MikeMike #Sonic (@MikeMikecz) header
MikeMike #Sonic (@MikeMikecz) profile picture

crypto fan

Joined April 2021
Photo by MikeMike #Sonic (@MikeMikecz)Photo by MikeMike #Sonic (@MikeMikecz)Photo by MikeMike #Sonic (@MikeMikecz)Photo by MikeMike #Sonic (@MikeMikecz)Photo by MikeMike #Sonic (@MikeMikecz)Photo by MikeMike #Sonic (@MikeMikecz)Photo by MikeMike #Sonic (@MikeMikecz)Photo by MikeMike #Sonic (@MikeMikecz)Photo by MikeMike #Sonic (@MikeMikecz)Photo by MikeMike #Sonic (@MikeMikecz)
MikeMike #Sonic retweeted
As we prepare for the Tari L2 Ootle launch tonight, let’s reflect on the journey that brought us here. This is a monumental achievement for everyone who built, tested, believed, and kept showing up. This moment belongs to you. Look how far we’ve come. Imagine what we can build from here.
4
20
68
4,112
MikeMike #Sonic retweeted
The Ootle launches in 3 days! Pre-burn your XTM for TARI now so that you're ready to start Ootlin' as soon as the network goes live. You can do all of this from inside Tari Universe. One app to rule them all!
3
17
2
51
5,275
MikeMike #Sonic retweeted
Russians 🌚
42
942
29
14,097
676,514
MikeMike #Sonic retweeted
Even the best-planned AI lab can't book all its compute 18 months out. When demand outruns the contracts, it needs GPUs that are ready now. That's the gap Aethir fills: on-demand compute for training and inference, without the long wait.
Anthropic spent $5.0B on short-term compute in Q3 alone. That's the line in the new S-1 I keep coming back to. Anthropic books most of its compute on multi-year deals, 12 to 18 months before it's available. When demand runs past that plan, it buys short-term capacity, and that capacity costs significantly more. So in the quarter revenue hit $17.0B, demand ran far enough past what they'd booked that they paid up for $5.0B of extra compute. Along with a $2.0B non-cash charge, that's what swung them to a $567M operating loss after a $974M operating profit in Q2. And it's going to keep happening. They expect to use more short-term compute in the near term. The bears will call this a margin problem. To me it's the clearest demand signal in the whole filing. It also tells you where the next dollar goes. The biggest buyers of compute are already running past their long-term plans, and every provider with GPUs live today is part of the answer
3
11
1
212
63,151
MikeMike #Sonic retweeted
Prohlášení Ing. Antonína Blaníka k současné situaci.
119
1,139
103
7,151
253,289
MikeMike #Sonic retweeted
AMD Helios is shipping: 72 MI455X GPUs per rack, 6 GW from OpenAI, up to 2 GW from Anthropic, and a $1.2B Vultr order. It brings a second-source GPU to a market that has run almost entirely on NVIDIA. A second supplier adds capacity and pricing pressure, but software and networking still decide how easily workloads move. However, GPU portability decides how far lock-in loosens. For Aethir’s decentralized GPU cloud, this means routing work across a more diverse market. Read our new blog for all the details👇
8
22
2
135
11,681
MikeMike #Sonic retweeted
An accelerator retires from frontier training long before it stops being useful. Serving a mature model is a different job: a smaller memory footprint per request, real tolerance for looser interconnect, and throughput that matters more than tight synchronization. Hardware displaced by the next training generation usually has years of productive inference ahead of it. The industry talks about GPUs aging on a calendar. They age against a workload, and inference is the workload that keeps older silicon busy. Aethir coordinates a great deal of that second career.
9
12
86
2,550
MikeMike #Sonic retweeted
The blank page is the scariest part. Start with Tari’s WASM templates for tokens, NFTs, swaps, LPs, and more. Explore the code, then turn it into your own confidential concoction. Pick your starting point: github.com/tari-project/wasm…
1
11
27
3,667
MikeMike #Sonic retweeted
Tokenized stocks were a $32M niche in January 2025. Today RWA.xyz tracks $3.14B across 4M holders. Here’s what changed🧵
12
21
155
13,505
MikeMike #Sonic retweeted
🚨 SECURITY ALERT: PROTECT YOUR ASSETS & STAY SAFE 🚨 Fake airdrops, claim pages, and accounts impersonating admins are common. Before connecting your wallet or approving a transaction, check the link against the list below. Official Aethir links: - Website: aethir.com - Documentation: docs.aethir.com - Blog: blog.aethir.com - X - Main: @AethirCloud - X - Ecosystem: @AethirEco - Discord: discord.gg/aethircloud - Telegram: t.me/aethir_cloud - Checker Node Portal: app.aethir.com - User Dashboard: user.aethir.com - GPU Dashboard: dashboard.aethir.com Before connecting your wallet or signing transactions at any link, verify its safety or ask for help in our official Discord or Telegram communities.
