What are my options when handling high throughput data streams apart from kafka

Question

I have a close to 50k to 100k data being streamed realtime by multiple threads with each thread having a upper limit of 1k data per sec. The individual data itself will only