Details

Written by Super User

Category: Latest

Published: 12 May 2013

Hits: 8115

Cloud computing is unquestionably changing the landscape of how traditional enterprise software is being licensed and implemented. Enabling ERP functionality into service-based, on-demand workloads is not only transformingtime toimplement, but the economics is dramatically changing where much of the costs are spread over longer periods of time rather than large upfront costs. This can be a tremendous benefit to organizations. Additionally, although there are security, compliance, and privacy considerations, moving critical applications to industry leading data centers can help improve organizations business continuity, disaster recovery, performance and scalability capabilities.

The Role of Big Data Analytics 

The application for Big Data is extending into many different areas. ERP is no exception.However, with any analytics initiative, it’s important to understand the underlying performance and trending metrics the organization is focused on capturing. From there, they can better align their big data programs and be better positioned to improve their ERP process and system changes.

Fostering Innovation 

Firstly, organizations need to determine how they want to foster innovation. What are the areas they are lagging behind in? How can they improve their go-to-market services, differentiate themselves, and improve their competitive advantages? Once those business models are identified, they can look to ERP to see how it can support these new processes.

My experience as a CIO as well as working with other CIOs is that business value has to take center stage to an organization’s continuous improvement

For example, here at Hitachi Solutions, we are changing business models to support the continuing shift towards cloud computing. This new model requires changes in ERP to support areas such as subscription based billingand new customer support services. We are adopting and driving new functionality in our enterprise systems to support our evolving capabilities.

The Rising Technologies 

I encourage organization’s to first look at what business improvements they need and are willing to make. This can be in the form of process improvements, customer experiences, organizational communication, and collaboration. From there, specific technologies, whether in be in the areas of financial management, trade and logistics, relationship management, customer service, social, or mobility can be leveraged and implemented.

The Changing Role of CIOs

My experience as a CIO as well as working with other CIOs is that business value has to take center stage to an organization’s continuous improvement. Technology moves quickly and there is a lot of it to choose from. Understand the organization challenges and associated opportunities, develop business stakeholder commitment, take time to understand process, the people, and company priorities.

As a modern day CIO it is important to continue developing both my business and technical acumen. Of the two, I find the business and people skills are critical to influencing transformation and driving fundamental improvement. Technical acumen is still very important and I recommend continuing to evolve these skills even if the focus is wider across a broader set of areas versus deeper in one specific technology.

Curbing the Cost Exigency 

There are varieties of ways to help mitigate the cost of ERP. Licensing is clearly a big component to the overall expense of running enterprise software. It is important to regularly assess appropriate licensing. When possible, try to limit intrusive customizations that add complexity, more maintenance, and risk upgradeability.  These all contribute to increased costs. A move to cloud services can also help reduce costs, however this is not always the case. It depends on which cloud services are leveraged.  Moving to a full-fledged SaaS solution eliminates the need for expensive hardware that is compounded when running dev, test, and production environments. However, using IaaS to simply move on-premise servers to cloud servers becomes a little trickier when analyzing costs. It may not be cheaper, but dollar for dollar you will likely experience more value when running in the cloud. This value comes in ease of scalability, ability to dial-up and down performance, improve disaster recovery and business continuity through geo redundancy.

The IoT Factor

IoT as it relates to ERP will have different meaning to different organizations. I think where there is commonality is in IoT’s ability to provide real time, intelligent connectivity to backend processes that can drive proactive process improvements.

Driving the Values 

Driving more value out of ERP doesn’t always require technology changes. I encourage our customers to frequently take inventory of how well their user education is, pay attention to change management, prioritize business enhancements, and where possible, drive more process consistency and standardization. Improvements in these areas alone will drive more value out of ERP.

Author: Mark Veronda, CIO – Global Solutions Group, Hitachi Solutions America, Ltd 

Details

Written by Super User

Category: Latest

Published: 12 May 2013

Hits: 8747

AMD unveiled a great deal of information at Hot Chips about its upcoming “Zen” CPU core and architecture. The new chip has been the subject of an enormous amount of speculation for more than a year, but things have heated up over the past few weeks asleaked benchmarks surfaced and AMD conducted its own public test.

Today’s information dump is the most detail AMD has shared to date — in fact, it’s significantly more information than I expected the company to share until Zen actually launched. Let’s get started.

Zen’s design goals

Zen is best understood as a response to the problems that plagued Bulldozer. AMD’soriginal goal with that architecture was to intelligently share resources between CPU cores, while simultaneously hitting higher frequencies and higher execution efficiencies than AMD’s previous CPU core, K10. Bulldozer’s failure to deliver left AMD in an ugly position: Should it try to repair its old core or return to the drawing board and build something completely new?

