Advancing AI innovation together. Built with devs, for devs. Supported through an open ecosystem. Powered by AMD. #TogetherWeAdvance
- Tweets 1,892
- Following 110
- Followers 57,494
- Likes 1,191
A 50 TFLOPS system for #DeepLearning, powered by two Radeon Vega Frontier Editions, #ROCm and @AMDRyzen. medium.com/intuitionmachine/…
17
279
2,103
.@AMD showcased what can be achieved when @AMDServer #EPYC is combined with @RadeonInstinct: Project 47. instinct.radeon.com/en-us/p4…
10
183
1,245
Overheard at AMD AI Dev Day... if you had access to unlimited compute, what would you build?
From AI speedrunners to smarter systems, AMD Devs had big ideas!
50
247
2,150
211,352
Overheard at AMD AI Dev Day @AnushElangovan takes on a quick round of “This or That." Supervised vs. unsupervised? Heap sort vs. merge sort? You might be surprised at his picks.
25
196
1,879
148,309
When developers have choice, everyone wins.
@maddsey shares how openness drives the future of AI. 🤝
40
225
1,315
508,007
Devs: ROCm 7.0 isn’t just ‘another’ upgrade. It’s your ticket to AI on your terms. Explore what ROCm 7 has to offer.
46
336
1,545
908,936
Want to know how the Radeon Vega Frontier Edition sizes up? @PCWorld gets their hands on the graphics card. pcworld.com/article/3202659/…
3
84
764
ROCm 6.4 is here — built by devs, for devs:
🧠 Faster PyTorch
🐳 Prebuilt containers
⚡ Optimized #AI inference flows
➗ Faster math libs
🔗 Better GPU comms
👁️🗨️ CV boosts
🛠️ Dev tool upgrades
☁️ k8s GPU management
🧩 Modular driver
📖 Full README → rocm.blogs.amd.com/ecosystem…
38
100
759
755,359
Just dropped: Start running pre-trained, pre-quantized LLMs with Ryzen AI using Lemonade SDK. 🍋
▶️ Watch the series on YouTube: bit.ly/4lPa0KP
22
99
762
346,475
We're going to need bolder projects—and newer hardware to support them.”
@GPU_MODE @marksaroufim shares how this 17,000+ open-source community is scaling performance-focused AI with help from AMD.
50
242
1,189
232,054
Visit @AMD at #SC17, booth #825 & learn how AMD + partners are bringing a new era in the #datacenter. bit.ly/2h1asWS
15
93
689
Excited to announce AMD Ryzen AI 1.4 software:
🧠 Unified LLM, CNN, NLP support
⚙️ INT8 + BF16 model deployment
⚡ Hybrid/NPU-only execution
🧰 Opensource Tools: Digest AI, Lemonade SDK, GAIA, TurnkeyML
Start building for AMD AI PCs.
🔗 bit.ly/4iriWml
54
104
696
437,014
You can build so many different #AI applications with DeepSeek models — the possibilities are endless. This quick walkthrough gets you up and running fast with AMD Instinct GPUs. Catch it (and more) on the AMD Developer Central YouTube channel: piped.video/@AMDDevCentral/
28
82
664
451,518
Vega's High-Bandwidth Cache, the most advanced memory architecture, ensures Radeon Instinct won’t sacrifice capacity piped.video/watch?v=ZSMaMKbw…
1
51
484
AI is moving fast—but we’re still early. @LemurianLabs CEO Jay Dawani on why AMD’s open ecosystem could future-proof how devs build #AI at scale. Watch: piped.video/9f4A8L5Lx1E?si=Q_U-…
19
48
478
463,026
Introducing the @AMDInstinct MI100, engineered for the era of exascale computing. Supercharge your HPC and AI projects using the new AMD #CDNA architecture and AMD ROCm open software platform.
