What's new

MERL - UIT paving the way to make Pakistan's first indigenous processor

.
That is excellent news...You may want to put this video and may be a brief explanation on the "Middle East Cutting edge tech" Thread in the middle east forum... I started this thread but and it will be great if you are a second nation to add info like the above to it..you can up-date it as it progresses,..
 
.
That is excellent news...You may want to put this video and may be a brief explanation on the "Middle East Cutting edge tech" Thread in the middle east forum... I started this thread but and it will be great if you are a second nation to add info like the above to it..you can up-date it as it progresses,..
To be honest, I am not an expert in the field to do a write up. But I will copy and post from their own website.
Ghazi System on a Chip
An SoC (System on a Chip) design for Google sponsored Open MPW shuttles for SKY130. The processor core is the 3-stage version of the Buraq Core RV32IMC. The hardware implementation incorporates options such as IRQ, Multiply, Divide and the compressed (16 bit) ISA for embedded applications. The SoC has peripherals such as GPIO, UART, a platform level interrupt controller (PLIC) as well as a timer and a debug module all connected using the Tilelink Interconnect and is going to be fabricated using a 130nm process in collaboration with Efabless and SkyWater which will be funded by Google.
FEATURES:
  • Support for M extension with a single cycle "Fast" multiplier
  • Separate instruction and data memories
  • TileLink Un-Cached Lightweight (TL-UL) Bus Protocol
  • 32 GPIO with configurable interrupts and option for masked writing
  • 2 pin full duplex UART • RISC-V compliant interrupt controller
  • 64-bit timer with 12-bit prescaler and 8-bit step register
  • JTAG Test Access Port (TAP) for debug

ghz_design.jpg

Ibtida System on a Chip
Ibtida - ابتدا means "The Beginning", this System on a Chip (SoC) is the start of many RISC-V based SoCs to come. It is the first CHISEL-based chip to be taped out from Pakistan, and has been designed by Muhammad Hadir Khan, Sajjad Ahmed, and Usman Zain; engineering graduate and undergraduate students respectively. The physical layout of the design is achieved by Aireen Amir Jalal, who is also an engineering graduate. Ibtida is a simple SoC, with GPIO as a peripheral, external instruction and data memories, connected with the TileLink interconnect. It is built around RISC-V based 5 stage pipelined core Buraq-Mini, all developed from scratch using CHISEL HCL.
FEATURES:
  • RV32IM extension support.
  • 5 stage pipelined core.
  • Separate instruction and data memories (each 256 Bytes).
  • TileLink Un-Cached Lightweight (TL-UL) Bus Protocol.
  • GPIO peripheral with 30 I/Os connected to the I/O pads.
ibt_design.jpg
 
Last edited:
.
To be honest, I am not an expert in the field to do a write up. But I will copy and post from their own website.
Ghazi System on a Chip
An SoC (System on a Chip) design for Google sponsored Open MPW shuttles for SKY130. The processor core is the 3-stage version of the Buraq Core RV32IMC. The hardware implementation incorporates options such as IRQ, Multiply, Divide and the compressed (16 bit) ISA for embedded applications. The SoC has peripherals such as GPIO, UART, a platform level interrupt controller (PLIC) as well as a timer and a debug module all connected using the Tilelink Interconnect and is going to be fabricated using a 130nm process in collaboration with Efabless and SkyWater which will be funded by Google.
FEATURES:
  • Support for M extension with a single cycle "Fast" multiplier
  • Separate instruction and data memories
  • TileLink Un-Cached Lightweight (TL-UL) Bus Protocol
  • 32 GPIO with configurable interrupts and option for masked writing
  • 2 pin full duplex UART • RISC-V compliant interrupt controller
  • 64-bit timer with 12-bit prescaler and 8-bit step register
  • JTAG Test Access Port (TAP) for debug

ghz_design.jpg

Ibtida System on a Chip
Ibtida - ابتدا means "The Beginning", this System on a Chip (SoC) is the start of many RISC-V based SoCs to come. It is the first CHISEL-based chip to be taped out from Pakistan, and has been designed by Muhammad Hadir Khan, Sajjad Ahmed, and Usman Zain; engineering graduate and undergraduate students respectively. The physical layout of the design is achieved by Aireen Amir Jalal, who is also an engineering graduate. Ibtida is a simple SoC, with GPIO as a peripheral, external instruction and data memories, connected with the TileLink interconnect. It is built around RISC-V based 5 stage pipelined core Buraq-Mini, all developed from scratch using CHISEL HCL.
FEATURES:
  • RV32IM extension support.
  • 5 stage pipelined core.
  • Separate instruction and data memories (each 256 Bytes).
  • TileLink Un-Cached Lightweight (TL-UL) Bus Protocol.
  • GPIO peripheral with 30 I/Os connected to the I/O pads.
ibt_design.jpg
That is great..
 
. . . . . .