Sources we’ve spoken to at AMD suggest that the difficulty of repairing Bulldozer was significant enough that AMD opted to build a new core from scratch with none of Bulldozer’s baggage. That doesn’t mean there’s no Bulldozer DNA in Zen — in fact, AMD has stated that the expertise it gained from improving Steamroller and Excavator’s energy efficiency was put to good use for its newest architecture. Say instead that what design elements AMD does borrow from its previous architectures will be the components of the chip that actually worked well rather than the problematic ones that dominated its performance.

Cache architecture

Much of what went wrong with Bulldozer was linked to its cache subsystem and overall architecture, so that’s a good place to start diving into Zen.

Where Bulldozer used the concept of a CPU module (defined as a pair of cores that shared resources), Zen uses complexes. One CPU complex (CCX) contains four cores, 2MB of L2 cache (512KB per core), and 8MB of L3 cache. That means AMD’s highest-end consumer Zen contains eight cores and 16MB of L3 cache in total, split into 2x8MB chunks. AMD has stated that the two CCXs on an eight-core chip can communicate with each other via the on-chip fabric, though there’s likely a performance penalty for doing so.

Zen’s L3 cache operates as a victim cache for the L1 and L2, meaning data evicted from those caches is stored in the L3 instead. It’s also 16-way associative, which is a significant change from Bulldozer’s 64-way associative L3. A cache with a higher set associativity has a greater likelihood of containing the information the CPU is looking for, but takes longer to search — and one of the issues that crippled Bulldozer was its cache latency at nearly every stage.

We don’t know anything about clock speeds on either the L3 cache or the integrated memory controller. Historically, AMD’s Bulldozer-derived CPUs and APUs have used a clock between 1.8 – 2.2GHz for the L3 cache and IMC.

AMD has stated that L1 and L2 bandwidth is nearly 2x Excavator while L3 bandwidth is supposedly 5x higher. These changes should keep the core fed and support higher performance. The L1 cache is write-back instead of write-through — that’s a significant change that should improve performance and reduce cache contention (Bulldozer’s write-through cache meant that L1 performance could be constrained by L2 cache write speed in some cases).

The CPU core

We’ve already tackled caches, so let’s check out the CPU core itself.

Here’s Zen’s high-level core diagram. There are several significant differences compared with AMD’s older Bulldozer core, including the addition of an op cache, a micro-op queue, and a larger number of integer pipelines per core.

Here’s an expanded view of how the core gets fed. This was another major problem area with Bulldozer — Bulldozer and Piledriver’s shared logic meant that the dispatch unit could only send work to one core or the other every clock cycle. Steamroller later fixed this issue by doubling up dispatch units, but this only resulted in a modest performance improvement.

AMD has taken a page from Intel’s book and implemented an op cache with Zen, even if we don’t know much about the specifics of the feature. This allows the CPU to cache decoded operations that it may need to dispatch repeatedly rather than requiring it to repeatedly decode and dispatch the same instructions. Each Zen core can decode four instructions per clock cycle, but the micro-op queue can dispatch six instructions per cycle. Clearly AMD anticipates that its cache will relieve pressure on the decode units and help keep the core fed while reducing power consumption. Steamroller had a macro-op queue that could hold up to 40 macro-ops but its usefulness was limited to tiny loops.

Like the Bulldozer family, Zen can theoretically fetch 32 bytes of data at a time, though CPU analyst Agner Fog found that the Bulldozer family of cores was practically limited to 21 bytes of data when both cores were in use or 16 bytes if one core was used. He theorized that this limit may have been why doubling up on Steamroller’s dispatch units yielded relatively limited results. Resolving this in Zen could be part of why AMD has significantly improved its IPC.

The integer cores have been rebalanced from the Bulldozer family. Prior to Bulldozer, AMD’s K10 paired three ALUs with three AGUs (address generation units). Bulldozer trimmed this to two ALUs and two AGUs per core. This, combined with the limited dispatch ability in the BD/PD cores, was thought to be a major performance bottleneck until Steamroller added additional dispatch capabilities and slashed the penalty Kaveri took when scaling across multiple cores. (Piledriver and Bulldozer achieved roughly 1.8x of the scaling you’d expect from a “true” dual-core, while Steamroller hit approximately 1.9x.) Four ALUs and two AGUs could boost overall performance compared with Bulldozer’s narrow design, but we’ll have to see how the chip performs in benchmarks.

