<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: muxamilian</title><link>https://news.ycombinator.com/user?id=muxamilian</link><description>Hacker News RSS</description><docs>https://hnrss.org/</docs><generator>hnrss v2.1.1</generator><lastBuildDate>Fri, 01 May 2026 22:08:37 +0000</lastBuildDate><atom:link href="https://hnrss.org/user?id=muxamilian" rel="self" type="application/rss+xml"></atom:link><item><title><![CDATA[Elon Musk's xAI discussed partnership with Mistral to try and rival OpenAI]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.euronews.com/next/2026/04/24/elon-musks-xai-discussed-partnership-with-mistral-to-try-and-rival-openai-and-anthropic-re">https://www.euronews.com/next/2026/04/24/elon-musks-xai-discussed-partnership-with-mistral-to-try-and-rival-openai-and-anthropic-re</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=47915134">https://news.ycombinator.com/item?id=47915134</a></p>
<p>Points: 5</p>
<p># Comments: 0</p>
]]></description><pubDate>Sun, 26 Apr 2026 22:00:55 +0000</pubDate><link>https://www.euronews.com/next/2026/04/24/elon-musks-xai-discussed-partnership-with-mistral-to-try-and-rival-openai-and-anthropic-re</link><dc:creator>muxamilian</dc:creator><comments>https://news.ycombinator.com/item?id=47915134</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47915134</guid></item><item><title><![CDATA[Cash-strapped AI chip startup Hailo sees valuation halved ahead of urgent IPO]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.calcalistech.com/ctechnews/article/rj000qzaowx">https://www.calcalistech.com/ctechnews/article/rj000qzaowx</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=47628207">https://news.ycombinator.com/item?id=47628207</a></p>
<p>Points: 2</p>
<p># Comments: 1</p>
]]></description><pubDate>Fri, 03 Apr 2026 15:52:20 +0000</pubDate><link>https://www.calcalistech.com/ctechnews/article/rj000qzaowx</link><dc:creator>muxamilian</dc:creator><comments>https://news.ycombinator.com/item?id=47628207</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47628207</guid></item><item><title><![CDATA[BlackRock Backs Dutch Chipmaker Axelera AI in $250M Round]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.bloomberg.com/news/articles/2026-02-24/blackrock-backs-dutch-chipmaker-axelera-ai-in-250-million-round">https://www.bloomberg.com/news/articles/2026-02-24/blackrock-backs-dutch-chipmaker-axelera-ai-in-250-million-round</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=47137076">https://news.ycombinator.com/item?id=47137076</a></p>
<p>Points: 1</p>
<p># Comments: 1</p>
]]></description><pubDate>Tue, 24 Feb 2026 13:43:55 +0000</pubDate><link>https://www.bloomberg.com/news/articles/2026-02-24/blackrock-backs-dutch-chipmaker-axelera-ai-in-250-million-round</link><dc:creator>muxamilian</dc:creator><comments>https://news.ycombinator.com/item?id=47137076</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=47137076</guid></item><item><title><![CDATA[Qualcomm scores big win over Arm in contentious lawsuit]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.tomshardware.com/pc-components/cpus/qualcomm-scores-big-win-over-arm-in-contentious-lawsuit-u-s-court-rejects-arms-lawsuit-confirms-qualcomms-can-use-oryon-cores-acquired-via-nuvia">https://www.tomshardware.com/pc-components/cpus/qualcomm-scores-big-win-over-arm-in-contentious-lawsuit-u-s-court-rejects-arms-lawsuit-confirms-qualcomms-can-use-oryon-cores-acquired-via-nuvia</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=45437098">https://news.ycombinator.com/item?id=45437098</a></p>
<p>Points: 2</p>
<p># Comments: 0</p>
