<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Hacker News: wafngar</title><link>https://news.ycombinator.com/user?id=wafngar</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Sat, 16 May 2026 03:06:30 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=wafngar" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[New comment by wafngar in "Terrain Diffusion: A Diffusion-Based Successor to Perlin Noise"]]></title><description><![CDATA[
<p>Very common in ML research these days - claim novelty / cite prior work in an obfuscated way and so on.</p>
]]></description><pubDate>Wed, 10 Dec 2025 21:48:24 +0000</pubDate><link>https://news.ycombinator.com/item?id=46224342</link><dc:creator>wafngar</dc:creator><comments>https://news.ycombinator.com/item?id=46224342</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=46224342</guid></item><item><title><![CDATA[50 years of weather forecasting at the ECMWF]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.nature.com/articles/s41467-025-65837-2">https://www.nature.com/articles/s41467-025-65837-2</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=45958476">https://news.ycombinator.com/item?id=45958476</a></p>
<p>Points: 5</p>
<p># Comments: 0</p>
]]></description><pubDate>Mon, 17 Nov 2025 21:21:35 +0000</pubDate><link>https://www.nature.com/articles/s41467-025-65837-2</link><dc:creator>wafngar</dc:creator><comments>https://news.ycombinator.com/item?id=45958476</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45958476</guid></item><item><title><![CDATA[New comment by wafngar in "Magistral — the first reasoning model by Mistral AI"]]></title><description><![CDATA[
<p>But they have built a fully “independent” pipeline. Deepseek and others probably trained in gpt4, o1 or whatever data.</p>
]]></description><pubDate>Tue, 10 Jun 2025 21:46:24 +0000</pubDate><link>https://news.ycombinator.com/item?id=44241828</link><dc:creator>wafngar</dc:creator><comments>https://news.ycombinator.com/item?id=44241828</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=44241828</guid></item><item><title><![CDATA[New comment by wafngar in "Nvidia adds native Python support to CUDA"]]></title><description><![CDATA[
<p>Why not use torch.compile()?</p>
]]></description><pubDate>Fri, 04 Apr 2025 17:54:06 +0000</pubDate><link>https://news.ycombinator.com/item?id=43585720</link><dc:creator>wafngar</dc:creator><comments>https://news.ycombinator.com/item?id=43585720</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43585720</guid></item><item><title><![CDATA[New comment by wafngar in "Project Aardvark: reimagining AI weather prediction"]]></title><description><![CDATA[
<p>They train with ERA5 and observations.</p>
]]></description><pubDate>Mon, 24 Mar 2025 19:03:50 +0000</pubDate><link>https://news.ycombinator.com/item?id=43464341</link><dc:creator>wafngar</dc:creator><comments>https://news.ycombinator.com/item?id=43464341</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43464341</guid></item><item><title><![CDATA[New comment by wafngar in "The FFT Strikes Back: An Efficient Alternative to Self-Attention"]]></title><description><![CDATA[
<p>Should be a relevant reference: <a href="https://arxiv.org/abs/2111.13587" rel="nofollow">https://arxiv.org/abs/2111.13587</a><p>Adaptive Fourier Neural Operators: Efficient Token Mixers for Transformers
John Guibas, Morteza Mardani, Zongyi Li, Andrew Tao, Anima Anandkumar, Bryan Catanzaro</p>
]]></description><pubDate>Wed, 26 Feb 2025 14:54:16 +0000</pubDate><link>https://news.ycombinator.com/item?id=43184150</link><dc:creator>wafngar</dc:creator><comments>https://news.ycombinator.com/item?id=43184150</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43184150</guid></item><item><title><![CDATA[New comment by wafngar in "EU to mobilize 200B Euros to invest in AI"]]></title><description><![CDATA[
<p>Brussels is quite small. Probably less than Washington.</p>
]]></description><pubDate>Wed, 12 Feb 2025 12:35:19 +0000</pubDate><link>https://news.ycombinator.com/item?id=43024741</link><dc:creator>wafngar</dc:creator><comments>https://news.ycombinator.com/item?id=43024741</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43024741</guid></item><item><title><![CDATA[New comment by wafngar in "The future of Deep Learning frameworks"]]></title><description><![CDATA[
<p>Depends. PyTorch on the other hand has a large user base and well defined and tested api. So should be doable; and is already progressing and rapid speed..</p>
]]></description><pubDate>Sat, 17 Aug 2024 18:13:25 +0000</pubDate><link>https://news.ycombinator.com/item?id=41276634</link><dc:creator>wafngar</dc:creator><comments>https://news.ycombinator.com/item?id=41276634</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=41276634</guid></item><item><title><![CDATA[New comment by wafngar in "The future of Deep Learning frameworks"]]></title><description><![CDATA[
<p>Probably unfair as a reaction to the unfair statements in the blog…</p>
