Mark Heaps
@lifebypixels
Followers
3K
Following
13K
Media
1K
Statuses
12K
Professional Creative keeping the human soul in technology brands. Director of Developer Engagement/Community at NVIDIA. Prev: Groq, Google, Duarte, and Apple.
Austin, TX
Joined June 2010
Want to try your own classic "strawberry" test or other prompt evaluation? Give it a spin at: https://t.co/j04Zz9Tqbd NO LOGIN REQUIRED! So far it's gotten all of my tests right!
🏎️ ️We accelerated @OpenAI’s new GPT open weight models -- gpt-oss-20b and gpt-oss-120b -- for leading inference performance on NVIDIA Blackwell architecture, delivering up to 1.5 million tokens per second on an NVIDIA GB200 NVL72 system.⏱️🏁 👀 Technical deep dive on how we
2
0
9
so about those reveals...
At #NVIDIAGTC Washington, D.C., we’re unveiling the newest NVIDIA partner and how together we are rolling out a first-ever, fully integrated software stack with a native data integration engine, operational applications and frameworks, and best-in-class AI combined with
0
0
0
When an API endpoint isn't enough and the data you are running needs to be completely private for your research...
🚀 SGLang In-Depth Review of the NVIDIA DGX Spark is LIVE! Thanks to @NVIDIA’s early access program, SGLang makes its first ever appearance in a consumer product, the brand-new DGX Spark. The DGX Spark’s 128GB Unified Memory and Blackwell architecture set a new standard for
0
0
1
ComfyUI is supported on NVIDIA DGX Spark! Create with speed, locally. 🎨⚡ Congratulations on the new release! @NVIDIA_AI_PC #SparkSomethingBig
12
27
182
📢#NVIDIADGX Spark arrives for the world's AI developers. #SparkSomethingBig Check out the specs and when it'll start shipping by reading our announcement below. ⬇️
🚨NVIDIA and its partners will start shipping DGX Spark — the world’s smallest AI supercomputer. Early recipients are testing, validating and optimizing their tools, software and models for DGX Spark. Built on the NVIDIA Grace Blackwell architecture, DGX Spark integrates
4
19
122
🚀🚀Just spun up gpt-oss-20B with vLLM on NVIDIA’s brand-new DGX Spark machine — a surprisingly powerful little box! This isn’t a rigorous benchmark, but early numbers look promising on a quick test (512×128 tokens) with stable serving and smooth setup!
47
74
956
🚨 NVIDIA just did the impossible. They trained a 12B-parameter language model on 10 trillion tokens entirely in 4-bit precision. It’s called NVFP4, and it might redefine how frontier AI models are trained. Here’s why this matters: • NVFP4 delivers 2–3× faster math
68
329
2K
Today is a day I've been waiting for most of this year - @NVIDIA DGX Spark is finally available!!! I've been fortunate enough to get early access to it, and for the past week I've been playing with it and testing it. Even one week is not long enough time to really put it to a
14
10
123
🚨NVIDIA and its partners will start shipping DGX Spark — the world’s smallest AI supercomputer. Early recipients are testing, validating and optimizing their tools, software and models for DGX Spark. Built on the NVIDIA Grace Blackwell architecture, DGX Spark integrates
46
200
1K
DGX Spark in the wild! @NVIDIAAIDev build for AI everywhere...
We're collaborating with @nvidia to push the boundaries of local AI development with the new DGX Spark, bringing the power of the Grace Blackwell Superchip to developers' desktops. 🚀 As JetBrains is one of the first organizations to receive DGX Spark, we’re excited to
0
0
1
DGX Spark in the wild!
1
0
2
Excited to release new repo: nanochat! (it's among the most unhinged I've written). Unlike my earlier similar repo nanoGPT which only covered pretraining, nanochat is a minimal, from scratch, full-stack training/inference pipeline of a simple ChatGPT clone in a single,
632
3K
23K
We’re honored to share that the NVIDIA DGX Spark was named one of TIME's Best Inventions of 2025 🏆 Our new desktop AI supercomputer will deliver an unprecedented 1 petaflop of AI compute, right to your workspace, making state-of-the-art AI more accessible than ever. 🙌 We
70
159
935
Small but mighty, the applications for the Spark are really exciting. I'm looking forward to these getting into the hands of the @NVIDIAAIDev Community and seeing what they build with it. Source: Time Magazine
time.com
Find out why Nvidia DGX Spark is on TIME's list of 2025's best inventions.
0
0
2
Today we are launching InferenceMAX! We have support from Nvidia, AMD, OpenAI, Microsoft, Pytorch, SGLang, vLLM, Oracle, CoreWeave, TogetherAI, Nebius, Crusoe, HPE, SuperMicro, Dell It runs every day on the latest software (vLLM, SGLang, etc) across hundreds of GPUs, $10Ms of
Going to be dropping something huge in 24 hours I think it'll reshape how everyone thinks about chips, inference, and infrastructure It's directly supported by NVIDIA, AMD, Microsoft, OpenAI, Together AI, CoreWeave, Nebius, PyTorch Foundation, Supermicro, Crusoe, HPE, Tensorwave,
110
144
2K
📣 NVIDIA Blackwell sets the standard for AI inference on SemiAnalysis InferenceMAX. Our most recent results on the independent benchmarks show NVIDIA’s Blackwell Platform leads AI factory ROI—— see how NVIDIA Blackwell GB200 NVL72 can yield $75 million in token revenue over
42
108
640
Thanks again to @mattshumer_ for joining us for a rad convo about models, agents, Nemotron from @NVIDIAAI and much more... Watch it here if you missed the stream.
Ask the Experts: Nemotron Nano, Tool Calling, and Agent Efficiency with @mattshumer_, @lifebypixels and @llm_wizard
2
1
8
Matt Shumer is incredible for this.
Introducing: Sora Extend! Generate Sora 2 videos of INFINITE length :) Bypasses OpenAI's 12-second limit with intelligent prompt enhancement and additional last-frame context. Open source, link here:
10
6
108
The Sora API is *really* good. It just generated this movie trailer in one shot. Total cost: $1.20
148
67
878