The last hop is the one you can control
We receive shreds at the network edge in Frankfurt and forward them. How quickly they reach you after that depends on where you are, and that part is not something we can fix from our side. Across the public internet we add latency neither of us controls. In the same building, it is a cross connect.
That is the honest version of what proximity buys. Not a faster feed: the same feed, with less of your own network between it and your process.
- Bots, single strategies, stagingVirtual serversShared tenancy on the same network as the relay. The cheapest way to remove the public internet from your path.
- Production desksDedicated virtual serversDedicated cores and NVMe with no noisy neighbours, so your p99 stops depending on somebody else's workload.
- High frequency and market makingBare metalA whole machine, your kernel tuning, your NIC choice, and a cross connect to the feed rather than a network path.
Why there are no prices here
Because a configuration is built to a workload, and because hardware pricing moves faster than a web page does. A stale price list would waste your time and ours. Tell us what you need to run and we will send the current catalogue, which is a shorter conversation than it sounds.
Email sam@shredstream.shUseful things to include: what you are running, how much memory and storage it wants, whether you need a dedicated NIC, and whether you are already taking the feed from us.