]]></description><pubDate>Wed, 01 Oct 2025 12:52:52 +0000</pubDate><link>https://www.tomshardware.com/pc-components/cpus/qualcomm-scores-big-win-over-arm-in-contentious-lawsuit-u-s-court-rejects-arms-lawsuit-confirms-qualcomms-can-use-oryon-cores-acquired-via-nuvia</link><dc:creator>muxamilian</dc:creator><comments>https://news.ycombinator.com/item?id=45437098</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45437098</guid></item><item><title><![CDATA[New comment by muxamilian in "Orange Pi RV2 $40 RISC-V SBC: Friendly Gateway to IoT and AI Projects"]]></title><description><![CDATA[
<p>They get this number by simply adding up all CPUs' processing speed. There is no NPU or similar: 
<a href="https://medium.com/@zlodeibaal/orange-pi-rv2-ai-board-scam-7c71a93d63f1" rel="nofollow">https://medium.com/@zlodeibaal/orange-pi-rv2-ai-board-scam-7...</a></p>
]]></description><pubDate>Thu, 18 Sep 2025 08:22:29 +0000</pubDate><link>https://news.ycombinator.com/item?id=45287000</link><dc:creator>muxamilian</dc:creator><comments>https://news.ycombinator.com/item?id=45287000</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=45287000</guid></item><item><title><![CDATA[New comment by muxamilian in "Is Gemini 2.5 good at bounding boxes?"]]></title><description><![CDATA[
<p>I'm rather puzzled by how bad the COCO ground truth is. This is the benchmark dataset for object detection? Wow. I would say Gemini's output is better than the ground truth in most of the example images.</p>
]]></description><pubDate>Thu, 10 Jul 2025 16:25:54 +0000</pubDate><link>https://news.ycombinator.com/item?id=44522735</link><dc:creator>muxamilian</dc:creator><comments>https://news.ycombinator.com/item?id=44522735</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=44522735</guid></item><item><title><![CDATA[Chip design is a RISC-y business: Codasip puts itself up for sale]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.theregister.com/2025/07/01/codasip_sale/">https://www.theregister.com/2025/07/01/codasip_sale/</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=44448679">https://news.ycombinator.com/item?id=44448679</a></p>
<p>Points: 4</p>
<p># Comments: 0</p>
]]></description><pubDate>Wed, 02 Jul 2025 20:51:20 +0000</pubDate><link>https://www.theregister.com/2025/07/01/codasip_sale/</link><dc:creator>muxamilian</dc:creator><comments>https://news.ycombinator.com/item?id=44448679</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=44448679</guid></item><item><title><![CDATA[New comment by muxamilian in "Ironwood: The first Google TPU for the age of inference"]]></title><description><![CDATA[
<p>In-memory computing (analog or digital). Still doing SIMD matrix multiplication but using more efficient hardware:
<a href="https://arxiv.org/html/2401.14428v1" rel="nofollow">https://arxiv.org/html/2401.14428v1</a>
<a href="https://www.nature.com/articles/s41565-020-0655-z" rel="nofollow">https://www.nature.com/articles/s41565-020-0655-z</a></p>
]]></description><pubDate>Wed, 09 Apr 2025 16:16:47 +0000</pubDate><link>https://news.ycombinator.com/item?id=43633782</link><dc:creator>muxamilian</dc:creator><comments>https://news.ycombinator.com/item?id=43633782</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43633782</guid></item><item><title><![CDATA[New comment by muxamilian in "Europe bets once again on RISC-V for supercomputing"]]></title><description><![CDATA[
<p>Why? Genuinely curious</p>
]]></description><pubDate>Sun, 09 Mar 2025 18:12:40 +0000</pubDate><link>https://news.ycombinator.com/item?id=43311962</link><dc:creator>muxamilian</dc:creator><comments>https://news.ycombinator.com/item?id=43311962</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43311962</guid></item><item><title><![CDATA[Europe bets once again on RISC-V for supercomputing]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.theregister.com/2025/03/07/dare_europe_risc_v_project/">https://www.theregister.com/2025/03/07/dare_europe_risc_v_project/</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=43311091">https://news.ycombinator.com/item?id=43311091</a></p>
