TPA-11/580 - Revision history

Jnc: New category

2024-01-01T22:38:56Z

New category

Jnc: /* External links */ typo

2019-12-31T09:30:28Z

‎External links: typo

Jnc: +cat

2019-12-12T17:21:18Z

+cat

Jnc: +cat

2018-12-19T23:58:25Z

+cat

Jnc: Proper cat, proper format

2018-10-21T18:43:35Z

Proper cat, proper format

Neozeed at 04:35, 9 May 2011

2011-05-09T04:35:44Z

Neozeed at 04:34, 9 May 2011

2011-05-09T04:34:50Z

Neozeed: Created page with "Great information to be found ... [http://hampage.hu/tpa/e_tpa11580.html here]. "The TPA-11/580 is the fastest and high-performance 32-bit, open architecture member of the KFKI..."

2011-05-09T04:32:54Z

Created page with "Great information to be found ... [http://hampage.hu/tpa/e_tpa11580.html here]. "The TPA-11/580 is the fastest and high-performance 32-bit, open architecture member of the KFKI..."

New page

Great information to be found ... [http://hampage.hu/tpa/e_tpa11580.html here].

"The TPA-11/580 is the fastest and high-performance 32-bit, open architecture member of the KFKI TPA-family. With a microcoded central processing unit, the speed and performance is due to its efficient instruction set and the paralell execution of independent procedures. Its operating system is the VAX/VMS virtual memory management system that can use up to 4 GBs of memory and can work with intrinsically infinte programlengths" (KFKI TPA-11/580 Hardware Handbook)

After producing PDP-11 clones, KFKI decided to build a "clone" of the Digital VAX-11/780. There was great demand on computers of this size and performance, but due to the restrictions of the "western countries", no real VAX was allowed to be shipped to Hungary or to any other country of the "eastern bloc". The VAX had an appeal with its PDP-11 compatibility mode, and it was microcoded, so the main goal was to construct an "iron" that runs the original microcode.

The project started in 1983, and ended in 1987. Cloning the [[VAX-11/780]] was hard, because the development team didn't had a machine to examine, they used what they could get access to, so the process took rather long time.

The main specifications of the TPA-11/580:

*Microcoded 32-bit CPU:

244 native instruction + PDP-11 instructions for compatibility mode
*16 x 32-bit general purpose register
*4 CPU operating modes:

kernel, executive, supervisor, user
*32-bit, high-speed, asynchronous system bus:

'''SBI''' (Synchronous Backplane Interconnect)
Devices are connected to the SBI via so-called NEXI (plural for NEXUS). There are three kinds of devices on the bus:
*The CPU
*The memory
*Adapters of other busses like UNIBUS (UBA), MASSBUS (MBA)

The CPU had a cycle time of 200 ns. It was constructed from SSI and MSI ICs (TTL Schottky and ECL). The processor consists of 26 (or 29 depending on options) cards with Intel, AMD, TI and other discrete components. There's no such thing as card-for-card compatibility with the VAX-11/780, so you cannot exchange cards between the TPA CPU and the VAX CPU, but the whole unit operates and looks like from the software's view as if it were an original DEC machine. It run VAX/VMS, the original diagnositics and all software developed for VAX/VMS.
The memory subsystem consists of the memory controller (connected to the SBI) and the storage modules (connected to the controller). The TPA-11/580 could handle two memory controllers, so the maximal memory ammount with the MS580-C controller was 2x4=8 MB. There was another controller, the MS580-F that was to be able to use 64 MB, but I don't know if that was ever manufactured. The subsystems stores everything in 64-bit quadwords with 8 bit of ECC (Error Correcting Code) information attached to each quadword (which gives 72 bits total).

The UBA (Unibus Adapter) connects the main system bus to the asychronous UNIBUS. The adapter has its own arbitration and it also translates the 30-bit SBI addresses to 18-bit UNIBUS addresses, handles DMA and data puffering (The UBA does 4 UNIBUS-cycles on one SBI cycle).

