Background — High Performance Attached Processors Handicapped By Architecture

The application of high-performance accelerators, notably GPUs, GPGPUs (APUs in AMD terminology) to a variety of computing problems has blossomed over the last decade, resulting in ever more affordable compute power for both horizon and mundane problems, along with growing revenue streams for a growing industry ecosystem. Adding heat to an already active mix, Intel’s Xeon Phi accelerators, the most recent addition to the GPU ecosystem, have the potential to speed adoption even further due to hoped-for synergies generated by the immense universe of x86 code that could potentially run on the Xeon Phi cores.

However, despite any potential synergies, GPUs (I will use this term generically to refer to all forms of these attached accelerators as they currently exist in the market) suffer from a fundamental architectural problem — they are very distant, in terms of latency, from the main scalar system memory and are not part of the coherent memory domain. This in turn has major impacts on performance, cost, design of the GPUs, and the structure of the algorithms:

• Performance — The latency for memory accesses generally dictated by PCIe latencies, which while much improved over previous generations, are a factor of 100 or more longer than latency from coherent cache or local scalar CPU memory. While clever design and programming, such as overlapping and buffering multiple transfers can hide the latency in a series of transfers, it is difficult to hide the latency for an initial block of data. Even AMD’s integrated APUs, in which the GPU elements are on a common die, do not share a common memory space, and explicit transfers are made in and out of the APU memory.

EMC's Project Bourne morphed into ViPR at the EMC World 2013 event at Las Vegas last week. It seems like everyone has a different take on what should be included in SDS, and my definition and implementation guidelines can be found in this report. Like other vendors, EMC is promising to revolutionize the way customers will provision, manage and create storage resources using ViPR, which will become a key component in the vendor's Software Defined Data Center strategy for virtualizing compute, networking, and storage resources.  Unlike other years, where EMC bombarded its attendees with dozens of product launches, this year's show focused almost entirely on ViPR, which makes sense given the importance of this technology. ViPR is expected to become generally available in the latter half of 2013, and like all other SDS implementations, ViPR is designed to reduce the number of administrators it takes to manage rapidly growing data repositories by using automation and self-service provisioning. So what's under ViPR's covers?

This blog has been kindly provided by Sandy Winschief, a vendor/supplier management specialist at TCS, who previous provided: A Late New Year’s Resolution: Be Nice To A Supplier And See What Happens. They are Sandy’s words with just a little editing on my part. Enjoy…

Service integration?

Service integration (SI) is already one of the IT buzzwords of 2013; you might also hear service integration and management (SIAM) which brings in an IT service management (ITSM) perspective. However, just because it is one the most talked about ideas in the IT industry does not mean it is understood.

For now let’s just say that if you could take your E2E IT operations and the complex multi-supplier environment in which it sits and give that pain to someone else to manage while you concentrate on what you do best, on your customers and their needs, then why wouldn’t you? This, in essence, is what pure-play SI is (Stephen – at Forrester we are also speaking to clients about internally operated SI).

Sounds attractive, doesn’t it?  The good news is that it can be. The bad news is that my experience to date has shown SI implementation can be a painful experience if those involved are not prepared.

Planning for service integration

The transition to an SI model is a lot easier if time is invested upfront to:

• Housekeeping key areas of the IT environment
• Engaging a potential SI provider earlier to work with the organization – to help organize and plan for transition readiness.

Yesterday, BMC Software announced that has signed a definitive agreement to be acquired by a private investor group led by Bain Capital and Golden Gate Capital together with GIC Special Investments Pte Ltd (“GIC”) and Insight Venture Partners (collectively, the “Investor Group”).

Under the terms of the agreement, affiliates of the Investor Group will acquire all outstanding BMC common stock for $46.25 per share in cash, or approximately $6.9 billion.

This is one of the largest M&A operations in a long time. Significantly, it has been prepared for quite some time, which culminated in a restructuring a month ago, by which the five product groups operating under BMC Software became one. Instead of having several categories reporting their gains (or losses) we have now one happy family where the gain of one member balances the loss of another. We have also a unique opportunity to have these former product lines working together for a better integration of BMC Software solutions with a corollary prospect of having more R&D investments in previously “weak” categories. Being free of the short term mandatory “good results to satisfy the street” will also participate in building a better BMC Software.

Although fourth quarter results were below the Street expectation by a hair (-$.06 per share and -.04% in Revenue), BMC Software bookings grew 14% from a year ago, with an encouraging result for ESM which was up 9% from a year ago.

Over the past ten years, BMC Software has made its mark on the IT Management Software (ITMS) market, and is today only second to CA Technologies. From what we can see, the privatization of BMC Software provides an opportunity to invest into the future of ITMS and to become a serious contender for first place in the years to come.