reading back

The main problems with render target readback are bandwidth, and stalling. Generally PCI-e/AGP bandwidth for reading back data is limited, so reading back large render targets can be slow. Downscaling to a smaller size can help a lot. Stalling occurs because the CPU and GPU have to synchronize, and then the GPU has to flush all commands so that the render target actually has the data you want. You can avoid the stall by triple-buffering your render target, as long as having 2 frames of latency is acceptable. 

评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值