Sorry but you cannot call this SoC processor as "indigenous" because simply it is just an implementation of the international ( originally American ) project RISC-V. This is similar to someone coming out with yet another distro of Linux.

Such "indigenous" processors have been brought out by China and India as well. So no great shakes really.

It would have been wonderful if the Ibtida processor ( meaning "The Beginning" ) mentioned above had actually been a new ISA / architecture. Something from Pakistan to the world.
 
Last edited:
.
Sorry but you cannot call this SoC processor as "indigenous" because simply it is just an implementation of the international ( originally American ) project RISC-V. This is similar to someone coming out with yet another distro of Linux.

Such "indigenous" processors have been brought out by China and India as well. So no great shakes really.

It would have been wonderful if the Ibtida processor ( meaning "The Beginning" ) mentioned above had actually been a new ISA / architecture. Something from Pakistan to the world.
No one will do independent ISA. Even China is doing x86, ARM and MIPS spin-offs. As long we know exactly what’s baked into it, it can serve as an important pillar in sensitive applications and also for small applications of local appliance industry for IOT.
 
.
No one will do independent ISA. Even China is doing x86, ARM and MIPS spin-offs. As long we know exactly what’s baked into it, it can serve as an important pillar in sensitive applications and also for small applications of local appliance industry for IOT.

1. Well, I have been designing an independent ISA and an OS to go with it, for the last many years. The processor arch. is a simplified and clock-less one. The OS is based on microkernel architecture. Took me years to simplify them. I want to start a company to commercially market them, especially them used within a wearable computer. The processor and the OS being open source and the wearable being priced.

2. Yes, China has been working with x86, MIPS, ARM and probably SPARC simply because it did not spend the intellectual time in realizing its own architecture. Day before yesterday one Indian member posted an article which detailed a finding that the Chinese company, Huawei, saying that its HarmonyOS was a new thing but it turned out to be simply a copy-paste thing of Android.
 
Last edited:
.
1. Well, I have been designing an independent ISA and an OS to go with it, for the last many years. The processor arch. is a simplified and clock-less one. The OS is based on microkernel architecture. Took me years to simplify them. I want to start a company to commercially market them, especially them used within a wearable computer.

2. Yes, China has been working with x86, MIPS, ARM and probably SPARC simply because it did not spend the intellectual time in realizing its own architecture. Day before yesterday one Indian member posted an article which detailed a finding that the Chinese company, Huawei, saying that its HarmonyOS was a new thing but it turned out to be simply a copy-paste thing of Android.
It’s not just about spending time. It’s about the developing the whole stack and the eco-system around it and then having it accepted and adopted by the wider community. It would be much easy to achieve commercial success based on a widely accepted ISA with a large development community around it. Having a lab project is one thing but for a country like us, the resources would be better spent on something that would be both technically and commercially viable.

Also it’s open source. So no royalties 🙂
 
.
It’s not just about spending time. It’s about the developing the whole stack and the eco-system around it. It would be much easy to achieve commercial success based on a widely accepted ISA with a large development community around it.

I agree that the ecosystem is also important so I will tell you what I am going to do. For my OS I will write an emulator for the IBM PC architecture just like the Bochs emulator ( webpage ). The emulator will allow running of current x86 OS' like Windows and Linux, albeit probably slowly, but the ecosystem is thus enabled on my new platform.

Having a lab project is one thing but for a country like us, the resources would be better spent on something that would be both technically and commercially viable.

TBH Pakistan did not spend monetary, material and human resources well. Take your space agency SUPARCO. Instead of dreaming of Pakistanis in space SUPARCO was guided towards military applications by a succession of chiefs who were military generals.

About thirty years ago, Pakistan should have been the leader of a consortium of front-line Muslim-majority countries like Iraq, Syria, Libya, Algeria, Indonesia, Nigeria, Turkey, Iran, even Afghanistan etc whose aim was to build a space station similar to the Soviet Mir. I am not doing vague talk. If a feudal, monarchic country like Russia ( and the rest of the USSR ) could become an industrialized one in one generation and send a human to space in 1961 it is shameful that 50 Muslim-majority countries in the world are not able to send a man to space in 2021. Though it appears that Iran is trying to do that by what 2025 ? If the much sanctioned Iran can attempt this why can't Pakistan ?
 
Last edited:
.
it is shameful that 50 Muslim-majority countries in the world are not able to send a man to space in 2021.

most of these muslim countries arent exactly democracies. more money is spent into keeping it that way instead of things like going to space. the West wants it that way because it scared of political islam.

If the much sanctioned Iran can attempt this why can't Pakistan ?

we have bigger issues than space. but Iran can do it because it is a dictatorship and the sanctions force them to be creative and do creative endeavours like space because it helps in many other industries. Pakistan isnt a dictatorship, its a democracy where the corrupt take their turns to steal.
 
.

Country Latest Posts

Back
Top Bottom