The console system was a qbus machine called TPA-11/110 or TPA-Janus. It was based on the russian MicroPDP-11 compatible Elektronika MC1201.01 card, and contained a 5.25" mini-floppy unit, an EPROM that emulated another floppy and an RX01-compatible floppy disk formatter. This subsystem was able to do everything that an LSI-11-based VAX-console could do, including running diagnostics, doing ODT or bootstrapping the TPA-11/580.

More than 20 TPA-11/580s were sold.

{{stub}}

@@ Line 3: / Line 3: @@
 From hampage.hu:
-: ''The TPA-11/580 is the fastest and high-performance 32-bit, open architecture member of the KFKI TPA-family. With a microcoded central processing unit, the speed and performance is due to its efficient instruction set and the paralell execution of independent procedures. Its operating system is the VAX/VMS virtual memory management system that can use up to 4 GBs of memory and can work with intrinsically infinte programlengths" (KFKI TPA-11/580 Hardware Handbook)
+: ''The TPA-11/580 is the fastest and high-performance 32-bit, open architecture member of the KFKI TPA-family. With a microcoded central processing unit, the speed and performance is due to its efficient instruction set and the paralell execution of independent procedures. Its operating system is the VAX/VMS virtual memory management system that can use up to 4 GBs of memory and can work with intrinsically infinite program lengths" (KFKI TPA-11/580 Hardware Handbook)
 : ''After producing PDP-11 clones, KFKI decided to build a "clone" of the Digital VAX-11/780. There was great demand on computers of this size and performance, but due to the restrictions of the "western countries", no real VAX was allowed to be shipped to Hungary or to any other country of the "eastern bloc". The VAX had an appeal with its PDP-11 compatibility mode, and it was microcoded, so the main goal was to construct an "iron" that runs the original microcode.
-: ''The project started in 1983, and ended in 1987. Cloning the [[VAX-11/780]] was hard, because the development team didn't had a machine to examine, they used what they could get access to, so the process took rather long time.
+: ''The project started in 1983, and ended in 1987. Cloning the [[VAX-11/780]] was hard, because the development team didn't have a machine to examine, they used what they could get access to, so the process took rather long time.
 : ''The main specifications of the TPA-11/580:
@@ Line 23: / Line 23: @@
 ::* ''Adapters of other busses like UNIBUS (UBA), MASSBUS (MBA)
-: ''The CPU had a cycle time of 200 ns. It was constructed from SSI and MSI ICs (TTL Schottky and ECL). The processor consists of 26 (or 29 depending on options) cards with Intel, AMD, TI and other discrete components. There's no such thing as card-for-card compatibility with the VAX-11/780, so you cannot exchange cards between the TPA CPU and the VAX CPU, but the whole unit operates and looks like from the software's view as if it were an original DEC machine. It run VAX/VMS, the original diagnositics and all software developed for VAX/VMS.
+: ''The CPU had a cycle time of 200 ns. It was constructed from SSI and MSI ICs (TTL Schottky and ECL). The processor consists of 26 (or 29 depending on options) cards with Intel, AMD, TI and other discrete components. There's no such thing as card-for-card compatibility with the VAX-11/780, so you cannot exchange cards between the TPA CPU and the VAX CPU, but the whole unit operates and looks like from the software's view as if it were an original DEC machine. It run VAX/VMS, the original diagnostics and all software developed for VAX/VMS.
-: ''The memory subsystem consists of the memory controller (connected to the SBI) and the storage modules (connected to the controller). The TPA-11/580 could handle two memory controllers, so the maximal memory ammount with the MS580-C controller was 2x4=8 MB. There was another controller, the MS580-F that was to be able to use 64 MB, but I don't know if that was ever manufactured. The subsystems stores everything in 64-bit quadwords with 8 bit of ECC (Error Correcting Code) information attached to each quadword (which gives 72 bits total).
+: ''The memory subsystem consists of the memory controller (connected to the SBI) and the storage modules (connected to the controller). The TPA-11/580 could handle two memory controllers, so the maximal memory amount with the MS580-C controller was 2x4=8 MB. There was another controller, the MS580-F that was to be able to use 64 MB, but I don't know if that was ever manufactured. The subsystems stores everything in 64-bit quadwords with 8 bit of ECC (Error Correcting Code) information attached to each quadword (which gives 72 bits total).
-: ''The UBA (Unibus Adapter) connects the main system bus to the asychronous UNIBUS. The adapter has its own arbitration and it also translates the 30-bit SBI addresses to 18-bit UNIBUS addresses, handles DMA and data puffering (The UBA does 4 UNIBUS-cycles on one SBI cycle).
+: ''The UBA (Unibus Adapter) connects the main system bus to the asynchronous UNIBUS. The adapter has its own arbitration and it also translates the 30-bit SBI addresses to 18-bit UNIBUS addresses, handles DMA and data buffering (The UBA does 4 UNIBUS-cycles on one SBI cycle).
 : ''The console system was a qbus machine called TPA-11/110 or TPA-Janus. It was based on the russian MicroPDP-11 compatible Elektronika MC1201.01 card, and contained a 5.25" mini-floppy unit, an EPROM that emulated another floppy and an RX01-compatible floppy disk formatter. This subsystem was able to do everything that an LSI-11-based VAX-console could do, including running diagnostics, doing ODT or bootstrapping the TPA-11/580.
 : ''More than 20 TPA-11/580s were sold.
 ==External links==
 * [http://hampage.hu/tpa/e_tpa11580.html]
 {{Nav VAX}}
-[[Category: VAXen]]
+[[Category: Other VAX Systems]]
 [[Category: VAX Processors]]

