After years of working in robotics, we kept running into the same problem: the best models never fit the hardware we actually had available.
The models that performed best were usually designed around datacenter assumptions: large GPUs, lots of memory bandwidth, and reliable network access. But most physical systems have the opposite constraints.
That led us down the path of figuring out how much of a frontier model could be preserved while still making it practical to run on edge hardware.
As part of that work, we recently open sourced InstinctRazor (https://github.com/General-Instinct/InstinctRazor)
One result we're excited about is compressing Qwen3.5-122B-A10B, a roughly 245 GB BF16 MoE model, into a 48 GiB GGUF. The resulting model is actually smaller than Gemma-4-26B-A4B while outperforming it on benchmarks like MMLU-Pro and GPQA-D etc. we preserve the parts that are always active (router, norms, Gated-DeltaNet/SSM layers, vision pathway, etc.) and quantize the routed experts much more aggressively. We then use on-policy distillation to recover capability lost during quantization.
The model can also run in a "small GPU" configuration where experts are streamed from system RAM. With an 8k context window, peak VRAM usage is around 7.6\u20138 GB.
If you're interested in the technical details, we wrote up the approach here (https://general-instinct.com/blog/frontier-moe-sub-4-bit)
We're especially interested in hearing from people deploying models onto robots or other edge devices. What models are you trying to run locally today? What has been the biggest bottleneck in getting them into production?"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"Launch HN: General Instinct (YC P26) \u2013 Frontier models on edge devices"}},"_tags":["story","author_guanming0717","story_48414869","launch_hn"],"author":"guanming0717","children":[48415146,48415719,48415809,48416203,48416355,48416807,48416977,48418270,48419990,48426187],"created_at":"2026-06-05T16:33:00Z","created_at_i":1780677180,"num_comments":16,"objectID":"48414869","points":63,"story_id":48414869,"story_text":"Hey HN, Guanming and Bill here from General Instinct (https://general-instinct.com/).
After years of working in robotics, we kept running into the same problem: the best models never fit the hardware we actually had available.
The models that performed best were usually designed around datacenter assumptions: large GPUs, lots of memory bandwidth, and reliable network access. But most physical systems have the opposite constraints.
That led us down the path of figuring out how much of a frontier model could be preserved while still making it practical to run on edge hardware.
As part of that work, we recently open sourced InstinctRazor (https://github.com/General-Instinct/InstinctRazor)
One result we're excited about is compressing Qwen3.5-122B-A10B, a roughly 245 GB BF16 MoE model, into a 48 GiB GGUF. The resulting model is actually smaller than Gemma-4-26B-A4B while outperforming it on benchmarks like MMLU-Pro and GPQA-D etc. we preserve the parts that are always active (router, norms, Gated-DeltaNet/SSM layers, vision pathway, etc.) and quantize the routed experts much more aggressively. We then use on-policy distillation to recover capability lost during quantization.
The model can also run in a "small GPU" configuration where experts are streamed from system RAM. With an 8k context window, peak VRAM usage is around 7.6\u20138 GB.