15
59
464
Devs, it's time to push the boundaries of open-source innovation. #ExperienceOpen with Void Run powered by AMD Instinct MI325X. Will open-source triumph? Sound off in the comments. #TogetherWeAdvance @higgsfield_ai @tensorwave
10
48
480
91,480
With the Radeon Vega Frontier Edition, developers can leverage the power of “Vega” architecture for #machinelearning pro.radeon.com/en-us/frontie…
3
55
447
A guide on how to use @AMD’s #ROCm framework with Docker container technology. instinct.radeon.com/en-us/am…
49
457
Check out 6 #deeplearning projects for AMD Radeon Instinct accelerators. instinct.radeon.com/en/6-dee…
4
76
452
"The future of AI is going to be open sourced."
In this clip, @kernelcdub of @RedHat explains why open source, open collaboration, and global ecosystems will define the next wave of AI, and how Red Hat empowers developers everywhere
13
62
430
92,042
Our passion is gaming. Our work is professional. Our instinct has always been computing. More on Radeon Instinct: piped.video/watch?v=rpRVlgOI…
3
91
388
We've excelled at understanding machines, now with Radeon Instinct machines will understand us. instinct.radeon.com/en-us/in…
2
102
382
.@AMD brings Radeon Instinct server accelerators, the MI6, MI8 and MI25, to heterogeneous computing and #HPC systems instinct.radeon.com/en-us/un…
12
88
374
Coding at dawn or midnight? @AnushElangovan answered this question and more in a lightning round at #AdvancingAI!
10
43
369
53,940
The Radeon Vega Frontier Edition, combined with #ROCm delivers performance leadership for #deeplearning. pro.radeon.com/en-us/vega-fr…
4
47
345
Accelerating the machine intelligence era in computing. Are you ready for Radeon Instinct?
7
116
329
“No gatekeeping. Just high-performance compute and real community.”
We couldn’t have said it better ourselves. If you, like @EvanKirstel, are ready to join a group of like-minded, forward thinking AI experts and tech professionals – follow us here for more. Because #TogetherWeAdvance.
8
33
346
119,369
In a world where innovation is siloed, something breaks through. Is it an anomaly or the beginning of something bigger?
🎬 Watch the teaser for VOID RUN now
👩💻 Made for developers. Inspired by possibilities.
#ExperienceOpen #TogetherWeAdvance
15
42
336
294,916
“Developers, developers, developers! That is the new mantra of ROCm. We are serious about bringing ROCm everywhere and to everyone.”
@AnushElangovan introduced the new Developer Cloud Portal at #AdvancingAI. Here’s how to get started: bit.ly/4n4lWsI
7
53
353
55,337
📢 @AndrewYNg's message to developers: build, build, build—and keep AI open.
Gatekeepers could slow innovation. Open tools like AMD’s ROCm stack help keep progress in everyone’s hands.
Watch to hear why Andrew believes openness in AI matters.
8
51
318
59,686
"We're still in the early innings of AI."
On the SHIFT podcast, AMD SVP of AI, Vamsi Boppana, talks with @StrongReporter about:
• Cloud to client AI strategy
• The future of cooperative, distributed intelligence
• Why we need to build and regulate responsibly
🎧 Listen: podcasts.apple.com/us/podcas…
22
66
512
137,970
Radeon Instinct is about improving our world - turning us into better artists, better researchers and better humans. radeon.com/en-us/radeon-inst…
4
77
309
“Real developers want access. Speed. Freedom to build—not fill out forms or chase approvals.”
Watch as @EvanKirstel discusses the importance of ROCm, an open-source AI stack that works with tools such as PyTorch, TensorFlow, JAX, and more. #TogetherWeAdvance.
7
36
304
72,509
Radeon Instinct is designed to tackle the compute challenge, synthesizing and harnessing the data we create. piped.video/watch?v=FUktXMzU…
3
51
287
AMD is excited for the emergence of the #ONNX format, giving our customers simpler paths to explore their networks. instinct.radeon.com/en/onnx-…
3
40
284
Bringing CPUs & GPUs together under one roof, @AMD enables true high-performance heterogeneous applications. piped.video/watch?v=ZSMaMKbw…
48
259
AMD partners with the open-source community to accelerate innovation.
