Originally posted by tuxd3v
View Post
The sample of 3A4000 is already being tested for a long period. And It's going to be released in October with products in my estimate.
The processing power of 3A4000 is fascinating for me. As the official announcement, it can reach 10 points/Ghz base in SPECint2006 benchmark.
But you might be disappointed since they have no plan for a 3A4000 laptop as for now. Because the design power consumption of 3A4000 can be up to 80 Watt@2Ghz, that is almost unacceptable for a portable device. As far as I know, single 3A4000 M-ATX board is already in sample stage, Multi ways 2/4/8 nodes 3A4000 systems are also in the plan.
Also, there are still some concerns about this processor. Firstly, the bridge chip it paired with, Loongson-7A1000, is pretty buggy. It has a Vivante GC1000 GPU, we tried to port open-source etnaviv driver, but soon we found that it's display controller had tons of issues that can't even display desktop correctly. Secondly, the quality of the documents is very bad. It's all in Chinese, but I can't even fully understand their words as a native Chinese speaker. Without asking the designer of the chip, it is nearly impossible to do low-level development.
The only good news is while Loongson officially stopped their steps of Kernel version on 3.10 from Redhat, Lemote is still trying to tracing the mainline kernel. We just brought up mainline kernel on that processor. But submit the patches to upstream would still be a long journey.
It's very surprising for me that there are still oversea Loongson users after Loongson-2F, wish you have fun with it : )
Leave a comment: