• Prime bitslice processor

    From Jim Wilcoxson@21:1/5 to All on Fri Apr 24 07:44:00 2020
    Is this the chip that Prime used to implement its CPUs?

    http://www.righto.com/2020/04/inside-am2901-amds-1970s-bit-slice.html

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Dennis Boone@21:1/5 to All on Fri Apr 24 11:04:56 2020
    Is this the chip that Prime used to implement its CPUs? http://www.righto.com/2020/04/inside-am2901-amds-1970s-bit-slice.html

    The 2900 family was used in at least some of the designs.

    There's another common ALU, the 74181, which might be a candidate
    for the earliest machines.

    De

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Bob Allison@21:1/5 to Jim Wilcoxson on Wed Oct 21 19:12:37 2020
    On Friday, April 24, 2020 at 10:44:01 AM UTC-4, Jim Wilcoxson wrote:
    Is this the chip that Prime used to implement its CPUs?

    http://www.righto.com/2020/04/inside-am2901-amds-1970s-bit-slice.html
    The 2901 bit slices were used in the early ecl machines but once we got to the 6000 series we were
    designing our own chips using Motorola MCA2500 ECL Macrocell Arrays on all the boards of the CPU.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Daiyu Hurst@21:1/5 to bobbya on Sat Jul 10 19:55:40 2021
    On Wednesday, October 21, 2020 at 10:12:38 PM UTC-4, bobbya wrote:
    On Friday, April 24, 2020 at 10:44:01 AM UTC-4, Jim Wilcoxson wrote:
    Is this the chip that Prime used to implement its CPUs?

    http://www.righto.com/2020/04/inside-am2901-amds-1970s-bit-slice.html
    The 2901 bit slices were used in the early ecl machines but once we got to the 6000 series we were
    designing our own chips using Motorola MCA2500 ECL Macrocell Arrays on all the boards of the CPU.

    Were the P300 and P400 ECL machines? I hadn't thought so. I have the P300 microcode manual, and we once had the microcode boards for the P400 at Greater Clark County Schools. I remember looking at them, and I'm pretty sure that's what I recall was used.

    On Friday, April 24, 2020 at 12:05:02 PM UTC-4, Dennis Boone wrote:
    Is this the chip that Prime used to implement its CPUs? http://www.righto.com/2020/04/inside-am2901-amds-1970s-bit-slice.html
    The 2900 family was used in at least some of the designs.

    There's another common ALU, the 74181, which might be a candidate
    for the earliest machines.

    I have the data sheets for the 54181/74181 ALU in my 1976 (second printing) Texas Instruments TTL Data Book. if they're not already online somewhere, let me know, I'll scan the pages.

    But I'm a stroked-out old lady; what do I know, lol.

    -Dai

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Dennis Boone@21:1/5 to All on Sun Jul 11 20:24:33 2021
    Were the P300 and P400 ECL machines? I hadn't thought so. I have the
    P300 microcode manual, and we once had the microcode boards for the P400
    at Greater Clark County Schools. I remember looking at them, and I'm
    pretty sure that's what I recall was used.

    No, pretty sure the first ECL machine was the 9950.

    I have the data sheets for the 54181/74181 ALU in my 1976 (second
    printing) Texas Instruments TTL Data Book. if they're not already
    online somewhere, let me know, I'll scan the pages.

    Check http://bitsavers.org/components/ti/_dataBooks/ - I suspect
    it's in there.

    De

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Daiyu Hurst@21:1/5 to Dennis Boone on Tue Jul 13 10:18:09 2021
    On Sunday, July 11, 2021 at 9:24:39 PM UTC-4, Dennis Boone wrote:
    Were the P300 and P400 ECL machines? I hadn't thought so. I have the
    P300 microcode manual, and we once had the microcode boards for the P400
    at Greater Clark County Schools. I remember looking at them, and I'm
    pretty sure that's what I recall was used.
    No, pretty sure the first ECL machine was the 9950.
    I have the data sheets for the 54181/74181 ALU in my 1976 (second
    printing) Texas Instruments TTL Data Book. if they're not already
    online somewhere, let me know, I'll scan the pages.
    Check http://bitsavers.org/components/ti/_dataBooks/ - I suspect
    it's in there.

