Arm Community
Site
Search
User
Site
Search
User
Arm Developer
Documentation
Learning Paths
On-Demand Videos
Groups
Arm Ambassadors
Education Hub
Open Source Software and Platforms
Research Collaboration and Enablement
Forums
AI forum
Architectures and Processors forum
Arm Development Platforms forum
Arm Development Studio forum
Automotive forum
Compilers and Libraries forum
Embedded and Microcontrollers forum
Internet of Things (IoT) forum
Keil forum
Laptops and Desktops forum
Mobile, Graphics, and Gaming forum
Morello forum
Operating Systems forum
Servers and Cloud Computing forum
SoC Design and Simulation forum
SystemReady Forum
Blogs
AI blog
Announcements
Architectures and Processors blog
Automotive blog
Embedded and Microcontrollers blog
Internet of Things (IoT) blog
Laptops and Desktops blog
Mobile, Graphics, and Gaming blog
Operating Systems blog
Servers and Cloud Computing blog
SoC Design and Simulation blog
Tools, Software and IDEs blog
Support
Arm Support Services
Documentation
Downloads
Training
Arm Approved program
Arm Design Reviews
Community Help
More
Cancel
Support forums
Mobile, Graphics, and Gaming forum
Mali 628 floating point precision
Jump...
Cancel
Locked
Locked
Replies
4 replies
Subscribers
136 subscribers
Views
6836 views
Users
0 members are here
Mali-T628
Mali-GPU
Options
Share
More actions
Cancel
Related
How was your experience today?
This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion
Mali 628 floating point precision
Maxim Mogilnitsky
over 11 years ago
Note: This was originally posted on 17th August 2013 at
http://forums.arm.com
Could you please elaborate on a floating point precision of Mali-628?
What part of the IEEE754-2008 is supported in GPU? I.e. how many bits of precision we can get (medump) ?
Thank you in advance
Maxim
http://blogs.arm.com/multimedia/318-arm-mali-t604-new-gpu-architecture-for-highest-performance-flexibility/
Peter Harris
over 11 years ago
Note: This was originally posted on 22nd August 2013 at
http://forums.arm.com
I'd suggest sticking to the Khronos precision minimums when picking qualifiers; it is very easy to write code which happens to work on one architecture which only works because that implementation provides more bits than technically necessary, which then fails on another device in weirdly wonderful and hard to debug ways.
FWIW on all of the Mali GPUs:
highp is fp32
mediump is fp16
lowp is fp16
Cancel
Vote up
0
Vote down
Cancel
Maxim Mogilnitsky
over 11 years ago
Note: This was originally posted on 26th August 2013 at
http://forums.arm.com
Thank you for the answer and for advice.
Just to be sure let me dig a bit into the question.
I understand that fp32 is a required and supported precision for the shaders.
#1 Just to make it clear: Is it for both vertex and fragment shader?
Let me also ask a question about HW itself:
#2 Is fp32 means that the GPU HW actually uses at least 32 bits of floating point precision throughout an entire GPU pipe?
Sorry for perhaps "blowing on water", but I got some "burns from milk".
Best Wishes
Maxim
Cancel
Vote up
0
Vote down
Cancel
Pete
over 11 years ago
Note: This was originally posted on 30th August 2013 at
http://forums.arm.com
Hi Maxim,
1) The Mali-T600 series GPUs have a unified architecture, so there is no distinction between the vertex shader and fragment shader - they run on the same execution units.
2) Yes, I believe so.
Hope that helps, Pete
Cancel
Vote up
0
Vote down
Cancel
Maxim Mogilnitsky
over 11 years ago
Note: This was originally posted on 30th August 2013 at
http://forums.arm.com
Many Thanks
Cancel
Vote up
0
Vote down
Cancel