If you're interested in the technical details, we wrote up the approach here (https://general-instinct.com/blog/frontier-moe-sub-4-bit)
We're especially interested in hearing from people deploying models onto robots or other edge devices. What models are you trying to run locally today? What has been the biggest bottleneck in getting them into production?","title":"Launch HN: General Instinct (YC P26) \u2013 Frontier models on edge devices","updated_at":"2026-07-29T20:35:28Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"jelliclesfarm"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"In the animal kingdom, the astonishing power of the number instinct"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"https://thereader.mitpress.mit.edu/animal-kingdom-power-of-the-number-instinct/"}},"_tags":["story","author_jelliclesfarm","story_24439533"],"author":"jelliclesfarm","children":[24439788,24440076,24440477,24440567,24440809,24440831,24441277,24442888],"created_at":"2020-09-11T05:46:33Z","created_at_i":1599803193,"num_comments":36,"objectID":"24439533","points":55,"story_id":24439533,"title":"In the animal kingdom, the astonishing power of the number instinct","updated_at":"2024-09-20T06:56:27Z","url":"https://thereader.mitpress.mit.edu/animal-kingdom-power-of-the-number-instinct/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"fariszr"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"AMD Launches Instinct MI300X AI Accelerator, Up to 60% Faster Than Nvidia H100"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"https://wccftech.com/amd-launches-instinct-mi300x-ai-gpu-accelerator-up-to-60-percent-faster-nvidia-h100/"}},"_tags":["story","author_fariszr","story_38548330"],"author":"fariszr","children":[38548752,38549046,38550432],"created_at":"2023-12-06T19:24:45Z","created_at_i":1701890685,"num_comments":16,"objectID":"38548330","points":48,"story_id":38548330,"title":"AMD Launches Instinct MI300X AI Accelerator, Up to 60% Faster Than Nvidia H100","updated_at":"2024-09-20T15:50:35Z","url":"https://wccftech.com/amd-launches-instinct-mi300x-ai-gpu-accelerator-up-to-60-percent-faster-nvidia-h100/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"lifefeed"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"My application programmer instincts failed when debugging assembler"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"https://landedstar.com/blog/posts/how-my-application-programmer-instincts-failed-when-debugging-assembler/"}},"_tags":["story","author_lifefeed","story_47264742"],"author":"lifefeed","children":[47285479,47285622,47285727,47286453,47286921,47287774,47288435,47288647,47301630,47301723],"created_at":"2026-03-05T17:43:33Z","created_at_i":1772732613,"num_comments":34,"objectID":"47264742","points":45,"story_id":47264742,"title":"My application programmer instincts failed when debugging assembler","updated_at":"2026-03-15T05:39:15Z","url":"https://landedstar.com/blog/posts/how-my-application-programmer-instincts-failed-when-debugging-assembler/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"mhb"},"story_text":{"matchLevel":"none","matchedWords":[],"value":""},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"We're happier when busy but our instinct is for idleness"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"http://bps-research-digest.blogspot.com/2010/07/were-happier-when-busy-but-our-instinct.html"}},"_tags":["story","author_mhb","story_1537928"],"author":"mhb","children":[1537930,1537931,1538012,1538270,1538379],"created_at":"2010-07-22T12:28:06Z","created_at_i":1279801686,"num_comments":8,"objectID":"1537928","points":41,"story_id":1537928,"story_text":"","title":"We're happier when busy but our instinct is for idleness","updated_at":"2024-09-19T17:08:35Z","url":"http://bps-research-digest.blogspot.com/2010/07/were-happier-when-busy-but-our-instinct.html"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"benbreen"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"An Instinct for Dragons"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"https://en.wikipedia.org/wiki/An_Instinct_for_Dragons"}},"_tags":["story","author_benbreen","story_32458927"],"author":"benbreen","children":[32493737,32493765,32493817,32494039,32494115,32494280,32494446,32494473,32494480,32494498,32494555],"created_at":"2022-08-14T13:06:35Z","created_at_i":1660482395,"num_comments":32,"objectID":"32458927","points":40,"story_id":32458927,"title":"An Instinct for Dragons","updated_at":"2024-09-20T11:51:01Z","url":"https://en.wikipedia.org/wiki/An_Instinct_for_Dragons"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"kristianp"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"AMD