly complex and costly nature of much of today’s IT. However, the move from using dedicated IT resources that are tied to a specific business function or process, towards a more standardised and shared environment, introduces a number of system-management considerations.
The degree of system management that a customer will have to undertake depends on the type of arrangement they have with the service provider. For example, IBM and HP allow customers to rent time on their hardware. System management will often be included as part of the package, although some will elect to handle this element themselves.
One of the first areas to feel the influence of the utility model was grid computing, which has principally been concerned with accelerating the processing of compute-heavy scientific and engineering problems. With such tasks it is generally possible to make sensible judgements at the outset about how much hardware is required and over what timescale.
However, the grid computing problems being tackled in the enterprise domain can entail a much greater level of unpredictability. For example, the loading on a server farm that is being used to handle the transaction processing associated with a website will follow a much more erratic pattern. This may involve unforeseen spikes in usage that are difficult to predict or accommodate, except by over-provisioning the system. “With these enterprise systems, you really need an insurance policy of over-provisioning”, said Shahin Khan of Azul Systems. The idea of on-demand computing expresses a current goal of hardware and system management software developers, to allow hardware to be provisioned automatically, in real-time, as and when it is needed.
One radical approach that has been proposed by the start-up company Azul is aimed at server systems that handle large quantities of Java processing. Dubbed ‘network-attached processing’, the scheme relies on a dedicated compute appliance that is optimised to provide processing for Java and similar object-oriented, virtual machine-based application software. Each server runs specialised proxy software that directs them to send any Java-intensive tasks to this dedicated appliance, which has a large amount of capacity to handle spill-over situations. The scheme offers one approach to automating the task of optimising the utilisation of a collection of servers. However, it does require some changes to the architecture of the system.
An element of automation that is currently being introduced to the management of IT equipment, such as server or storage farms, is that of business service management (BSM). The term is being used to refer to a layer of functions that sit above the existing system-management layer, supporting the effort to manage components in the system in a way that takes account of their business function. “Today’s service providers and enterprises tend to have very large systems”, said Jim White of Managed Objects. “In a typical day, there might be several conflicting demands for provisioning.”
With traditional, component-level system management, a change in the loading on a system might cause one application to be automatically provisioned at the expense of another, on a first-come-first-served basis, irrespective of their relative importance from a business point-of-view. The service-level view of the system supported by BSM makes it possible to assign priorities to tasks so that the system will provision them appropriately. Software solutions to support BSM are incorporated in IBM’s Tivoli product range as well as products from Managed Objects and Computer Associates, for example.
Where multiple users share the computing resources of a common hardware infrastructure, an important priority is to automate the way resources are allocated to these users. One possibility that is being explored here is the use of market-based methods, whereby users can bid for time on a shared computing resource. The practicalities of this kind of resource-brokering are currently being worked out by a number of groups. HP, for example, is evaluating this method of resource allocation with SE3D, a project it is running in conjunction with Alias, a supplier of 3D graphics software tools.
Twelve UK animation groups are participating in the project, which gives them access to HP’s 3D rendering computing infrastructure over the Internet. Sun is also exploring the idea of a commodities marketplace for computing resources. The company is currently developing a platform to support this kind of ‘CPU-cycle exchange’ inconjunction with Archipelago Holdings, which operates the Archipelago Exchange, an online marketplace for trading equity securities.
The utility computing market has a long way to go and will explore many more options before it settles on the few approaches that customers actually buy into. |