<p>Points: 219</p>
<p># Comments: 219</p>
]]></description><pubDate>Sun, 09 Mar 2025 17:04:28 +0000</pubDate><link>https://www.theregister.com/2025/03/07/dare_europe_risc_v_project/</link><dc:creator>muxamilian</dc:creator><comments>https://news.ycombinator.com/item?id=43311091</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43311091</guid></item><item><title><![CDATA[Nvidia and Broadcom testing chips on Intel manufacturing process, sources say]]></title><description><![CDATA[
<p>Article URL: <a href="https://www.reuters.com/technology/nvidia-broadcom-testing-chips-intel-manufacturing-process-sources-say-2025-03-03/">https://www.reuters.com/technology/nvidia-broadcom-testing-chips-intel-manufacturing-process-sources-say-2025-03-03/</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=43241384">https://news.ycombinator.com/item?id=43241384</a></p>
<p>Points: 2</p>
<p># Comments: 3</p>
]]></description><pubDate>Mon, 03 Mar 2025 13:11:38 +0000</pubDate><link>https://www.reuters.com/technology/nvidia-broadcom-testing-chips-intel-manufacturing-process-sources-say-2025-03-03/</link><dc:creator>muxamilian</dc:creator><comments>https://news.ycombinator.com/item?id=43241384</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43241384</guid></item><item><title><![CDATA[New comment by muxamilian in "Coral USB Accelerator with Google's Edge TPU"]]></title><description><![CDATA[
<p>It seems more than dead and only supports small neural networks. Viable alternatives are Hailo and Axelera (<a href="https://www.axelera.ai" rel="nofollow">https://www.axelera.ai</a>), which is a newer.</p>
]]></description><pubDate>Fri, 21 Feb 2025 12:48:28 +0000</pubDate><link>https://news.ycombinator.com/item?id=43126854</link><dc:creator>muxamilian</dc:creator><comments>https://news.ycombinator.com/item?id=43126854</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=43126854</guid></item><item><title><![CDATA[New comment by muxamilian in "YOLOv5 on FPGA with Hailo-8 and 4 Pi Cameras"]]></title><description><![CDATA[
<p>Maybe not 10 but in principle it’s possible:  <a href="https://www.lannerinc.com/products/edge-ai-appliance/deep-learning-accelerators/falcon-h8" rel="nofollow">https://www.lannerinc.com/products/edge-ai-appliance/deep-le...</a></p>
]]></description><pubDate>Sat, 01 Jun 2024 12:22:44 +0000</pubDate><link>https://news.ycombinator.com/item?id=40545149</link><dc:creator>muxamilian</dc:creator><comments>https://news.ycombinator.com/item?id=40545149</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=40545149</guid></item><item><title><![CDATA[New comment by muxamilian in "YOLOv5 on FPGA with Hailo-8 and 4 Pi Cameras"]]></title><description><![CDATA[
<p>Google Coral seems abandoned by Google. Not official but last news on their page from May 2022</p>
]]></description><pubDate>Fri, 31 May 2024 13:35:20 +0000</pubDate><link>https://news.ycombinator.com/item?id=40534941</link><dc:creator>muxamilian</dc:creator><comments>https://news.ycombinator.com/item?id=40534941</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=40534941</guid></item><item><title><![CDATA[New comment by muxamilian in "YOLOv5 on FPGA with Hailo-8 and 4 Pi Cameras"]]></title><description><![CDATA[
<p>It’s a competitor to Google Coral (seems abandoned) and NVIDIA Jetson. I’ve been using it for more than a year and the hardware seems to be one of the best on the market. The software (how to actually do inference on the chip) is subpar though.</p>
]]></description><pubDate>Fri, 31 May 2024 13:06:13 +0000</pubDate><link>https://news.ycombinator.com/item?id=40534664</link><dc:creator>muxamilian</dc:creator><comments>https://news.ycombinator.com/item?id=40534664</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=40534664</guid></item><item><title><![CDATA[New comment by muxamilian in "YOLOv5 on FPGA with Hailo-8 and 4 Pi Cameras"]]></title><description><![CDATA[