AMD’s floating point unit will still use 128-bit registers for AVX and AVX2, but latency on some FP operations has been decreased and there are now four pipes instead of three to feed the FPU. The CPU isn’t capable of executing 256-bit AVX instructions in a single cycle. Whether this will prove a detriment in real-world code is an open question, but AVX/AVX2 haven’t boosted general application performance the way SSE2 once did.

Details

Written by Super User

Category: Latest

Published: 12 May 2013

Hits: 8731

Spreadsheets are cheap, easy to manage and can accommodate many different types of data. Therefore, it’s no surprise that, according to an FEI Canada report, 71 percent of U.S. companies still rely on spreadsheets for budgeting and planning. For many small businesses, spreadsheets are a system of first choice when just starting out.

However, something happens when your business matures and becomes more successful. As an organization grows, so much data is generated that it makes manipulating figures and creating accurate financial statements a time-consuming and error-prone task when using spreadsheets.

The biggest problem with spreadsheets is often their isolation from the rest of the enterprise, making collaboration fraught with challenges. Data cannot be automatically or even easily incorporated in the system. Even when it is, it is nearly impossible to prevent multiple versions of the same data from distorting results.

Most organizations accept that spreadsheets are going the way of landlines, encyclopedias and floppy disk drives. However, after making this tough decision, businesses face a challenging prospect: how to get all that data from Excel and into their ERP software of choice. Modern-day ERP offers some powerful tools for seamless integration – and the resulting solution is well worth the time and effort.

Here are six tips for making your transfer from spreadsheets to ERP as easy and painless as possible.

Evaluate Your Current Data Processes

When businesses have used the same systems for years, they can fall into a pattern of habituation, where tasks and reports become automatic, filled out without much thought to their need and use. Only when an issue with an internal procedure/process arises does a system get reevaluated and updated.

The transition from a system of Excel spreadsheets to an ERP solution can be an ideal time to take another look at the processes your business has in place. Evaluate what information you currently gather, how that information is used and what additional information would be beneficial. The added capabilities provided by ERP systems can also provide you with a new lens through which to consider your current processes — and may unearth new procedures you wouldn’t have considered otherwise.

Determine if You Need All the Data

In addition to reviewing processes and procedures, take a closer look at historical data and consider how often it gets used – and what might no longer be needed. Determining a specific date and only transferring files from after this date can optimize the process and save a lot of time and energy. You can always go back and add additional data if needed.

Speaking with an expert who has experience in this area can be helpful, as they will already understand potential pitfalls and help you avoid them. Check with other business owners in your circle who may have already completed their own transitions, or ask the companies you are considering purchasing your system from for references.

Break Data up into Sections

Just like anything else in business, you need a plan to be successful when transitioning data over to a new finance system. Decide if it makes more sense to attempt a mass transition, or a slow and gradual process. Keep in mind that most companies benefit from a gradual process, but there are instances where a quicker transition might be the better option. An example of this would be if the data is used regularly and needs to be accessible in its entirety, even during the transition.

Make a list of the milestones you need to achieve. Include steps like picking out the right product, testing the system and training end users. This list will ease some of the stress associated with the migration.

Choose ERP Product Based on Your Specific Needs

Before choosing an ERP solution, you need to understand all of your options, both customized and out-of-the-box solutions. First, however, you should understand which kind of systems your organization needs. Once you have this information, it will be easier to compare software programs and choose the right solution.

For example, a company that outsources its marketing functions may have little use for an ERP that prioritizes marketing. However, that same company may thrive on the increased knowledge it has from automated financial reports that highlight just where improvement is needed and which areas of the business are performing best.

It sounds like a no-brainer, but when making a decision, pick a finance system that meets your exact needs. Try not to make any assumptions at this stage in the process, as the answer to a quick question can often change the direction that your company chooses to go in. It is important to uncover what exactly each vendor offers and how it could benefit your company’s operations. There are many systems out there capable of managing your finance system, but powerful capabilities are only beneficial if they can be directly applied to your business.

Prepare Your Team

Most likely, you aren’t transitioning to a new financial system alone. Consider how you want to prepare your team to help in the transition. You have two types of people to prepare as part of your transition: team members who will help you migrate your information, and end users of the software programs selected. Get feedback from both of these groups on what they need out of the new system early in the process so they remain more invested over the long run.

Regardless of how you choose to transition the data over, allow extra time in your plans. Something unexpected may come up and you don’t want to be caught off guard. By anticipating that the project may take a little extra time and preparing your team for this, you may be pleasantly surprised with an earlier-than-planned finish.