Arm Community
Site
Search
User
Site
Search
User
Groups
Arm Research
DesignStart
Education Hub
Graphics and Gaming
High Performance Computing
Innovation
Multimedia
Open Source Software and Platforms
Physical
Processors
Security
System
Software Tools
TrustZone for Armv8-M
中文社区
Blog
Announcements
Artificial Intelligence
Automotive
Healthcare
HPC
Infrastructure
Innovation
Internet of Things
Machine Learning
Mobile
Smart Homes
Wearables
Forums
All developer forums
IP Product forums
Tool & Software forums
Support
Open a support case
Documentation
Downloads
Training
Arm Approved program
Arm Design Reviews
Community Help
More
Cancel
Developer Community
Tools and Software
Software Tools
Jump...
Cancel
Software Tools
Arm Development Studio forum
Any equivalent NEON instruction to SMULWy?
Tools, Software and IDEs blog
Forums
Videos & Files
Help
Jump...
Cancel
New
Replies
2 replies
Subscribers
126 subscribers
Views
1422 views
Users
0 members are here
Related
Any equivalent NEON instruction to SMULWy?
Offline
Jake Lee
over 7 years ago
Note: This was originally posted on 6th July 2013 at http://forums.arm.com
Hi everybody,
I'm currently working on 7x7 gaussian blur filter for NEON.
And since everything bigger than 3x3 is hard to handle with 2D algorithms, I made it to 2*1D algorithm.
Running the filter horizontally first, every pixel (y value) is temporarily stored in 16bit uq8. So far so good. It's super fast with zero latency and dual issue everywhere possible.
The problem begins when doing it vertically.
I know WIDE model isn't available for mul instructions. It's no problem widening the coefficients to 16bit for 16bit*16bit multiplications. With VMULL.u16, the result is then 32bit so I have to do narrowing twice in order to get the final result in 8bit, and I really don't like it.
I read through the assembly reference several times, but there seems to be no mul instruction giving the upper 16bit as the result. Am I right on this?
Do I really have to accept having to do narrowing twice?
I badly need something like SMULWy....
VQDMULH, which gives the upper half as result won't do the trick since it works only with signed values and doubles the result, if I understood correctly.
I'm really curious :
What is VQDMULH good for?
I can hardly imagine anything where that doubling might be useful. Can someone enlighten me?
Thanks in advance
More questions in this forum
By title
By date
By reply count
By view count
By most asked
By votes
By quality
Descending
Ascending
All recent questions
Unread questions
Questions you've participated in
Questions you've asked
Unanswered questions
Answered questions
Questions with suggested answers
Questions with no replies
Answered
Failed to read contents of Internal RAM L1-I_DATA in ARM DS
0
Arm Development Studio
Cache
Debug and Trace Services Layer (DTSL)
13529
views
23
replies
Latest
4 months ago
by
Boon Khai
Answered
DS-5 connect fail when cortex-r5 is in lock-step mode
+1
11225
views
10
replies
Latest
5 months ago
by
Stuart Hirons
Answered
On Cortex-M4F microcontrollers: is fixed point math faster or floating point?
+1
10779
views
10
replies
Latest
5 months ago
by
Ronan Synnott
Answered
Debugging kernel: OS support not working for Linux 5.4
+1
Kernel Developers
External Hardware Debug
Debugger
9649
views
5
replies
Latest
5 months ago
by
sgoldschmidt
Answered
DS-5 bare metal wait error after run "debug"
+1
DS-5 Development Studio
Debugging
Arm Compiler 5
Memory
32547
views
14
replies
Latest
5 months ago
by
prasadghole
<
>
View all questions in Arm Development Studio forum