Instinct MI325X to Feature 256GB HBM3E Memory, CDNA4-Based MI355X with 288GB"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"https://videocardz.com/newz/amd-instinct-mi325x-to-feature-256gb-hbm3e-memory-cdna4-based-mi355x-with-288gb"}},"_tags":["story","author_kristianp","story_41808351"],"author":"kristianp","children":[41808912,41808970,41809342,41809374,41809779],"created_at":"2024-10-11T10:59:19Z","created_at_i":1728644359,"num_comments":18,"objectID":"41808351","points":40,"story_id":41808351,"title":"AMD Instinct MI325X to Feature 256GB HBM3E Memory, CDNA4-Based MI355X with 288GB","updated_at":"2025-04-10T01:40:01Z","url":"https://videocardz.com/newz/amd-instinct-mi325x-to-feature-256gb-hbm3e-memory-cdna4-based-mi355x-with-288gb"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"prostoalex"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"There is no language instinct (2014)"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"https://aeon.co/essays/the-evidence-is-in-there-is-no-language-instinct"}},"_tags":["story","author_prostoalex","story_10612447"],"author":"prostoalex","children":[10612878,10612918,10612983,10613047,10613085,10613358,10613456,10613509,10613779,10614411],"created_at":"2015-11-23T02:01:27Z","created_at_i":1448244087,"num_comments":30,"objectID":"10612447","points":37,"story_id":10612447,"title":"There is no language instinct (2014)","updated_at":"2023-09-07T00:09:42Z","url":"https://aeon.co/essays/the-evidence-is-in-there-is-no-language-instinct"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"jonbaer"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"AMD, HP Unveil 2-Exaflop Supercomputer with Epyc, Radeon Instinct"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"https://www.extremetech.com/computing/307004-amd-hpe-el-capitan-2-exaflop-supercomputer-epyc-radeon-instinct"}},"_tags":["story","author_jonbaer","story_22495487"],"author":"jonbaer","children":[22496312,22497275],"created_at":"2020-03-05T17:05:32Z","created_at_i":1583427932,"num_comments":2,"objectID":"22495487","points":37,"story_id":22495487,"title":"AMD, HP Unveil 2-Exaflop Supercomputer with Epyc, Radeon Instinct","updated_at":"2024-09-20T05:47:11Z","url":"https://www.extremetech.com/computing/307004-amd-hpe-el-capitan-2-exaflop-supercomputer-epyc-radeon-instinct"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"bx376"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"AMD Instinct MI300X Accelerators"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"https://www.amd.com/en/products/accelerators/instinct/mi300/mi300x.html"}},"_tags":["story","author_bx376","story_38760072"],"author":"bx376","children":[38762265,38769765],"created_at":"2023-12-25T05:22:39Z","created_at_i":1703481759,"num_comments":27,"objectID":"38760072","points":36,"story_id":38760072,"title":"AMD Instinct MI300X Accelerators","updated_at":"2024-12-27T22:24:33Z","url":"https://www.amd.com/en/products/accelerators/instinct/mi300/mi300x.html"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"jonbaer"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"Is there a war instinct?"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"http://aeon.co/magazine/society/human-beings-do-not-have-an-instinct-for-war/"}},"_tags":["story","author_jonbaer","story_9483879"],"author":"jonbaer","children":[9484136,9484138,9484174,9484199,9484220,9484225,9484453,9484774,9489009],"created_at":"2015-05-04T01:34:55Z","created_at_i":1430703295,"num_comments":54,"objectID":"9483879","points":29,"story_id":9483879,"title":"Is there a war instinct?","updated_at":"2023-09-07T02:39:48Z","url":"http://aeon.co/magazine/society/human-beings-do-not-have-an-instinct-for-war/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"guanming0717"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"Hey HN, Guanming here, cofounder of General Instinct. We just released InstinctFlash, a high-performance serving framework for robotics models. It\u2019s licensed under AGPL-3.0.
On Jetson Thor, we see speedups about 1.2x to 7.9x from runtime optimizations alone and up to 33.78x for LingBot-VA when we combine those runtime optimizations with a distilled few-step diffusion scheduler, going from the original 25 visual / 50 action steps to 2 / 4 steps. Across 50 Robotwin2.0 tasks, we evaluated 1,153 episodes per configuration, LingBot-VA with InstinctFlash at 2 visual / 4 action steps achieved a 90.5% success rate, compared with 92.1% for the baseline at 25 visual / 50 action steps.