<p>From what I know it doesn’t have memory but streams everything to the chip. So there’s no limit regarding the size of the the neural network (unlike Google Coral)</p>
]]></description><pubDate>Fri, 31 May 2024 13:04:17 +0000</pubDate><link>https://news.ycombinator.com/item?id=40534646</link><dc:creator>muxamilian</dc:creator><comments>https://news.ycombinator.com/item?id=40534646</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=40534646</guid></item><item><title><![CDATA[New comment by muxamilian in "Only smart LLMs can understand good essays"]]></title><description><![CDATA[
<p>I analyzed several LLMs and how well they "understand" university essays: For all LLMs, bad essays are hard to understand. But to <i>smart</i> LLMs, good essays make sense. This could be a way to <i>measure an LLM's reasoning ability</i>.</p>
]]></description><pubDate>Sat, 09 Mar 2024 17:37:45 +0000</pubDate><link>https://news.ycombinator.com/item?id=39653281</link><dc:creator>muxamilian</dc:creator><comments>https://news.ycombinator.com/item?id=39653281</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=39653281</guid></item><item><title><![CDATA[Only smart LLMs can understand good essays]]></title><description><![CDATA[
<p>Article URL: <a href="https://muxamilian.github.io/essay-entropy-llms/">https://muxamilian.github.io/essay-entropy-llms/</a></p>
<p>Comments URL: <a href="https://news.ycombinator.com/item?id=39653280">https://news.ycombinator.com/item?id=39653280</a></p>
<p>Points: 1</p>
<p># Comments: 1</p>
]]></description><pubDate>Sat, 09 Mar 2024 17:37:45 +0000</pubDate><link>https://muxamilian.github.io/essay-entropy-llms/</link><dc:creator>muxamilian</dc:creator><comments>https://news.ycombinator.com/item?id=39653280</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=39653280</guid></item><item><title><![CDATA[New comment by muxamilian in "Low Latency, Low Loss, and Scalable Throughput (L4S) Internet Service: RFC 9330"]]></title><description><![CDATA[
<p>While it's a step in the right direction, there's a problem if there's at least one 'malicious' actor, who ignores the congestion feedback and just wants a larger share of bandwidth. Then all other actors will retreat and the unfair actors get what they want. Unfortunately it is hard to know for a good actor if the other actors are playing nicely or not. Only if a good actor knows that there's fair queuing, they can trust L4S to treat them fairly.<p>This can be solved by complementing L4S with fair queuing (e.g. fq_codel) and by making sure that congestion control can detect the presence of fair queuing (<a href="https://github.com/muxamilian/fair-queuing-aware-congestion-control">https://github.com/muxamilian/fair-queuing-aware-congestion-...</a>).</p>
]]></description><pubDate>Mon, 11 Dec 2023 12:26:53 +0000</pubDate><link>https://news.ycombinator.com/item?id=38600003</link><dc:creator>muxamilian</dc:creator><comments>https://news.ycombinator.com/item?id=38600003</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=38600003</guid></item><item><title><![CDATA[New comment by muxamilian in "A PCIe Coral TPU Finally Works on Raspberry Pi 5"]]></title><description><![CDATA[
<p>The software is actively maintained yes. The price is between 100 to 200$.</p>
]]></description><pubDate>Sat, 18 Nov 2023 09:45:37 +0000</pubDate><link>https://news.ycombinator.com/item?id=38317482</link><dc:creator>muxamilian</dc:creator><comments>https://news.ycombinator.com/item?id=38317482</comments><guid isPermaLink="false">https://news.ycombinator.com/item?id=38317482</guid></item></channel></rss>