    De

    Yup. Minus the faux-leather covers. Al seems not to favor the covers.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Dennis Boone@21:1/5 to All on Tue Jul 13 20:18:04 2021
    Yup. Minus the faux-leather covers. Al seems not to favor the covers.

    Given the volume he scans, if it won't go through the sheet feeder,
    I suspect they sometimes get skipped. :)

    De

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Bob Allison@21:1/5 to Dennis Boone on Sun Jan 22 22:55:37 2023
    On Sunday, July 11, 2021 at 9:24:39 PM UTC-4, Dennis Boone wrote:
    Were the P300 and P400 ECL machines? I hadn't thought so. I have the
    P300 microcode manual, and we once had the microcode boards for the P400
    at Greater Clark County Schools. I remember looking at them, and I'm
    pretty sure that's what I recall was used.
    No, pretty sure the first ECL machine was the 9950.
    I have the data sheets for the 54181/74181 ALU in my 1976 (second
    printing) Texas Instruments TTL Data Book. if they're not already
    online somewhere, let me know, I'll scan the pages.
    Check http://bitsavers.org/components/ti/_dataBooks/ - I suspect
    it's in there.

    De

    Yes, the 9950 and all 9000s were ECL with the 2900 family of bit-slice chips. There were 6 boards in the CPU: Execute Units E1 & E2, Instruction I-Unit, S-Unit (Cache), CS (Control Store), and MC (Memory Controller).

    The 6000 series was ECL with custom Prime-designed Motorola MCA 2500 ECL VLSI for the Execution Unit, Instruction/Cache Unit, Control Store, and Memory Controller with multiple designs on each of the 4 boards that made up the CPU.

    The next ECL CPU (The Lion project) was to be with the MCA 25000 VLSI but that project was canceled after the takeover attempt and the buyout by JH Whitney and DR Holdings.

    The 5000 series was with custom Prime-designed CMOS from LSI Logic. That project survived the takeover and the 5370 dual CPU was the last machine to ship.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Bob Allison@21:1/5 to Dennis Boone on Sun Jan 22 22:50:15 2023
    On Sunday, July 11, 2021 at 9:24:39 PM UTC-4, Dennis Boone wrote:
    Were the P300 and P400 ECL machines? I hadn't thought so. I have the
    P300 microcode manual, and we once had the microcode boards for the P400
    at Greater Clark County Schools. I remember looking at them, and I'm
    pretty sure that's what I recall was used.
    No, pretty sure the first ECL machine was the 9950.
    I have the data sheets for the 54181/74181 ALU in my 1976 (second
    printing) Texas Instruments TTL Data Book. if they're not already
    online somewhere, let me know, I'll scan the pages.
    Check http://bitsavers.org/components/ti/_dataBooks/ - I suspect
    it's in there.

    De

    Yes, the 9950 and all 9000s were ECL with the 2900 family of bit-slice chips. There were 6 boards in the CPU: Execute Units E1 & E2, Instruction I-Unit, S-Unit (Cache), CS (Control Store), and MC (Memory Controller).

    The 6000 series was ECL with custom Prime-designed Motorola MCA 2500 ECL VLSI for the Execution Unit, Instruction/Cache Unit, Control Store, and Memory Controller with multiple designs on each of the 4 boards that made up the CPU.

    The next ECL CPU (The Lion project) was to be with the MCA 25000 VLSI but that project was canceled after the takeover attempt and the buyout by JH Whitney and DR Holdings.

    The 5000 series was with custom Prime-designed CMOS from LSI Logic. That project survived the takeover and the 5370 dual CPU was the last machine to ship.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Bob Allison@21:1/5 to Dennis Boone on Sun Jan 22 23:05:06 2023
    On Sunday, July 11, 2021 at 9:24:39 PM UTC-4, Dennis Boone wrote:
    Were the P300 and P400 ECL machines? I hadn't thought so. I have the
    P300 microcode manual, and we once had the microcode boards for the P400 at Greater Clark County Schools. I remember looking at them, and I'm pretty sure that's what I recall was used.
    No, pretty sure the first ECL machine was the 9950.
    I have the data sheets for the 54181/74181 ALU in my 1976 (second printing) Texas Instruments TTL Data Book. if they're not already
    online somewhere, let me know, I'll scan the pages.
    Check http://bitsavers.org/components/ti/_dataBooks/ - I suspect
    it's in there.

