801103 Honeywell Level 6

This document, a "Competitive Update" from Digital Equipment Corporation (DEC) dated November 3, 1980, provides an in-depth competitive analysis of Honeywell's Level 6 Series minicomputers.

Purpose: The primary goal is to equip DEC's sales and marketing teams with information on Honeywell's Level 6 systems, identify their strengths and weaknesses, and outline strategies for effective competition.

Honeywell Level 6 Overview: Honeywell introduced the Level 6 Series in 1976 to re-enter the minicomputer market after its merger with General Electric. The series includes models like the 6/23, 6/33, 6/43, 6/47, 6/53, and 6/57, positioned to compete with systems such as DEC's PDP-11 and Prime's offerings across various market segments (entry-level, mid-range, and high-end commercial/scientific applications). By June 1979, over 5,000 Level 6 systems had been installed, with a manufacturing capacity of 400 systems per month.

Honeywell Strengths (as perceived by DEC):

• Hardware: Modular design, compatibility across models, easy field upgrades, generally lower-priced peripherals, and reliable hardware.
• Software: Offers a CODASYL DBMS (IDS II), transaction processing capabilities, shareable/reentrant code, and comprehensive accounting/financial/manufacturing packages. Their operating systems (MOD 200, 400, 600) typically do not require SYSGEN (system generation), which can be attractive to less sophisticated users.
• Key Application: Honeywell's "Data Entry Facility" (DEF) is noted as a significant selling point, enabling successful penetration into their large mainframe customer base by allowing interactive inquiry into host files.
• Support: Generally good hardware and software field support.

Honeywell Weaknesses & DEC Opportunities:

• Performance: Lacks high-performance real-time interfaces; multiuser interactive performance is noted to degrade with 16 or more users, and Honeywell lacks public multiuser benchmark data. Certain controller failures can "lock up" the entire system.
• Memory & Architecture: Limited 20-bit address space, high memory costs, and the non-virtual memory operating system (GCOS-6) makes running large programs challenging, often requiring more physical memory than comparable DEC systems (programs tend to be 30-35% larger due to 2-word addresses). No 32-bit architecture.
• Software Limitations: MOD 600 lacks true real-time support and is not a virtual memory system. Overall, Honeywell's OS lacks modern features like password/login procedures and robust task swapping. Language implementations (FORTRAN, COBOL) are often lower-level ANSI standards, lacking features like symbolic debuggers.
• Networking: Offers only basic file transfer capabilities, significantly less advanced than DEC's DECnet. Limited support for connecting to other vendors' mainframes (e.g., CDC, UNIVAC, SNA, though 3271 support was announced).
• Market Penetration/References: Honeywell's Level 6 MOD 600 has not gained wide user acceptance, and Honeywell may struggle to provide strong customer references for this system. DEC has a significantly larger installed base.

DEC's Competitive Strategies: The document advises DEC to:

1. Stress Networks & DDP: Highlight DECnet's advanced capabilities against Honeywell's limited file transfer.
2. Emphasize Ease-of-Use & References: Leverage DEC's 20+ years of interactive systems experience, user-friendly commands (DCL), and vast installed base (over 235,000 systems) with strong references, particularly where Honeywell lacks them (e.g., MOD 600).
3. Promote Multiuser Performance: Challenge Honeywell's single-user benchmark claims by emphasizing the proven multiuser performance and scalability of DEC's VAX/VMS, RSTS/E, and high-end PDP-11s (supporting 32+ users).
4. Highlight Virtual Memory: Stress the advantages of VMS over Honeywell's non-virtual memory GCOS-6 for large program capability.
5. Compare COBOL/FORTRAN Performance: Argue that PDP-11 COBOL on an 11/44 can match or exceed 6/57 performance at lower cost, and VAX-11/750 outperforms Level 6. Highlight Honeywell's low-level language implementations.
6. Showcase Customer Support: Emphasize DEC's comprehensive support (educational, CSS, remote diagnostics, Telephone Support Center) compared to Honeywell's more localized GCOS-6 software support.
7. Address Honeywell Objections: Provide counter-arguments for Honeywell's claims regarding timesharing, large program sizes, distributed processing, and compatibility by contrasting them with DEC's superior offerings and more robust architecture. Advise against certain benchmarking types where Honeywell has an advantage (e.g., compile/link times).

In essence, the document positions DEC as the superior choice due to its advanced architecture (virtual memory, 32-bit), robust operating systems, comprehensive networking, proven multiuser performance, and extensive support infrastructure, especially when compared to the perceived limitations and high costs of Honeywell's Level 6 series in more demanding environments.