Here\u2019s an optimized 5B world action model, running in real time on a Jetson Thor:\nhttps://youtu.be/nku65iyL5Fw
InstinctFlash currently supports 8 VLA / world-action model families, including pi0.5 and NVIDIA Cosmos Policy, across RTX 4090 / 5090 and Jetson Thor.
Just give it your fine-tuned checkpoint and InstinctFlash handles the rest, exposing the accelerated model through a Python runtime or an OpenPI-compatible WebSocket server.
We started working on this because we kept running into the same problem while deploying robot policies, the models were getting much better, but inference was often way too slow for the control loop we actually wanted.
For pi0.5, mixed-precision GEMMs and CUDA graphs speed up computation and reduce launch overhead. For Cosmos, caching avoids redundant computation across diffusion steps. World-action models\u2019 diffusion denoising step depends on the previous one which motivated our work on few-step distillation.
Right now, InstinctFlash contains 6 aspects of optimization.
- Graph: CUDA graph capture, memory planning and separating prefill from repeated execution.
- Cache: Reusing KV and conditioning state across diffusion steps and prediction calls.
- Attention: Specialized attention paths for different model architectures.
- Kernels: Fused operations and kernels tailored to specific backends and tensor layouts.
- Precision: FP8 and mixed-precision execution.
- Model: Few-step distillation for diffusion and action generation.
Teams at Samsung, Siemens, and other robotics startups have used InstinctFlash for model acceleration on VLAs, WAMs, and diffusion-based world models. Now we are opening up access to you.
Feel free to try it here:\nhttps://github.com/General-Instinct/InstinctFlash
More implementation details and benchmarks:\nhttps://general-instinct.com/blog/instinctflash-edge-inferen...
Would love to hear your feedback!"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"Show HN: InstinctFlash \u2013 High-Performance Serving Runtime for Robotics Models"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"https://github.com/General-Instinct/InstinctFlash"}},"_tags":["story","author_guanming0717","story_49802789","show_hn"],"author":"guanming0717","children":[49806509,49812340],"created_at":"2026-09-22T15:20:06Z","created_at_i":1790090406,"num_comments":4,"objectID":"49802789","points":27,"story_id":49802789,"story_text":"Hey HN, Guanming here, cofounder of General Instinct. We just released InstinctFlash, a high-performance serving framework for robotics models. It\u2019s licensed under AGPL-3.0.
On Jetson Thor, we see speedups about 1.2x to 7.9x from runtime optimizations alone and up to 33.78x for LingBot-VA when we combine those runtime optimizations with a distilled few-step diffusion scheduler, going from the original 25 visual / 50 action steps to 2 / 4 steps. Across 50 Robotwin2.0 tasks, we evaluated 1,153 episodes per configuration, LingBot-VA with InstinctFlash at 2 visual / 4 action steps achieved a 90.5% success rate, compared with 92.1% for the baseline at 25 visual / 50 action steps.
Here\u2019s an optimized 5B world action model, running in real time on a Jetson Thor:\nhttps://youtu.be/nku65iyL5Fw
InstinctFlash currently supports 8 VLA / world-action model families, including pi0.5 and NVIDIA Cosmos Policy, across RTX 4090 / 5090 and Jetson Thor.
Just give it your fine-tuned checkpoint and InstinctFlash handles the rest, exposing the accelerated model through a Python runtime or an OpenPI-compatible WebSocket server.
We started working on this because we kept running into the same problem while deploying robot policies, the models were getting much better, but inference was often way too slow for the control loop we actually wanted.
For pi0.5, mixed-precision GEMMs and CUDA graphs speed up computation and reduce launch overhead. For Cosmos, caching avoids redundant computation across diffusion steps. World-action models\u2019 diffusion denoising step depends on the previous one which motivated our work on few-step distillation.
Right now, InstinctFlash contains 6 aspects of optimization.
- Graph: CUDA graph capture, memory planning and separating prefill from repeated execution.
- Cache: Reusing KV and conditioning state across diffusion steps and prediction calls.
- Attention: Specialized attention paths for different model architectures.