Vamsi Boppana, SVP of AI at AMD, shares more about this journey and why we’re deeply grateful for the collaboration that makes this progress possible.
8
55
399
59,872
Speed 🤝 Scale
See how @Google Gemma uses AMD Instinct Accelerators for peak performance: bit.ly/4jjeg3g
#GoogleGemma #TogetherWeAdvance
16
74
252
53,940
The Radeon Instinct MI8 is a cost-sensitive, scalable accelerator for machine & #deeplearning inference applications instinct.radeon.com/product/…
3
54
257
The Radeon Instinct MI6 is a passively cooled accelerator for scalable #server deployments. instinct.radeon.com/product/…
2
46
262
Open is the AI foundation.
“Beyond CUDA to me is the democratization of compute... ROCm gave us a seamless training experience.”
PhD student Neha Prakriya is building smarter AI with AMD GPUs + ROCm — no walled gardens, just open innovation.
See more: bit.ly/4cGmcJo
16
63
244
46,551
Radeon Instinct MI25 brings the world’s most advanced GPU memory architecture and next-generation compute engine. instinct.radeon.com/en-us/pr…
1
39
249
Based on Microsoft's Project Olympus and powered by Radeon Instinct and #EPYC, @ZTSystems XPO200 3U PCIe Expansion System is ready to take on #AI and #deeplearning workloads along with outstanding virtualization density. bit.ly/2P0B7lX
2
31
254
The Radeon Instinct MI8 is ideal for #datacenter deployments of inference applications for #machineintelligence.
instinct.radeon.com/product/…
37
252
It’s fast, flexible, and ready to work.
Learn more about our new AI Tensor Engine for ROCm™ 👉: bit.ly/3ExlnGc #TogetherWeAdvance
10
108
234
60,179
.@AMD’s #ROCm initiative will accelerate deep learning development and leave competition in the dust. instinct.radeon.com/en-us/th…
2
23
225
The Radeon Instinct MI6 accelerator is a versatile training and inference accelerator for #machineintelligence. instinct.radeon.com/product/…
2
61
232
Devs, together we’re building and shaping the future of AI.
Introducing AMDevs, a movement for the open ecosystem.
AMDevs isn’t just AMD → it’s all of us.
It’s a community to share your stories, your builds, your wins, and your challenges.
We’ll learn through fireside chats, Q&As, and tutorials, and have some fun along the way.
Join the movement. Share your story. Collaborate.
This is only the beginning. There’s more to come.
29
132
604
584,023
🔓 Unlock expert insights from the frontlines of AI and open source. @iamthezack and Vamsi Boppana explore where AI is headed and what developers can do now to innovate and create.
Watch now on YouTube: bit.ly/4kaj4bo
4
57
210
59,496
An open community of Developers, focused on AI innovation using ROCm, leadership GPUs, and open standards-based infrastructure is better than a closed approach.
Open always wins.
#ExperienceOpen #TogetherWeAdvance
5
27
227
41,469
Today, 2-3 PM PST: join @AMD & @GFXChipTweeter's AMA on what Radeon Vega Frontier Edition means for #MachineLearning teddit.net/comments/6bklro
47
206
Introducing the Radeon Vega Frontier Edition: empowering a new generation of Pioneers to understand and analyze. pro.radeon.com/en-us/vega-fr…
1
28
195
Overheard at AMD #AdvancingAI: What does an open community mean for AI developers?
1
16
210
29,878
.@LinusTech gets a look at @AMD's #Project47, enabled by Radeon Instinct, @AMDServer #EPYC and #opensource #ROCm. piped.video/watch?v=tXjvoNta…
2
28
182
Hot off the press from #AdvancingAI, The AMD Developer Cloud is here!
Whether you're developing AI models, contributing to open-source projects, or experimenting with AMD hardware, this is your chance to innovate without worrying about infrastructure setup and get direct access to AMD Instinct MI300X GPUs.
Why Join the AMD Developer Cloud?
→ Complimentary GPU hours to start your projects*
→ Pay-as-you-go flexibility for scalable compute needs
→ Easy access to powerful GPU instances
→ Join an active, collaborative community of innovators
Learn more: amd.com/en/blogs/2025/introd…
5
47
202
49,387
Ready to deploy Meta’s Llama models at scale?