    De

    Yes, the 9950 and all 9000s were ECL with the 2900 family of bit-slice chips. There were 6 boards in the CPU: Execute Units E1 & E2, Instruction I-Unit, S-Unit (Cache), CS (Control Store), and MC (Memory Controller).

    The 6000 series was ECL with custom Prime-designed Motorola MCA 2500 ECL VLSI. There were 6 boards in the CPU: the Execution Unit, Instruction/Cache Unit, Control Store, and Memory Controller with multiple VLSI designs on each of the 4 boards that made up the CPU.

    The next ECL CPU (The Lion project) was to be with the MCA 25000 VLSI, but that project was canceled after the takeover attempt and the buyout by JH Whitney and DR Holdings.

    The 5000 series was with custom Prime-designed CMOS from LSI Logic. The single processor variations had multiple chips for the complete CPU on one board plus slots for memory modules on the same board. The 5370 dual processor version had two sets of
    CPU chips and a new Memory Controller all on one board with a second board for all the memory modules. Those projects survived the takeover and the 5370 dual CPU was the last machine designed and shipped from Prime.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Daiyu Hurst@21:1/5 to Bob Allison on Tue Jun 27 11:41:23 2023
    On Monday, January 23, 2023 at 2:05:07 AM UTC-5, Bob Allison wrote:
    On Sunday, July 11, 2021 at 9:24:39 PM UTC-4, Dennis Boone wrote:
    Were the P300 and P400 ECL machines? I hadn't thought so. I have the P300 microcode manual, and we once had the microcode boards for the P400 at Greater Clark County Schools. I remember looking at them, and I'm pretty sure that's what I recall was used.
    No, pretty sure the first ECL machine was the 9950.
    I have the data sheets for the 54181/74181 ALU in my 1976 (second printing) Texas Instruments TTL Data Book. if they're not already
    online somewhere, let me know, I'll scan the pages.
    Check http://bitsavers.org/components/ti/_dataBooks/ - I suspect
    it's in there.

    De
    Yes, the 9950 and all 9000s were ECL with the 2900 family of bit-slice chips.
    There were 6 boards in the CPU: Execute Units E1 & E2, Instruction I-Unit, S-Unit (Cache), CS (Control Store), and MC (Memory Controller).

    The 6000 series was ECL with custom Prime-designed Motorola MCA 2500 ECL VLSI.
    There were 6 boards in the CPU: the Execution Unit, Instruction/Cache Unit, Control Store, and Memory Controller with multiple VLSI designs on each of the 4 boards that made up the CPU.

    The next ECL CPU (The Lion project) was to be with the MCA 25000 VLSI, but that project was canceled after the takeover attempt and the buyout by JH Whitney and DR Holdings.

    The 5000 series was with custom Prime-designed CMOS from LSI Logic. The single processor variations had multiple chips for the complete CPU on one board plus slots for memory modules on the same board. The 5370 dual processor version had two sets of
    CPU chips and a new Memory Controller all on one board with a second board for all the memory modules. Those projects survived the takeover and the 5370 dual CPU was the last machine designed and shipped from Prime.

    I really thought I remembered seeing the AM2900 in the microcode boards for the P400.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Dennis Boone@21:1/5 to All on Wed Jun 28 02:01:17 2023
    I really thought I remembered seeing the AM2900 in the microcode boards
    for the P400.

    If you compare pdf page 75 here -

    https://sysovl.info/pages/blobs/prime/archhw/MAN1857_microcodingCodingHbk_Sep74.pdf

    to the function table for the 74181 here -

    https://en.wikipedia.org/wiki/74181

    they seem to be the same thing.

    De

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Dennis Boone@21:1/5 to All on Wed Jun 28 02:46:15 2023
    I really thought I remembered seeing the AM2900 in the microcode boards for the P400.

    I have a board set from a 150. It's a little after the 400, of course,
    but the cpu board has several 74s181 chips on board.

    https://yagi.h-net.org/p150_cpu_74s181/

    De

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)