Files
yann64andClaude Opus 5.5 b274874139 All machines linked to their forum thread (full comment import)
1,289 discussions, 9,527 archived comments; pages no longer carry the comments themselves.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01YDRjtmAiNpo5mK5N3d7QwZ
2026-10-13 17:26:43 +02:00

101 lines
3.9 KiB
Markdown

---
title: Amstrad PC 1640
categories:
- Computers
types:
- Professional Computer
origins:
- United Kingdom
years:
- '1986'
cpus:
- Intel 8086
params:
name: PC 1640
image: /media/museum/photos/amstrad_pc1640.jpg
company: /companies/amstrad
company_name: Amstrad
specs:
- label: NAME
value: PC 1640
- label: MANUFACTURER
value: Amstrad
- label: TYPE
value: Professional Computer
- label: ORIGIN
value: United Kingdom
- label: YEAR
value: '1986'
- label: CPU
value: Intel 8086
- label: SPEED
value: 8 mHz
- label: RAM
value: 640 KB
- label: VRAM
value: 256 KB ?
- label: ROM
value: 64 KB
- label: TEXT MODES
value: 40 x 25 / 80 x 25
- label: GRAPHIC MODES
value: |-
All EGA graphic modes (maximum : 640 x 350)\
Hercules mode for monochrome versions (maximum 720 by 350)
- label: COLORS
value: 16 among 64
- label: SOUND
value: Beeper
- label: I/O PORTS
value: Four 8-bit ISA slots (one being reserved of the internal HD controller), Centronics, RS232c, Mouse port (proprietary)
- label: BUILT IN MEDIA
value: one or two 5.25'' disk-drives, 20 or 30 MB HDD
- label: OS
value: MS DOS 3.2
- label: POWER SUPPLY
value: PSU built-in
- label: PERIPHERALS
value: Most of the standard IBM/PC cards
tabs:
- links
forum_thread: 39
links:
- title: Intel GEM page
url: http://www.seasip.demon.co.uk/Gem/index.html
text: a page dedicated to the GEM graphical user interface
related:
- /museum/amstrad-pc-1512
- /museum/amstrad-ppc-512-640
- /museum/ibm-pc-at
- /museum/ibm-pc-model-5150
- /museum/ibm-pc-portable-model-5155
- /museum/ibm-pc-xt-286
- /museum/ibm-pc-xt-model-5160
- /museum/ibm-ps-1
- /museum/micronique-hector-xt
oc_id: 19
source:
url: https://web.archive.org/web/20230520105232/http://www.old-computers.com/museum/computer.asp?c=19
date: '2023-05-20'
---
The Amstrad PC 1640 was the successor to the [Amstrad PC 1512](/museum/amstrad-pc-1512/). It had the same characteristics as its predecessor except for added memory (640 KB instead of 512 KB) and the EGA graphics standard.
It also had great success, but to a lesser extent than the PC 1512. As the PC 1512, the Amstrad PC 1640 came with the GEM graphical user interface, from Digital Research, an alternative to Windows.
The PC-1640 was marketed under the name PC-6400 in the USA. It was also sold in Germany and maybe some other European countries under the name Schneider.
Three different kinds of monitors where supported, monochrome (-> Hercules), low-res (max 640 x 200) and hi-res (max 640 x 350). The hi-res monitor had a fan for the power supply as opposed to the other monitors which where fan-less. With the low-res monitor you could choose between a full CGA compatible mode (required for many CGA games) and an EGA mode (used for 640 x 200 x 16 eg. for GEM or 320 x 200 x 16 for several games).
---
Contributors: André Janz
**Charles Da Silva** adds:\
The 1640 was first introduced in the US and after a few months in Europe (which infuriated the British media, some of them having already been harsh critics on the 1512 : fan problems - forgetting that since the PSU was in the monitor, it was not needed - not 'fully' compatible as it was then thought. All of this proved to come from IBM itself, which made Alan Sugar really angry).\
You can find all this info in 'Alan Sugar' from late D. Thomas.
**Ex Cathedra's** memories:\
I did a large amount of development on these machines in the late 80s. They were surprisingly good, with only a few bugs and incompatibilities with the IBM PC standard. Bearing in mind the price differential, we had no problems living with these.\
The RTC and BIOS settings were preserved by 4xAA batteries under the monitor - a configuration I wish we still had today! There was a minor (patchable) bug in the BIOS which caused a div/0 error at midnight each night if you'd left it running a long compilation...