summaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorKonstantin <const@const.me>2023-02-08 14:18:30 +0100
committerKonstantin <const@const.me>2023-02-08 14:18:30 +0100
commitfd6eb55d602f1020bc9f89a4322db3644615ec22 (patch)
tree2c4c2fe7e5abf9e25abbf7d497a47591fa6f4fa5
parent1da6f30fadaf1ebca3730bcc3f9f05d43e326c6b (diff)
Readme
-rw-r--r--Readme.md4
1 files changed, 1 insertions, 3 deletions
diff --git a/Readme.md b/Readme.md
index 8da9d30..d023716 100644
--- a/Readme.md
+++ b/Readme.md
@@ -101,9 +101,7 @@ I don’t know how to measure that, but I have a feeling the bottleneck is memor
Someone on Hacker News [has tested](https://news.ycombinator.com/item?id=34408429) on [3060Ti](https://en.wikipedia.org/wiki/GeForce_30_series#Desktop),
the version with GDDR6 memory.
Compared to 1080Ti, that GPU has 1.3x FP32 FLOPS, but 0.92x VRAM bandwidth.
-The app was about 10% slower on the 3060Ti.<br/>
-Would be interesting to test on AMD [RDNA2](https://en.wikipedia.org/wiki/Radeon_RX_6000_series#Desktop)
-or [RDNA3](https://en.wikipedia.org/wiki/Radeon_RX_7000_series#Desktop) GPUs, with their large 80 MB+ level 3 cashes.
+The app was about 10% slower on the 3060Ti.
## Further Optimisations