Having said that, we have 20-odd proof of concepts going on, right? And the revenue we talk about somewhere between $10 to $20 million in 2025 is driven by enterprise storage today and some royalty software license revenues as well because we also have sort of a a paired offering using our superior algorithms for better compression storage in the traditional systems. And then we have these compute servers that are being - actually, they're being demonstrated, right? We had the Atlanta Supercomputer Conference where we demonstrated with QCT, which is a major compute server manufacturer. There they use storage compression. But the key point is why do they need it? it's not just about compression and security for reducing the amount of storage space. It's actually a very, very powerful enabling capability where we reduce the latency of compute by really having a very, very low latency storage with very, very high density compression. So what that does is when you think about AI-type applications, whether it is inside the data center or outside the data center, it becomes very critical not to have bottlenecks in the, whether it's a training or inference systems. So I think at this point, to be honest with you, our revenues are coming from enterprise, which is basically storage appliances and compute servers. However, we have a lot of strong traction after our reference platform announcement with AMD, a joint development platform, So we're getting a lot of traction in the other no

Apr 23, 2025 / CEO / receipt, needs a key / Kishore Seendripu, CEO | earnings call 2025-04-23 | transcript 42eff6bc-5604-4f7d-abf8-350ea8afa947 | chars 25449-26949 | segment 26