| View previous topic :: View next topic |
| Author |
Message |
xsecret
Joined: 01 Feb 2004
Posts: 1846
Location: France
|
 Posted: Fri Aug 03, 2018 2:15 pm Post subject: [UCA] Sneak peek for 2019 ! |
 |
|
I can't resist to share some news about the very early PGA UCAS that just started to work today  It is fitted with debug connectors and some rework has been done, but the basic features worked as expected in just one week!
The PGA UCAS is a complex shield with an adapter card to be plugged on top. The adapter just includes a Socket and a couple caps. I wanted to start working on the most advanced CPU that would be supported by the UCA. The 486 is much (much) more advanced than the 8086. It uses a 32-bit bus, pipelining, wide instruction prefetch, burst, embedded L1 cache , etc.
And here it is: 
It's far from being fully usable with any x86 code and still require a lot of work on hardware and software, but it works. Right now, I tested up to 100 MHz (with 33 MHz bus). It works with the full 32-bit data bus and not a stripped down 16-/8-bit. I was also be able to get CPUID information from the CPU and to read power consumption.
I've also tested 3.3V and 5V CPU (you just have to change a jumper) and it also work with a 486DX:
That's probably the biggest milestone for the UCA because it proves that the UCA is technically able to support almost anything from the 4004 to the 486DX4 (including similar non-x86 CPU like Motorola 680xx or odd ones like the Intel 80376).
If you have any question, I'll be happy to answer
_________________
ES-Only Collector : http://www.engineering-sample.com
Universal Chip Analyzer (UCA) : https://x86.fr/uca / http://www.cpu-world.com/forum/viewtopic.php?t=34349 |
|
| Back to top |
|
 |
Glory_Cloud
Joined: 24 Jul 2010
Posts: 2942
|
| Posted: Fri Aug 03, 2018 3:18 pm Post subject: |
|
|
| NICE!!! |
|
| Back to top |
|
|
debs3759
Joined: 18 Jan 2006
Posts: 9477
Location: Northampton, Divided Kingdom
|
| Posted: Fri Aug 03, 2018 4:30 pm Post subject: |
|
|
Brilliant! I'll be wanting it when it is ready
_________________
My graphics card database can be found at http://www.gpuzoo.com.
I can resist anything except temptation.
Debs |
|
| Back to top |
|
|
Neon
Joined: 04 Feb 2008
Posts: 1512
Location: Dallas, Texas, USA
|
| Posted: Fri Aug 03, 2018 4:44 pm Post subject: |
|
|
This is good.
Now I am interested.  |
|
| Back to top |
|
|
xsecret
Joined: 01 Feb 2004
Posts: 1846
Location: France
|
|
| Back to top |
|
|
edwardauskis
Joined: 26 Nov 2014
Posts: 770
Location: Lithuania
|
| Posted: Fri Aug 03, 2018 10:15 pm Post subject: |
|
|
Great!!!! 👍👍👍👍👍👍👍👍👍👍
_________________
Regards,
Edvardas |
|
| Back to top |
|
|
CPUShack
Joined: 16 Jun 2003
Posts: 34259
Location: State of Jefferson, USA
|
|
| Back to top |
|
|
Vegeta
Joined: 13 Apr 2013
Posts: 7049
|
| Posted: Sat Aug 04, 2018 3:52 am Post subject: |
|
|
Very nice!
_________________
Wanted: 4 x VT-25-A tubes |
|
| Back to top |
|
|
Vegeta
Joined: 13 Apr 2013
Posts: 7049
|
| Posted: Sat Aug 04, 2018 1:12 pm Post subject: |
|
|
| CPUShack wrote: | | If you want to help with the 486 UCA, let Xsecret know if ya find/know of any PGA 168/169 ZIF sockets, gonna need a lot |
I sometimes see on sale CPU modules for PC-98 with one or two 168/169-pin ZIF sockets. At a price of $20-30. Maybe they should be bought and desoldered?
_________________
Wanted: 4 x VT-25-A tubes |
|
| Back to top |
|
|
debs3759
Joined: 18 Jan 2006
Posts: 9477
Location: Northampton, Divided Kingdom
|
| Posted: Sat Aug 04, 2018 1:24 pm Post subject: |
|
|
Will you be detecting the reset code (same as CPUID) on processors without the CPUID instruction?
_________________
My graphics card database can be found at http://www.gpuzoo.com.