- Kernels: Fused operations and kernels tailored to specific backends and tensor layouts.
- Precision: FP8 and mixed-precision execution.
- Model: Few-step distillation for diffusion and action generation.
Teams at Samsung, Siemens, and other robotics startups have used InstinctFlash for model acceleration on VLAs, WAMs, and diffusion-based world models. Now we are opening up access to you.
Feel free to try it here:\nhttps://github.com/General-Instinct/InstinctFlash
More implementation details and benchmarks:\nhttps://general-instinct.com/blog/instinctflash-edge-inferen...
Would love to hear your feedback!","title":"Show HN: InstinctFlash \u2013 High-Performance Serving Runtime for Robotics Models","updated_at":"2026-09-24T14:56:09Z","url":"https://github.com/General-Instinct/InstinctFlash"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"nabla9"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"AMD sold $1B of Instinct GPUs in 2Q, driving 3-digit datacenter growth"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.theregister.com/2024/07/31/amd_q2_2024/"}},"_tags":["story","author_nabla9","story_41125145"],"author":"nabla9","children":[41125206,41125310,41125337],"created_at":"2024-08-01T00:48:18Z","created_at_i":1722473298,"num_comments":12,"objectID":"41125145","points":25,"story_id":41125145,"title":"AMD sold $1B of Instinct GPUs in 2Q, driving 3-digit datacenter growth","updated_at":"2024-09-20T17:31:58Z","url":"https://www.theregister.com/2024/07/31/amd_q2_2024/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"tintinnabula"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"Heard instinct: Deafness and the prosthetics for overcoming it"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.the-tls.co.uk/articles/hearing-happiness-jaipreet-virdi-review-jonathan-ree/"}},"_tags":["story","author_tintinnabula","story_26841647"],"author":"tintinnabula","children":[26850811],"created_at":"2021-04-17T04:48:28Z","created_at_i":1618634908,"num_comments":6,"objectID":"26841647","points":25,"story_id":26841647,"title":"Heard instinct: Deafness and the prosthetics for overcoming it","updated_at":"2024-09-20T08:22:59Z","url":"https://www.the-tls.co.uk/articles/hearing-happiness-jaipreet-virdi-review-jonathan-ree/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"pixelpoet"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"AMD's Instinct MI455X: Aiming for the Sun"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"https://chipsandcheese.com/p/amds-instinct-mi455x-aiming-for-the"}},"_tags":["story","author_pixelpoet","story_49025398"],"author":"pixelpoet","children":[49029763,49030048],"created_at":"2026-07-23T17:42:43Z","created_at_i":1784828563,"num_comments":4,"objectID":"49025398","points":24,"story_id":49025398,"title":"AMD's Instinct MI455X: Aiming for the Sun","updated_at":"2026-07-27T13:33:02Z","url":"https://chipsandcheese.com/p/amds-instinct-mi455x-aiming-for-the"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"Libertatea"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"Do we have an instinct for privacy?"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"http://www.aeonmagazine.com/living-together/do-we-have-a-privacy-instinct-or-are-we-wired-to-share/?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+AeonMagazineEssays+%28Aeon+Magazine+Essays%29"}},"_tags":["story","author_Libertatea","story_6174939"],"author":"Libertatea","children":[6175840,6175977,6176641,6177301],"created_at":"2013-08-07T19:11:32Z","created_at_i":1375902692,"num_comments":3,"objectID":"6174939","points":24,"story_id":6174939,"title":"Do we have an instinct for privacy?","updated_at":"2025-12-27T06:54:27Z","url":"http://www.aeonmagazine.com/living-together/do-we-have-a-privacy-instinct-or-are-we-wired-to-share/?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+AeonMagazineEssays+%28Aeon+Magazine+Essays%29"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"pcl"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"We instinctively add on new features and fixes. Why don\u2019t we subtract instead?"