Geekbench6 score breakdown: integer & floating point
Hi,
In the document, I see that the overall GB6 score is computed based on 65% integer score and 35% float-point score.
However, I didn't find which benchmarks are in integer group and which ones are in floating-point.
Is there any place I can find this info?
Thanks.
Keyboard shortcuts
Generic
? | Show this help |
---|---|
ESC | Blurs the current field |
Comment Form
r | Focus the comment reply box |
---|---|
^ + ↩ | Submit the comment |
You can use Command ⌘
instead of Control ^
on Mac
1 Posted by Paul Hoon on 24 Feb, 2024 04:07 PM
Hi to the Geekbench team:
I'm building a new computer for 702,307 cancer patients based on the outstanding Geekbench depository of users single and multi scores in Geekbench 5, 6, and 6.1. Will be purchasing soon for Win 11 pro.
I'm using matrix algebra on large matrices, sometimes 90 columns by 350,00 rows, square and rectangular in sizes. The operations include, post and pre-multiplications, divisions, finding determinants, eigenvalues, and eigen vectors and building transformation matrices for rotation of multivariate axes in multi-dimentional space. I believe this is mostly, almost exclusively, floating point arithmetic/mathematics. The monitor just sits still for hours.
Which of the new individual tests in
Geekbench 6.0.1 best represent floating point mathematics? During my work, I believe the key load is on multiple CPUs (with fastest cpu cycle most desirable) , read-write to RAM, and read-write to SSD and read-write to caches second, but almost equally important.
Thank you for your time and support of cancer research.
Dr. Pete in Colorado