editor's blog
Subscribe Now

Chip Design Tweaker

Last-minute chip design changes are always unfortunate, whether right before cutting masks or, worse yet, after you get silicon back. Some major tool environments provide engineering change order (ECO) support, some don’t. But it’s always a less-than-perfect scenario: an ideal top-down flow would maintain the chain of refinement from the most abstract representation down to the final details. Making a change only at the low level breaks that.

But the practical fact is that, if you’ve spent weeks and months getting things just the way you want them – with the exception of those annoying issues you just found – you don’t want to risk undoing all that careful work by taking a trip back to the abstract level and repeating the flow.

So you grit your teeth and push a couple polygons or change some transistor characteristics at the lowest level and chalk it up to “you gotta do what you gotta do.”

A small company called Dorado is trying to help out with these last minute ECO tweaks with a slightly unfortunately-named tool called Tweaker. (I suppose it can stay up for days in a row, but hopefully it doesn’t suffer from unpredictability or a nasty crash after a few days…)

Tweaker is set up to automatically fix a number of issues automatically and more in a manually-guided manner. Its scope includes functional, timing, and power tweaks. Changes may be made at the RTL level, where they attempt to minimize the scope of any resultant changes, or at the physical level.

In my discussion with them at DAC, it seems that low-level physical tweaks can’t be automatically preserved in the face of, say, a synthesis change. For example, if you make a low-level change and then have to make a functional (RTL) tweak that affects the same cells, you’ll have to redo the low-level changes after the RTL tweak re-synthesizes. But a TCL script can be used, which should save time (assuming no name changes, etc.).

It can work pre-mask by trying to preserve the chip area or post-silicon by taking advantage of spare cells.

They’ve managed to convince TSMC of its value…

An important note, however: if you want to convince your boss to make this tool available for you, make sure to say, “I need Tweaker to make some quick design changes.” Not, “I need a tweaker to make some quick design changes.” Big difference.

14 thoughts on “Chip Design Tweaker”

  1. Pingback: read what he said
  2. Pingback: Togel Shenzen
  3. Pingback: DMPK
  4. Pingback: zdporn.com
  5. Pingback: pezevenk
  6. Pingback: kari satilir
  7. Pingback: hash
  8. Pingback: bandar judi
  9. Pingback: domino online
  10. Pingback: Cheap

Leave a Reply

featured blogs
Jun 15, 2021
Samtec Flyover® Twinax Cable assemblies allow designers to extend signal reach and density, enabling 112 Gbps PAM4 performance. Samtec Flyover systems are commonly used in mid-board applications, with a cable assembly connector located next to the chip. The signal path ...
Jun 15, 2021
We share key automotive cybersecurity considerations for connected vehicle technology such as automotive WiFi & Bluetooth, along with NHTSA best practices. The post Closing the 'Door' on Remote Attackers by Securing Wireless Paths into Vehicles appeared first on From Si...
Jun 15, 2021
At the recent TSMC 2021 Online Technology Symposium, the keynote to open the show was delivered by Dr C. C. Wei, TSMC's CEO. In addition to C.C. himself, there were guest appearances by Lisa Su,... [[ Click on the title to access the full blog on the Cadence Community s...
Jun 14, 2021
By John Ferguson, Omar ElSewefy, Nermeen Hossam, Basma Serry We're all fascinated by light. Light… The post Shining a light on silicon photonics verification appeared first on Design with Calibre....

featured video

Kyocera Super Resolution Printer with ARC EV Vision IP

Sponsored by Synopsys

See the amazing image processing features that Kyocera’s TASKalfa 3554ci brings to their customers.

Click here for more information about DesignWare ARC EV Processors for Embedded Vision

Featured Paper

An FPGA-Based Solution for a Graph Neural Network Accelerator

Sponsored by Achronix

Graph Neural Networks (GNN) drive high demand for compute and memory performance and a software only based implementation of a GNN does not meet performance targets. As a result, there is an urgent need for hardware-based GNN acceleration. While traditional convolutional neural network (CNN) hardware acceleration has many solutions, the hardware acceleration of GNN has not been fully discussed and researched. This white paper reviews the latest GNN algorithms, the current status of acceleration technology research, and discusses FPGA-based GNN acceleration technology.

Click to read more

featured chalk talk

Accelerating Innovation at the Edge with Xilinx Adaptive System on Modules

Sponsored by Xilinx

The combination of system-on-module technology with advanced SoCs with programmable logic offer the ultimate in functionality, performance, flexibility, power efficiency, and ease of use. In this episode of Chalk Talk, Amelia Dalton chats with Karan Kantharia of Xilinx about the new Kira SOM, and how it enables faster time-to-deployment versus conventional component-based design.

Click here for more information about Kria Adaptive System-on-Modules