Compaq's secret VMS plans

Dear OpenVMS Engineering Colleagues,

It¡¦s been nearly four months since we announced the Alpha-to-IPF port. The OpenVMS Engineering team has made amazing progress in identifying the work needed and the direction we are taking for the overall VMS business. I want to give you some of the highlights of the porting effort in this memo.

We have a multi-phased plan for delivering IPF-based OpenVMS systems.

Phase 1 involves booting OpenVMS on an IPF system. We will be using cross compilers for C, Bliss, and Macro running on Alpha initially for this, creating a working boot environment, and then bringing in the native compilers. We are targeting H2 2002 for this phase.

Phase 2 is to ship a revenue-generating environment to key ISVs, partners, and early adopters. This will include a core set of compilers, network protocols, and development tools, and will be fully supported by services. We are targeting at least two shipments to key ISVs, partners and early adopters during this phase ƒ{ the first in early 2003 with the second following in the second half of the year.

Phases 3, 4, and Beyond are to ship production quality, revenue-generating environments to general distribution. We will roll out increasing suites of layered products and ISV solutions over these phases, and will offer increasing levels of clustering over these phases. The first production quality phase is scheduled for early 2004.

Some other highlights in the plan:

Our capital plan was approved by management, and we should have our first IPF-based systems arriving soon. It is uncertain whether we will use ProLiant DL590/64 platforms or HP hardware like the NSK developers are using for their porting project.

Our incremental headcount plan was approved, and we are hiring new developers.

We will have a binary translator from Alpha to IPF for VMS.

We will develop and offer Mixed Architecture VMSclusters. Technically we can offer mixed VAX-Alpha-IPF clusters. We need to figure out what customers will want in the 2004-and-future timeframe, and as always, what we can work into our test and qualification resource plan. The need for VAX support is somewhat questionable.

We will be porting almost the entire OpenVMS Alpha V7.2 layered product portfolio over to IPF.. We anticipate that we will be able to recompile and relink for the great majority of the layered products with the bulk of the porting effort devoted to testing and qualification. We don¡¦t think we need to port products in retirement, product in mature product support, and products that we never ported to Alpha.

Our key ISVs, including Oracle, are very supportive of the port. We now have quotes from over half of our top ISVs committing to port to IPF, and we have a goal of achieving 90 percent buy-in by the end of the quarter. We are working closely with our key ISVs to make sure we are doing everything possible to allow them to support OpenVMS on Alpha and IPF systems quickly, easily, and effectively. The Compaq/Intel deal has matching funds from Intel for ISV porting, and we will leverage this to bring significantly increased resources to our ISVs.

Marketing also has matching funds as part of the agreement, and we have put a plan forward to leverage our H2 2001 marketing spending with Intel funds. We have a detailed communication plan rolling out, and are working a large number of go to market plans. In the meantime we need to keep our attention focused on selling current and future Alpha systems, and we need to make sure we keep delivering key messages and features for Alpha while we do the port.

Thanks for all your hard work, both in developing the IPF plan and in keeping the engineering, support, and marketing going for the existing OpenVMS business. It has been an exciting four months and I know many of you have spent a lot of extra time helping with this critically important project.

Best Regards

Mark Gorham