3
12
81
2,489
MikeMike #Sonic retweeted
An AI agent has to stay running constantly. That makes the compute vendor a counterparty. CoreWeave disclosed $35.6 billion in total debt in its Q2 2026 filing. Depreciation now consumes roughly half of revenue across the leading neoclouds, tens of billions in GPU-collateralized loans mature between 2026 and 2028, and analysts expect only five to seven of these providers to remain standalone businesses in two years. Aethir has no single balance sheet under the workload. Supply comes from independent operators across 94 countries and 200+ locations. Our upcoming Aethir ACCELERATE mid-sized data centers are all demand-based, with sites chosen based on actual market demand. Concentration is a risk you inherit.
4
11
1
92
2,850
MikeMike #Sonic retweeted
A closed model and the data center serving it are one purchase. Open weights break that bundle. Portable weights are the precondition for compute being a market at all. This is the real open weights story👇
25
13
118
12,849
MikeMike #Sonic retweeted
San Francisco @Techweek_ is underway, and Aethir is in town. From October 5 to 11, hundreds of events across the city bring together the founders, engineers, and investors shaping what AI looks like next. It's a week built on direct conversations with the people building the products. That's why we're here. Every AI company hits the same question sooner or later: Where does the compute come from, and what does it cost at scale? We want to hear how founders are answering it, and show what a decentralized GPU cloud can do for teams that need capacity now. Building something in AI? Let's talk this week.
2
14
2
92
3,790
MikeMike #Sonic retweeted
Ootle package downloads more than doubled: 3,038 → 6,270. Those are 30-day totals across eight npm packages in the July and October snapshots. Curious what you can build? Start with the templates: github.com/tari-project/wasm…
3
9
45
5,094
MikeMike #Sonic retweeted
Lots of serious AI infra builders in SF for @Techweek_ Compute, data centers, agents teams are all here for exactly the conversations we came for The Aethir team is around all week! Come say hai
In town for SF Tech Week! Excited for a full week of AI infra + founder meetings Only 1 day in, and already overwhelmed by amount of quality folks working on AI datacenters, compute pricing, and agents If you are working in AI infra or just building something interesting, hit me and @AethirCloud team up! @Techweek_
6
8
88
21,882
MikeMike #Sonic retweeted
Gartner expects inference on a 1 trillion-parameter model to cost more than 90% less by 2030 than it does now. That is the direction of travel, but nobody serving users in 2026 gets to wait four years for it. The curve arrives early for whoever runs on the most efficient infrastructure available today. Aethir delivers equivalent GPU specs at 40% to 80% below hyperscaler pricing, with H100 access around 86% below comparable Google Cloud configurations. The future price is already available. It just requires switching.
12
24
171
44,930
MikeMike #Sonic retweeted
If you know you know Don’t say I didn’t tell you
24
8
1
92
15,784
MikeMike #Sonic retweeted
At production scale, inference cost comes down to two numbers: how many tokens you serve, and what you pay per GPU hour. Teams optimize the first one relentlessly with shorter prompts, smaller models, caching, and batching. The second gets treated as a fixed input, and it isn't. Aethir delivers equivalent GPU specs at 40% to 80% below hyperscaler pricing, with H100 access around 86% below comparable Google Cloud configurations. Cutting the rate you pay does more for gross margin than another round of prompt tuning.
4
12
83
2,690
‼️BREAKING: 🇨🇳 CHINA reportedly plans to lift its restrictions on cryptocurrencies by the end of 2026. Just imagine what could happen when MILLIONS of Chinese investors re-enter the crypto market during the biggest bull run in history! 🚀 THIS COULD BE HUGE! 🔥
201
461
62
2,724
336,499