gui (@guisalberto) header
gui (@guisalberto) profile picture

ceo @latitudesh the global bare metal platform

Sao Paulo
Joined October 2011
Photo by gui (@guisalberto)Photo by gui (@guisalberto)Photo by gui (@guisalberto)Photo by gui (@guisalberto)Photo by gui (@guisalberto)Photo by gui (@guisalberto)Photo by gui (@guisalberto)Photo by gui (@guisalberto)Photo by gui (@guisalberto)Photo by gui (@guisalberto)
Cloud-like experience while deploying and managing physical servers ,@latitudesh offers you access to globally distributed, dedicated enterprise-grade hardware Bare metal, billed hourly, with no virtualization Create an account and get access to a server in 15 seconds 馃У
19
14
55
15,480
Watch @latitudesh most recent deployment of a single giant computer capable of delivering trillions of tokens a year
Welcome to the AI Token Factory! 馃彮馃殌 Ever wondered what $130M of AI infrastructure looks like? 1,024 NVIDIA B300 GPUs. 40,000 strands of fiber. Nearly 2 petabits per second of aggregate network capacity. And the potential to generate over 6 million tokens per second!
1
7
372
The reason @latitudesh doesn't go offline in multiple regions simultaneously is that the entire network features independent internet uplinks at each location, and no region relies on another to stay up For instance, if you have computing resources in NYC and Miami, a failure in one region will never affect the other
1
10
1,002
Each site runs its own transit blend and its own edge. No global route reflectors pushing state between regions. A bad route in one market dies in that market.
1
113
27% of all staked SOL was sitting on a single ASN. 94% of it went dark at the same time. Not a Solana problem. An infrastructure concentration problem.
1
7
337
One BGP config change in Miami took down validator infra in London, Amsterdam, Frankfurt, Singapore and Tokyo. Solana got within 4.5pp of a finality halt. This is why ASN diversity matters more than datacenter diversity
1
4
308
Most Ethereum teams upgraded to cloud VMs and called it a day. But VMs come with hidden costs nobody talks about. Victor Chiea from @latitudesh is on the Redford Stage right now breaking down what Ethereum workloads actually need 馃敡 馃搷 Redford Stage | 12:50 - 13:10
4
21
1,287
Means so much to have such great brands and products backing our benchmarks. we have lots more to benchmark. not just sandboxes. 馃槑
Replying to @computesdk
We're thrilled to welcome @latitudesh, @TigrisData, and @browserbase as our newest sponsors, joining @archildata in supporting our benchmarking efforts.
1
1
8
882
Proud to be backing @computesdk Sandbox Benchmarks! Check out the latest results:
馃毃馃毃馃毃 Benchmark Friday is back and y'all have been 馃殺馃殺馃殺馃殺
2
2
1
5
1,214
anyone not able to use @OpenAI deep research or it's just me?
3
2
409
2026 datacenter capacity fully spoken for by the end of this quarter Next phase is deploying hardware on that capacity and we will see a shortage of CPU by Q2
1
8
731
Claude Code is again unusable/offline. Needs better infra asap
2
1
7
1,345
AI is eating all the memory I have bought this hardware for @latitudesh a year ago for ~$25k each unit today, the memories alone cost $50k
1
6
447
sandbox competition is hot interesting that providers ranked at the bottom are all at @awscloud
馃毃 we have new sandbox providers 馃毃 last run: February 25, 2026
4
442
gui retweeted
If you do not direct your attention, it will be directed for you.
844
2,421
180
19,922
698,630
What's the best AI tool / model for marketing?
2
1
253
I am also interested. Who else?
going to benchmark sandboxes i have - cloudflare - vercel - daytona - e2b.dev - exe.dev - sprites anyone else i should try
1
276
can relate, I've seen unprecedented demand for CPUs on @latitudesh
2024 was gpus 2025 was rams 2026 will be cpus Just had a call where a customer asked: "Can I spin up 5,000 sandboxes per second? And run 50,000 concurrently? Actually, ideally 500,000?" Every second, every day, for two weeks. Multiple frontier companies have also asked us for 500k+ concurrent sandboxes for RL training. The compute demand is so staggering that CPUs will soon become the next bottleneck.
3
441
Is it only me or @AnthropicAI API is down
1
142
Memory shortages are driving price increases across the cloud market, either through higher rates or new setup fees. @latitudesh charges $0 setup fees on all SKUs. We secured thousands of servers ahead of vendor price hikes and will hold pricing while inventory lasts.
6
176