Our latest blog shows you how to harness the power of ROCm + vLLM on AMD Instinct MI300X GPUs for blazing-fast LLM inference. From setup to scalable APIs with Llama Stack, we’ve got you covered.
🔗 bit.ly/42OHqR3
7
44
193
61,474
AMD is excited to announce Day 0 support for Meta’s latest Lllama4 models using vLLM. Learn more: bit.ly/4jCRlj7 #Llama4 #TogetherWeAdvance
8
43
198
33,945
The Radeon Instinct MI25, based on "Vega" architecture, is the world’s ultimate training accelerator for #DL. instinct.radeon.com/product/…
4
39
188
Lemonade Server is here: run local LLMs with OpenAI API compatibility and Ryzen AI acceleration—no code changes, just raw speed. It’s open-source, it’s hybrid (NPU + iGPU), and it just works.
Try it now → bit.ly/44wNz6D
10
38
191
78,983
ROCm on Windows? Yes, with WSL.
AI developers can now run AMD ROCm on Windows, unlocking easier access to AMD GPUs for AI training, inference, and experimentation — no dual booting required.
Get started with @ComfyUI: rocm.blogs.amd.com/software-…
11
42
213
32,182
🚨 Learn how to install and run Lemonade Server! 🚨
1️⃣ Integrate with Microsoft’s AI Toolkit
2️⃣ Use with Continue for a local LLM coding assistant
3️⃣ Integrate into Python apps
Watch on YouTube → bit.ly/3EDUa4S
5
61
185
48,728
The Radeon Instinct MI25 is the world's fastest training accelerator for #machineintelligence and #deeplearning. instinct.radeon.com/product/…
2
46
189
.@AMD's vision of investing heavily in heterogeneous system architectures gives our platform a distinct edge. instinct.radeon.com/en-us/th…
26
160
Airing on YouTube 5/13 at 1PM ET, @iamthezack and Vamsi Boppana dive into the future of AI, open-source, and what’s in store for developers. Set a reminder to be notified when the video drops here & jump in the conversation: bit.ly/4iSNX2N
#ExperienceOpen
9
72
175
38,306
Pretraining just got smarter.
Even a 7B model trained on 198B tokens can reflect—reasoning during pretraining, not just fine-tuning.
@KarimBhalwani shares how: bit.ly/3Sd3uQ5
7
67
160
54,832
The future of local AI is here: AMD teamed up with @OpenAI and @LMStudio to make it real!
Ryzen AI Max+ 395 is the first consumer AI PC processor to run GPT‑OSS 120B, right on your machine.
Build, prototype, and run powerful models without the cloud. Start here 👉 amd.com/en/blogs/2025/how-to…
22
30
185
44,081
Radeon Instinct: Enabling a global community of brilliant minds to change the world. instinct.radeon.com/en-us/ra…
26
158
AMD Quark is now open-source! Built to streamline model quantization across AMD GPUs, processors, & SoCs, Quark supports vLLM, SGLang, PyTorch, ONNX & more empowering users with the best quantization strategies for a wide range of deployment scenarios.
bit.ly/3EUwucH
7
25
161
46,507
Large models or small? @realSharonZhou answered this question and more in a lightning round at #AdvancingAI!
2
6
163
22,435
The next leap in AI isn’t only about bigger models. It’s also about smarter training.
Together with @DeepLearningAI, we’re helping developers master fine-tuning and reinforcement learning using the post-training techniques that power today’s frontier LLMs.
Learn from @realSharonZhou, and train smarter with AMD GPUs.
🔗 Enroll now: deeplearning.ai/courses/fine…
10
36
181
64,625
Radeon Instinct MI25, MI8 and MI6 combined with AMD’s ROCm platform can dramatically speed through deep learning workloads.
2
28
142
📣 @OpenAI just dropped their latest open models.
AMD offers Day 0 support—fine-tune and run inference from cloud to client with ROCm and Ryzen AI.