← Older revision		Revision as of 09:30, 31 December 2019
Line 42:		Line 42:

	[[Category: VAXen]]		[[Category: VAXen]]
−	[[Category: VAX Processors]	+	[[Category: VAX Processors]]

← Older revision		Revision as of 17:21, 12 December 2019
Line 42:		Line 42:

	[[Category: VAXen]]		[[Category: VAXen]]
		+	[[Category: VAX Processors]

← Older revision		Revision as of 23:58, 19 December 2018
Line 40:		Line 40:

	{{Nav VAX}}		{{Nav VAX}}
		+
		+	[[Category: VAXen]]

@@ Line 1: / Line 1: @@
 [[Image:mta580.jpg|right|200px|thumb|KFKI TPA-11/580]]
-Great information to be found ... [http://hampage.hu/tpa/e_tpa11580.html here].
+From hampage.hu:
-"The TPA-11/580 is the fastest and high-performance 32-bit, open architecture member of the KFKI TPA-family. With a microcoded central processing unit, the speed and performance is due to its efficient instruction set and the paralell execution of independent procedures. Its operating system is the VAX/VMS virtual memory management system that can use up to 4 GBs of memory and can work with intrinsically infinte programlengths" (KFKI TPA-11/580 Hardware Handbook)
+: ''After producing PDP-11 clones, KFKI decided to build a "clone" of the Digital VAX-11/780. There was great demand on computers of this size and performance, but due to the restrictions of the "western countries", no real VAX was allowed to be shipped to Hungary or to any other country of the "eastern bloc". The VAX had an appeal with its PDP-11 compatibility mode, and it was microcoded, so the main goal was to construct an "iron" that runs the original microcode.
-After producing PDP-11 clones, KFKI decided to build a "clone" of the Digital VAX-11/780. There was great demand on computers of this size and performance, but due to the restrictions of the "western countries", no real VAX was allowed to be shipped to Hungary or to any other country of the "eastern bloc". The VAX had an appeal with its PDP-11 compatibility mode, and it was microcoded, so the main goal was to construct an "iron" that runs the original microcode.
+: ''The project started in 1983, and ended in 1987. Cloning the [[VAX-11/780]] was hard, because the development team didn't had a machine to examine, they used what they could get access to, so the process took rather long time.
-The project started in 1983, and ended in 1987. Cloning the [[VAX-11/780]] was hard, because the development team didn't had a machine to examine, they used what they could get access to, so the process took rather long time.
+: ''The main specifications of the TPA-11/580:
-The main specifications of the TPA-11/580:
+:* ''Microcoded 32-bit CPU:
-*Microcoded 32-bit CPU:
