American Computer and Peripheral: Difference between revisions

Content deleted Content added
No edit summary
Line 22:
Various companies such as [[Daisy Systems]] and [[Valid Logic Systems]] manufactured software development workstations equipped with the 386 microprocessor and running Intel's own assembler, compilers, and software utilties as early as December 1985, when pre-production batches of 386es were manufactured. However, these workstations were large, cumbersome to set up and expensive, costing several thousands of dollars.{{sfn|Goering|1985|pp=33–34}} The 386 Translator, by comparison, cost $895 (equivalent to ${{Inflation|US|895|1986|r=0|fmt=c}} in {{Inflation/year|US}}) with a 386 included or $395 (${{Inflation|US|395|1986|r=0|fmt=c}} in {{Inflation/year|US}}) without.{{sfn|Chabal|Ranney|1986|p=8}} In addition, existing ATs could be equipped with the 386 using AC&P's module, avoiding the need for a dedicated workstation. Counter-intuitive to the nature of an upgrade module, however, the 386 Translator ran an AT computers 10 percent {{em|slower}} than a stock computer with a 286. This was due to the module inserting [[wait state]]s in order for slower AT-grade memory chips to work with the faster 386. Aside from this performance penalty, the 386 Translator allowed software developers with ATs to get a head start on learning 386's new [[virtual 8086 mode]].{{sfn|Nelson|1986|p=32}}
[[File:American 386 Turbo.jpg|thumb|386 Turbo, showing ISA card with 386 and processor module connected via two ribbon cables]]
Just three months after the release of the 386 Translator, in November, AC&P introduced the 386 Turbo expansion board. Like the Translator, the Turbo board allowed users to upgrade their existing ATs with the 386 processor, this time with the promise of increased speed over the AT's 286 processor. The company touted a 400 percent increase in software performance and claimed that the Turbo could double the clock speeds of ATs running between 6 and 12 MHz.{{sfn|Staff writer|1986c|p=16}} The company later revised their claim to only double the clock speeds of 6 and 8 MHz 286s,{{sfn|Miller|1986|p=77}} as 386 processors at the time were not rated for 24 MHz.{{sfn|Satchell|19871987b|p=56}} The 386 Turbo allowed users to switch the clock speed of the 386 on the fly, and it also included 1 MB of [[cache memory]].{{sfn|Staff writer|1986c|p=16}}
 
The board was comparable to [[Intel Inboard 386|Intel's Inboard 386]], which came out at the same time. Both boards plugged into one of the AT's 16-bit ISA expansion slots. While Intel offered a version of the Inboard that could work on XTs, the 386 Turbo could only be used by ATs.{{sfn|Crabb|1987|p=213}} The Turbo's 1 MB of memory was strictly used for cache, while the Inboard could accept up to 4 MB of memory chips to be used as conventional RAM, on top of having 64 KB of cache memory itself.{{sfn|Crabb|1987|pp=214–215}}
 
==Reception==
AC&P's peripherals received mixed reviews throughout the company's short life. Stephen Satchell of ''[[InfoWorld]]'' found that the Turbo 386 failed to double the performance of ATs with 6 MHz 286 processors, as claimed by the company, instead only increasing performance by 83 percent. He felt that this was the product's biggest downfall, because 6 MHz ATs had the slowest clock speeds of the AT class, and thus users with these computers would have been the perfect market for the Turbo. Satchell contrasted the Turbo with Intel's Inboard, which increased performance of 6 MHz ATs by 250 percent,{{sfn|Satchell|19871987b|p=56}} surmounting even the Turbo's boost on 8 MHz machines.{{sfn|Satchell|19871987b|p=57}} Howard Marks of ''[[PC Magazine]]'' found the increased performance adequate on his AT but panned the lack of memory beyond the 1 MB used for cache; access to memory on the computer's [[motherboard]] above 1 MB would be bottlenecked by the AT's 16-bit data bus, negating the processing speed of the 386.{{sfn|Marks|1987|p=245}} Satchell praised the company's Abovefunction [[multifunction board]] as a bargain, on the other hand.{{sfn|Satchell|1987a|p=60}}
 
The poor quality of the documentation provided with the company's products was a source of frequent criticism. Of the Turbo 386, Satchell wrote that its manual was "lacking in several key areas. Both the wording and the diagrams are unclear, and a user could easily damage the system board by removing the 286 chip as directed in the manual".{{sfn|Satchell|1988|p=48}} Marks found an error in the same manual: a [[Jumper (computing)|jumper]] on the board was factory-set to "slow mode" as stated in the manual, underclocking the 386 to be later configured for "fast mode" in a program provided with the Turbo 386. However Marks discovered that the Turbo 386 only worked on his PC AT with jumper configured for "fast mode"—the AT displayed nothing when he first turned it on with the Turbo 386 configured to "slow mode", leading him to believe that he had destroyed his computer.{{sfn|Marks|1987|p=245}} When the company released a [[computer mouse|mouse]] in 1987, Christopher Barr of ''PC Magazine'' found that it only worked with [[Mouse Systems]] drivers not included with the mouse, a trait not mentioned in its manual.{{sfn|Barr|1987|p=122}}
Line 264:
}}
* {{cite journal
| ref={{sfnRef|Satchell|1987a}}
| last=Satchell
| first=Stephen
| title=Six Pack Premium: EEMS Superset Differentiates Add-in Board from Rivals
| url=https://books.google.com/books?id=xTAEAAAAMBAJ&pg=PA56
| date=January 12, 1987
| work=InfoWorld
| publisher=CW Communications
| volume=9
| issue=24
| pages=58–60
| accessdate=October 14, 2021
| via=Google Books
}}
* {{cite journal
| ref={{sfnRef|Satchell|1987b}}
| last=Satchell
| first=Stephen