Dive in → rocm.blogs.amd.com/ecosystem…
6
25
158
28,318
Radeon Instinct is an open ecosystem for universities, labs & industries - technology designed to better our future. piped.video/watch?v=FUktXMzU…
1
26
148
Worlds are colliding. The silo is breaking. Can innovation accelerate? Will open-source triumph?
Devs, tune in June 9th. @higgsfield_ai @tensorwave
#VoidRun #ExperienceOpen #TogetherWeAdvance
4
9
150
38,717
The Radeon Instinct MI25 is the world’s fastest training accelerator for machine intelligence and deep learning.
1
31
135
No dual-boot, no hacks. Native PyTorch for AMD on Windows is here. AMD just published a beginner's guide for setting up and running LLMs locally. Works on Radeon 7000/9000 GPUs and Ryzen AI APUs. Get started: gpuopen.com/learn/pytorch-wi…
8
18
150
12,397
Fine-tune Llama-3.1 8B with Llama-Factory on AMD GPUs with this step-by-step guide: bit.ly/4k14ORL
Discover more fine-tuning tutorials on the ROCm AI Developer Hub: bit.ly/4kLQiOQ
2
17
145
57,505
In this video, we demonstrate how to install and run Lemonade Server, enabling users to deploy LLMs on Ryzen AI 300 series PCs.
Watch the full video: bit.ly/4jFJ8eu
4
62
123
46,994
Instella-Math is the first fully open math reasoning model that is trained on AMD GPUs. As part of AMD’s commitment to open innovation, we’re sharing the full model weights, training setup, codebase, and datasets to foster collaboration and transparency across the AI dev community.
@AMD first reasoning model!!! Expanding on Instella family of models, Instella-Math is based on Instella-3B-Instruct model with reasoning capabilities added with multi-stage training pipeline: 2 stages of SFT and 3 stages of RL. Trained on @AMD MI300X and fully open (training code, dataset, checkpoints and technical blog). Kudo to the team. @AIatAMD
rocm.blogs.amd.com/artificia…
2
16
136
6,466
Want to build an AI agent that actually does things? This tutorial walks you through:
• Qwen3 reasoning
• Custom tool calling
• Replacing tools with MCPs
• Multi-tool Airbnb agent
All GPU-accelerated with AMD + vLLM
🔗 rocm.docs.amd.com/projects/a…
6
23
129
58,563
Local LLMs just got smarter on AMD Ryzen AI PCs. Install Lemonade Server, connect it to @huggingface Tiny Agents, and start using tools like SQLite all locally, no cloud needed.
Read more to learn how MCP unlocks powerful agent behavior with zero friction: amd.com/en/developer/resourc…
3
11
132
51,364
Four Radeon Instinct MI25 server accelerators paired with @AMDServer #EPYC for 100TFLOPs of computing power, created by partners Inventec.
4
32
126
Ready to scale LLM inference with Meta’s Llama Stack on AMD Instinct GPUs?
Full guide walks you through vLLM Docker set up and integration with the Llama Stack CLI & Python SDK—powered by AMD ROCm—for a seamless, production-ready setup. bit.ly/3S224YT
5
26
112
30,064
When it comes to unlocking AI & software capabilities, we’re just getting started. Devs, tune into a new fireside chat with Vice President of AI Software at AMD, @AnushElangovan & the amazing TechFluencer himself, @EvanKirstel’s uncovering where AMD is headed in pushing the open ecosystem.
📺 Watch now on YouTube: piped.video/watch?v=VLzys9ys…
#TogetherWeAdvance
1
7
111
48,021
ROCm 7 enables Day-0 model support, optimized Triton kernels, and elastic scaling for engineering teams building AI at scale. Explore how ROCm 7.0 supports inference and fine-tuning at scale: bit.ly/4mGgWZF
2
19
115
4,772
Our CEO @LisaSu proudly unveils Project 47, a petaflop server rack, built with @AMAXInfoTech, powered by @AMDServer #EPYC & @RadeonInstinct.