I can resist anything except temptation.
Debs |
|
| Back to top |
|
|
molka
Joined: 17 Mar 2011
Posts: 2416
Location: Budapest, Hungary
|
| Posted: Mon Aug 06, 2018 3:20 am Post subject: |
|
|
It's incredible how powerful the FPGA is. It can emulate the 32bit environment at 33MHz for a 486?
Congrats for this nice project!
_________________
----------------------------------------------------------
My collection
http://www.x86-guide.net/molka/en/collection.html |
|
| Back to top |
|
|
H3nrik V!
Joined: 15 Apr 2014
Posts: 1246
Location: Denmark
|
| Posted: Mon Aug 06, 2018 3:43 am Post subject: |
|
|
| Will it run 50 MHz bus also? |
|
| Back to top |
|
|
xsecret
Joined: 01 Feb 2004
Posts: 1846
Location: France
|
| Posted: Tue Aug 14, 2018 12:24 pm Post subject: |
|
|
Sorry for the delay: I spent almost 10 days on the PGA UCAS, messing with timings constrains almost impossible to met inside the FPGA, but I think I've just reachead the sweet spot! The UCA is now able to execute any 8/16/32 instructions and read/write burst to RAM without any Wait State.
More interesting, I also implemented a much more advanced MCU inside the FPGA and I'm now able to communicate with the internals and read x86 memory from the USB port. It's possible to change frequency, reset the CPU and much more just by sending a command.
Here is some answers:
| debs3759 wrote: | | Will you be detecting the reset code (same as CPUID) on processors without the CPUID instruction? |
Absolutely. I save a copy of EDX (as well as EFLAGS and CPUID registers if available) to memory as soon as the reset vector is passed. These values can then be read by the internal MCU and sent outside. It works just fine right now. That will be very useful for CPU that doesn't support CPUID.
| molka wrote: | It's incredible how powerful the FPGA is. It can emulate the 32bit environment at 33MHz for a 486?
Congrats for this nice project! |
Thanks! Yes, it now works with the full 32 bit data bus. With all that HDL code, it should be possible (in a far future) to build a complete motherboard and use the FPGA as the chipset. But it would require a more advanced FPGA, because I need much more I/O (the PGA UCAS already uses 92% of the 100+ available I/O). With a bigger BGA FPGA like the Spartan-7 XC7S50 ($40), we can dream about a 486/P5 motherboard with DDR2/3/4 & PCI Express Ridiculous! 
| H3nrik V! wrote: | | Will it run 50 MHz bus also? |
I don't know, but unlikely. One week ago, I was unable to go higher than a 10 MHz bus without write memory corruption. Three days ago, I find a trick to reach 25 MHz, but with insane frequencies inside the FPGA (400 MHz). Yesterday, I finally found what's wrong with the design and tested a new address decoder. Now it works just fine with a DX2-66 or a DX4-100 at 33 MHz, and the maximum clock inside the FPGA is limited to 66 MHz.
Unfortunately, I don't have a 486 with a 40 or 50 MHz bus to try right now  But I will try to find one ASAP. Signal integrity might be an issue, especially with a 50 MHz bus.
_________________
ES-Only Collector : http://www.engineering-sample.com
Universal Chip Analyzer (UCA) : https://x86.fr/uca / http://www.cpu-world.com/forum/viewtopic.php?t=34349 |
|
| Back to top |
|
|
debs3759
Joined: 18 Jan 2006
Posts: 9477
Location: Northampton, Divided Kingdom
|
| Posted: Tue Aug 14, 2018 1:06 pm Post subject: |
|
|
| xsecret wrote: | | Unfortunately, I don't have a 486 with a 40 or 50 MHz bus to try right now But I will try to find one ASAP. Signal integrity might be an issue, especially with a 50 MHz bus. |
I can loan you an AMD A80486DX4-120SV8B for testing purposes if that helps. I'll pay shipping to you if you pay to return it when you are done testing.
_________________
My graphics card database can be found at http://www.gpuzoo.com.
I can resist anything except temptation.
Debs |
|
| Back to top |
|
|
CPUShack
Joined: 16 Jun 2003
Posts: 34259
Location: State of Jefferson, USA
|
|
| Back to top |
|
|
|
![[ Hidden ]](https://cdn.cpu-world.com/forum/templates/CPUWorld/images/lang_english/icon_email.gif "[ Hidden ]"){kind=icon}
