The chips will be the same from the company, but they will have different chips, eg. AKD500, 1000, 1500, 2000 etc
However, you can modify the neural network you want to deploy to the chip in any way you want, so you can reduce power consumption to increase inference rate or accuracy etc
If you want to see comparisons to other available inference hardware, just download the ADE and run the tests yourself ?
The ADE is very easy to use.
AKIDA benchmarking?, page-9
-
- There are more pages in this discussion • 133 more messages in this thread...
You’re viewing a single post only. To view the entire thread just sign in or Join Now (FREE)
Add BRN (ASX) to my watchlist
 (20min delay)
|
|||||
|
Last
19.5¢ |
Change
-0.005(2.50%) |
Mkt cap ! $361.9M | |||
| Open | High | Low | Value | Volume |
| 20.0¢ | 20.5¢ | 19.5¢ | $878.2K | 4.371M |
Buyers (Bids)
| No. | Vol. | Price($) |
|---|---|---|
| 64 | 1444914 | 19.5¢ |
Sellers (Offers)
| Price($) | Vol. | No. |
|---|---|---|
| 20.0¢ | 388739 | 9 |
View Market Depth
| Last trade - 16.10pm 15/07/2024 (20 minute delay) ? |
| BRN (ASX) Chart |