PVDM "We aim to saturate the market", page-89

First 89 Last

Share
Profile

Follow
kasta1

640 Posts.

121

10/06/21

11:52

Post #: 53765165
Share

Originally posted by BarrelSitter: ↑

Hi Kasta,

It's true that a 300 mm wafer has 2.25 times the area of a 200 mm wafer. However that does not really tell us how many more Akida 500 chips we will get off a wafer, because we are already using 300 mm wafers for Akida 1000. The 200 mm wafer was just one Peter had in the oven earlier for display or "conceptual" reasons.

The relevant measuremnt is the comparative area of the Akida 1000 chip and the Akida 500 chip, and that depends on the number of nodes in the Akida 500, and the number of peripherals in Akida 500 compared to Akida 1000, assuming both use the same ARM Cortex processor.

From Anil, we know there are 20 nodes in Akida 1000. We could make a guess at the number of nodes in Akida 500, but I would think it would be no more than 10, so my guess would be 8 although that's not strictly in compliance with the ratio of 500 to 1000.

If we assume 8 nodes, then the Akida 500 might be about half the size of Akida 1000. Because of the smaller area, we might have a smaller peripheral no-go zone. However, I'm guessing we could have about twice as many Akida 500 chips on a 300 mm wafer compared to the number of Akida 1000 chips on the 300 mm wafer.

On this basis, iIf the 300 mm wafer were to hold 2000 Akida 1000s, then it should hold about 4000 Akida 500s.

Of course, if there are 10 nodes in the Akida 500, then there would be significntly fewer Akida 500s, though still still significantly more than the Akida 1000s.

There is someone on HC with the name something like liesdamnedlies ... but it isn't me.

Expand

Good explanation Barrel .. and makes sense based on assumptions.
Thanks

BRN Price at posting: 56.3¢ Sentiment: Buy Disclosure: Held
3 Upvote

0 Great analysis

Reply