+: ''The memory subsystem consists of the memory controller (connected to the SBI) and the storage modules (connected to the controller). The TPA-11/580 could handle two memory controllers, so the maximal memory ammount with the MS580-C controller was 2x4=8 MB. There was another controller, the MS580-F that was to be able to use 64 MB, but I don't know if that was ever manufactured. The subsystems stores everything in 64-bit quadwords with 8 bit of ECC (Error Correcting Code) information attached to each quadword (which gives 72 bits total).
-native instruction + PDP-11 instructions for compatibility mode
+: ''The UBA (Unibus Adapter) connects the main system bus to the asychronous UNIBUS. The adapter has its own arbitration and it also translates the 30-bit SBI addresses to 18-bit UNIBUS addresses, handles DMA and data puffering (The UBA does 4 UNIBUS-cycles on one SBI cycle).
-kernel, executive, supervisor, user
+: ''The console system was a qbus machine called TPA-11/110 or TPA-Janus. It was based on the russian MicroPDP-11 compatible Elektronika MC1201.01 card, and contained a 5.25" mini-floppy unit, an EPROM that emulated another floppy and an RX01-compatible floppy disk formatter. This subsystem was able to do everything that an LSI-11-based VAX-console could do, including running diagnostics, doing ODT or bootstrapping the TPA-11/580.
-'''SBI''' (Synchronous Backplane Interconnect)
+: ''More than 20 TPA-11/580s were sold.
-The CPU had a cycle time of 200 ns. It was constructed from SSI and MSI ICs (TTL Schottky and ECL). The processor consists of 26 (or 29 depending on options) cards with Intel, AMD, TI and other discrete components. There's no such thing as card-for-card compatibility with the VAX-11/780, so you cannot exchange cards between the TPA CPU and the VAX CPU, but the whole unit operates and looks like from the software's view as if it were an original DEC machine. It run VAX/VMS, the original diagnositics and all software developed for VAX/VMS.
+==External links==
-The UBA (Unibus Adapter) connects the main system bus to the asychronous UNIBUS. The adapter has its own arbitration and it also translates the 30-bit SBI addresses to 18-bit UNIBUS addresses, handles DMA and data puffering (The UBA does 4 UNIBUS-cycles on one SBI cycle).
+* [http://hampage.hu/tpa/e_tpa11580.html]
-The console system was a qbus machine called TPA-11/110 or TPA-Janus. It was based on the russian MicroPDP-11 compatible Elektronika MC1201.01 card, and contained a 5.25" mini-floppy unit, an EPROM that emulated another floppy and an RX01-compatible floppy disk formatter. This subsystem was able to do everything that an LSI-11-based VAX-console could do, including running diagnostics, doing ODT or bootstrapping the TPA-11/580.
+{{semi-stub}}
 {{Nav VAX}}

← Older revision		Revision as of 04:35, 9 May 2011
Line 1:		Line 1:
		+	[[Image:mta580.jpg\|right\|200px\|thumb\|KFKI TPA-11/580]]
		+
	Great information to be found ... [http://hampage.hu/tpa/e_tpa11580.html here].		Great information to be found ... [http://hampage.hu/tpa/e_tpa11580.html here].

← Older revision		Revision as of 04:34, 9 May 2011
Line 36:		Line 36:

	{{stub}}		{{stub}}
		+	{{Nav VAX}}