4
20
107
FM2 (Liquid Foundation Model 2) from Liquid AI just dropped today with 3 model weights: 350M, 700M, 1.2B. LFM2 is specifically designed to provide the fastest on-device gen-AI experience across the industry.
What's more is LFM2 has been optimized on AMD Ryzen AI day 0 and works out of the box. It provides many advancements such as increased accuracy, efficiency, and the ability to generate more complex and diverse outputs. Try them now on the Liquid Playground, @huggingface, and @openrouter.
AMD is proud to support @liquidai.
Today, we release the 2nd generation of our Liquid foundation models, LFM2.
LFM2 set the bar for quality, speed, and memory efficiency in on-device AI.
Built for edge devices like phones, laptops, AI PCs, cars, wearables, satellites, and robots, LFM2 delivers the fastest on-device gen-AI experience on the market. 1/
4
20
121
24,858
New ROCm 7. New Instinct MI350 series. Brand-new AMD Developer Cloud.
Same vibe: open, fast, no lock-in. 🧪⚡️
If you're building with PyTorch, scaling models, or just tired of closed ecosystems—this one's for you.
Summary of #AdvancingAI by @TechAllyson:
🔗 techarena.ai/content/amd-cha…
7
8
101
23,211
Devs, #VoidRun is coming. Will open-source triumph?
Stay connected for the full episode release on June 9th. #ExperienceOpen
2
7
90
23,131
Hear from @simon_mo_, PhD student at @UCBerkeley and vLLM co-lead, at Advancing AI 2025 and learn how to overcome LLM inference challenges using memory-efficient techniques—unlocking fast, scalable AI for real-time applications across industries. web.cvent.com/event/c137f02f…
#AdvancingAI2025 #vLLM
5
10
94
24,467
The Radeon Instinct initiative combines a hardware platform with #ROCm to propel machine intelligence forward.
instinct.radeon.com/en-us/in…
3
18
82
We are proud to support DeepSeek-V3.2-Exp, the latest release from DeepSeek AI, optimized to run on our flagship AMD Instinct MI355X GPUs.
Follow the simple steps in this article to launch the model on your AMD system: amd.com/en/developer/resourc…
3
7
89
4,274
The era of exascale is here. Immense computational power coupled with the fusion of HPC and AI is enabling researchers to tackle grand challenges once thought beyond reach — powered by @AMD.
3
4
84
Premiering May 13 at 1PM ET!
Get ready for a conversation with @iamthezack and Vamsi Boppana, SVP of AI at AMD. If you’re looking towards the future of compute and AI inference, this one’s for you. #ExperienceOpen #TogetherWeAdvance
Set Reminder: piped.video/w0ji-LILi_4
6
35
79
23,376
FLUX + @huggingface on AMD Instinct MI300X = serious speed.
Unleash high-performance image generation using AMD hardware with minimal setup → bit.ly/4jjIeUR #TogetherWeAdvance
6
14
84
16,550
Developers move AI forward.
@roaner shares more on how AMD is showing up for the community through open-source work and hands-on collaboration.
65
16
84
27,430
Learn from one of the core maintainers of vLLM, @robertshaw21 (@RedHat), as he dives into the challenges of production-scale model serving.
💡Level up your serving game at #AdvancingAI: bit.ly/414PETQ
4
9
77
15,900
SF-based developers, new hackathon alert! 🚨
Join us tomorrow, Sunday 11/16 at 9:30 AM at the @liquidai office for a fast-paced, one-day innovation sprint pushing the boundaries of AI and edge computing.
All projects deploy on AMD NUCs with AMD ROCm acceleration — compact systems engineered for high-performance AI.
Bring your ideas, build something bold, and collaborate with fellow creators.
🏆 $5K in cash prizes and two Framework Desktop systems featuring the AMD Ryzen AI 395+ processor with 128GB RAM
Walk-ins welcome. Creativity required. See you at the edge!
📍 555 Mission St, San Francisco
🕤 Nov 16, 9:30 AM PST
🍴 Lunch and snacks provided
👉 Learn more and register: luma.com/smik3k94
4
15
82
36,082