Never miss a story from IC Packaging and SiP. Subscribe for in-depth analysis and articles.
Last week we looked at new features largely targeting your manufacturing flow. Layer-based degassing improvements, acute angle corrections, and tools to perform a layout-vs-layout comparison of your intended manufacturing mask against the original design.
Now, let’s look at a different class of improvements: things that will save you time while giving you improved visual access to your design. From how the design looks in the canvas and how fast you can move around it to recycling a known-good memory die stack from an existing design in a new one, there are many items to be found that, I believe, will excite you!
The view you spend the most time looking at is the 2D, top-down view of your package substrate. This is displayed in the canvas area of the Allegro® Package Designer Plus editor environment.
In early versions of the tool, graphics were drawn using X11 and other protocols. This didn’t allow for advanced rendering. This is the first image in the progression below. Notice the solid yellow color for the pins, vias, and clines. The shapes are difficult to see, being drawn with a stipple pattern to allow you to find the vias that are underneath them. The fillet shapes, while present, are very difficult to make out even at this moderate zoom level.
From here, Allegro layout editors moved to an OpenGL-based canvas rendering. With that, you gained important controls like transparency. In the second image, notice you can see the green via pads on the next layer which, in the original canvas, were hidden completely by the yellow. There are increased visual cues for you, such as the via span labels showing how many layers above/below the stack extends. Net names can be seen on shapes, traces, even pins. Those visual aspects provide a huge increase in the information you have direct access to without needing to rely on commands like the show element or the hover-over data tips.
Now, there is a third level with even more advanced rendering. The far-right image shows this. A screen capture like this doesn’t do justice to the improvements. Circles are now perfectly round in the display. Even shapes see additional smoothing, with the edges blending with the rest of the shape. Gone are any graphical artifacts when you zoom in very tightly on small segments. Only when you need to edit the outline of the shape are you forced to see the outline as a brighter line font.
If you choose to ignore the quality improvements of the displayed image, you won’t be so able to ignore the performance improvements. GPU-based rendering makes use of your dedicated graphics card’s power to smooth edges and present a better display, but it similarly makes the rendering itself much faster. Try actions like panning, zooming, even turning layers on and off. Today’s complex designs may have thousands of vias, millions of degassing holes, and many thousands of pins. Turning layers on and off is now nearly instantaneous using the power of the GPU. This also lowers the burden on your main CPU, letting it work on tasks it is better suited to.
While not everything will be faster – a faster rendering of the design doesn’t mean that the auto-router will connect pins up with greater speed – the movement around the design, finding the items you seek, isolating problems, are all done with increased haste.
Try it out for yourself! If you have a compatible graphics card, I am very confident you will be thrilled with the new display!
Die stacks are an object that I think many of you reuse from one design to the next. A common configuration of dies that provide the memory for the logic components of your design, having a stacked set of these components allows for a smaller area for placement by making the package molding cap slightly taller.
However, prior to now, reusing a total stack of memory devices has been more difficult than the frequency of its application belied. That’s why, with this release, dies and die stacks have been added to the objects supported in a module reusable sub-drawing. Creating a module takes only a few clicks to select the items and save them out to an MDD. Should you have already completed the bond shell in the source drawing, include that with the components and save yourself time and effort.
With your die stack module file created, you can work inside of the module. Edit the bond shell, optimize the number of tiers of fingers, and do anything else you need to. Why work here in the MDD database? Good question!
As you make changes, even perform ECOs on the die components, you will create additional versions of the module simply by renaming it. Then, in all the designs that reference this module, you can at any time use the Placement Edit application mode’s new Replace modules command to swap in the new version.
The module design includes the definition of the die stack – thickness of the stacked components, spacers, and interposers as well as the placement status (Surface mount? Open Cavity, Closed Cavity? Your choice!). When you instantiate it into your design database, this information comes along for free.
Save yourself some time! Move to the latest hotfix release and start reusing your dies and die stacks.
The layer compare toolset – which includes the interactive layer pair comparison in the same database, a batch tool for comparing sets of layers (or entire designs), and the difference walking engine to understand the items reported – has long existed in the tool as un unsupported prototype.
The feedback we have received from the user community has shown us that the tool has reached a stage of maturity that its time is come to become a full-fledged member of the Tool menu community. As shown above, you’ll see the new location for these powerful utilities.
The batch tool (shown in the next image) is probably the most frequently exercised because it supports OS command line runs, batch configuration files for common comparisons, and the ability to compare different designs. It also gives complex filters and pre/post-processing options for comparing against external formats such as DXF or GDS files read into an Allegro Package Designer Plus database.
Do not miss out on the difference walker, though. It is possible to load different sets from multiple databases at once, allowing you to see the changes in a design as it moves from revision 1 through the current revision if you need to. Find out when a change was made with this graphical differencing engine.
One additional tidbit: if you’re performing GDS LVL with the GDS Verification flow highlighted last week, the differences, if any, found by running that tool are also added to the difference engine’s changes you can visualize. I encourage you to try both out!
This is the end of the second part of our what’s new and interesting tale. Come back next week for the third and final act! We will help you extract the maximum value out of the powerful Allegro Package Designer Plus tool and, as always, value any feedback and suggestions that you have!
Do SUBSCRIBE to be updated about upcoming blogs. If you have any topic you want us to cover first or any feedback, you can write to us at firstname.lastname@example.org.