]]></description><pubDate>Sat, 17 Aug 2024 11:22:00 +0000</pubDate><link>https://news.ycombinator.com/item?id=41273654</link><dc:creator>wafngar</dc:creator><comments>https://news.ycombinator.com/item?id=41273654</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=41273654</guid></item><item><title><![CDATA[New comment by wafngar in "The future of Deep Learning frameworks"]]></title><description><![CDATA[
<p>PyTorch is developed by multiple companies / stake holders while jax is google only with internal tooling they don’t share with the world. This alone is a major reason not to use jax. Also I think it is more the other way around: with torch.compile the main advantage of jax is disappearing.</p>
]]></description><pubDate>Sat, 17 Aug 2024 07:43:10 +0000</pubDate><link>https://news.ycombinator.com/item?id=41272654</link><dc:creator>wafngar</dc:creator><comments>https://news.ycombinator.com/item?id=41272654</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=41272654</guid></item><item><title><![CDATA[New comment by wafngar in "XLSTMTime: Long-Term Time Series Forecasting with xLSTM"]]></title><description><![CDATA[
<p>AIFS is transformer based (Graphcast is pure GNN) so different architecture and is already running operationally, see:<p><a href="https://www.ecmwf.int/en/about/media-centre/aifs-blog/2024/its-raining-data" rel="nofollow">https://www.ecmwf.int/en/about/media-centre/aifs-blog/2024/i...</a></p>
]]></description><pubDate>Wed, 17 Jul 2024 12:25:50 +0000</pubDate><link>https://news.ycombinator.com/item?id=40985099</link><dc:creator>wafngar</dc:creator><comments>https://news.ycombinator.com/item?id=40985099</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=40985099</guid></item><item><title><![CDATA[New comment by wafngar in "Analysts estimate Nvidia owns 98% of the data center GPU market"]]></title><description><![CDATA[
<p>For AI you have to support PyTorch. That works well now on the big AMD GPUs. Consumer gpu capabilities do not matter here.</p>
]]></description><pubDate>Thu, 01 Feb 2024 20:00:49 +0000</pubDate><link>https://news.ycombinator.com/item?id=39220548</link><dc:creator>wafngar</dc:creator><comments>https://news.ycombinator.com/item?id=39220548</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=39220548</guid></item><item><title><![CDATA[New comment by wafngar in "IRIS²: The EU’s Response to Musk’s Starlink"]]></title><description><![CDATA[
<p>US has no bureaucrats? All the EU member states, no bureaucrats? How many projects in member states cost the tax payer billions? Btw the EU budget is tiny in comparison.</p>
]]></description><pubDate>Sun, 19 Feb 2023 21:33:37 +0000</pubDate><link>https://news.ycombinator.com/item?id=34862081</link><dc:creator>wafngar</dc:creator><comments>https://news.ycombinator.com/item?id=34862081</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=34862081</guid></item><item><title><![CDATA[New comment by wafngar in "IRIS²: The EU’s Response to Musk’s Starlink"]]></title><description><![CDATA[
<p>The UK would do much better had they not chosen Brexit. Basically everyone here admits that (bank of England etc) apart from some zealot politicians. And the health system has collapsed. The problems are not solely caused by Brexit but it exacerbates nearly all of them.</p>
]]></description><pubDate>Sun, 19 Feb 2023 21:11:17 +0000</pubDate><link>https://news.ycombinator.com/item?id=34861908</link><dc:creator>wafngar</dc:creator><comments>https://news.ycombinator.com/item?id=34861908</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=34861908</guid></item><item><title><![CDATA[New comment by wafngar in "IRIS²: The EU’s Response to Musk’s Starlink"]]></title><description><![CDATA[
<p>What concrete advantage would it have for you?</p>
]]></description><pubDate>Sun, 19 Feb 2023 21:05:34 +0000</pubDate><link>https://news.ycombinator.com/item?id=34861853</link><dc:creator>wafngar</dc:creator><comments>https://news.ycombinator.com/item?id=34861853</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=34861853</guid></item><item><title><![CDATA[New comment by wafngar in "Building a semantic search engine in Rust"]]></title><description><![CDATA[
<p>sure, my point was more that it uses pytorch -> also lots of c++ :-)</p>
]]></description><pubDate>Wed, 09 Nov 2022 20:10:18 +0000</pubDate><link>https://news.ycombinator.com/item?id=33537161</link><dc:creator>wafngar</dc:creator><comments>https://news.ycombinator.com/item?id=33537161</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=33537161</guid></item><item><title><![CDATA[New comment by wafngar in "Building a semantic search engine in Rust"]]></title><description><![CDATA[
<p>But isn’t that PyTorch with a few extra steps?</p>
]]></description><pubDate>Wed, 09 Nov 2022 15:43:50 +0000</pubDate><link>https://news.ycombinator.com/item?id=33533213</link><dc:creator>wafngar</dc:creator><comments>https://news.ycombinator.com/item?id=33533213</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=33533213</guid></item></channel></rss>