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.washingtonpost.com/outlook/2021/04/15/psychology-innovation-subtraction-addition/"}},"_tags":["story","author_pcl","story_26850150"],"author":"pcl","children":[26850288,26850293,26850352,26850362,26850478,26850495,26850593,26851096,26851955,26852245,26852873,26857889],"created_at":"2021-04-18T05:57:46Z","created_at_i":1618725466,"num_comments":21,"objectID":"26850150","points":23,"story_id":26850150,"title":"We instinctively add on new features and fixes. Why don\u2019t we subtract instead?","updated_at":"2025-10-06T21:02:55Z","url":"https://www.washingtonpost.com/outlook/2021/04/15/psychology-innovation-subtraction-addition/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"gdiamos"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"AI startup Lamini bets future on AMD's Instinct GPUs"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"https://www.theregister.com/2023/09/26/amd_instinct_ai_lamini/"}},"_tags":["story","author_gdiamos","story_37661836"],"author":"gdiamos","children":[37661837,37662515,37662574,37663161],"created_at":"2023-09-26T16:35:35Z","created_at_i":1695746135,"num_comments":6,"objectID":"37661836","points":22,"story_id":37661836,"title":"AI startup Lamini bets future on AMD's Instinct GPUs","updated_at":"2024-09-20T15:14:22Z","url":"https://www.theregister.com/2023/09/26/amd_instinct_ai_lamini/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"luord"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"AMD Announces New \u201cInstinct MI100\u201d GPU, Breaks the 10 Tflops Barrier in FP64"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"https://www.cloudsavvyit.com/8032/amd-announces-new-instinct-mi100-gpu-breaks-the-10-tflops-barrier-in-fp64/"}},"_tags":["story","author_luord","story_25145600"],"author":"luord","children":[25147583],"created_at":"2020-11-19T02:36:37Z","created_at_i":1605753397,"num_comments":1,"objectID":"25145600","points":21,"story_id":25145600,"title":"AMD Announces New \u201cInstinct MI100\u201d GPU, Breaks the 10 Tflops Barrier in FP64","updated_at":"2024-09-20T07:18:17Z","url":"https://www.cloudsavvyit.com/8032/amd-announces-new-instinct-mi100-gpu-breaks-the-10-tflops-barrier-in-fp64/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"rzk"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"Doomsday Prepping: Reactionary Behavior or Inherited Instinct?"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"https://seattleanxiety.com/psychiatrist/2023/1/12/doomsday-prepping-reactionary-behavior-or-inherited-instinct"}},"_tags":["story","author_rzk","story_40899562"],"author":"rzk","children":[40899945,40900002,40900179,40900395,40900420,40900481,40900554,40900596],"created_at":"2024-07-07T18:45:44Z","created_at_i":1720377944,"num_comments":49,"objectID":"40899562","points":18,"story_id":40899562,"title":"Doomsday Prepping: Reactionary Behavior or Inherited Instinct?","updated_at":"2025-08-14T22:35:36Z","url":"https://seattleanxiety.com/psychiatrist/2023/1/12/doomsday-prepping-reactionary-behavior-or-inherited-instinct"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"langer"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"Like most people who will read this, I'm in my 20s and working on a startup I founded. Almost all the advice I've read and listened to regarding decision-making is consistent:
- follow your instincts
- don't get too emotionally attached to your startup
But in order to hone your instincts, surely you want them to be as in touch with your startup as possible? If this is true, and instincts and emotions were the same thing, then this would lead to a contradiction in the advice.
Therefore, I'm beginning to think that there must be a difference between instinct and emotion. However, I'm struggling to distinguish between the two inside myself.
So - is there a difference between instinct and emotion? If there is, what is it?"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"Ask PG: The difference between instinct and emotion?"}},"_tags":["story","author_langer","story_196668","ask_hn"],"author":"langer","children":[196686,196698,196702,196714,196717,196727,196728,196731,196779,196788,196808,196886,196964,197037,198266],"created_at":"2008-05-21T23:35:56Z","created_at_i":1211412956,"num_comments":23,"objectID":"196668","points":18,"story_id":196668,"story_text":"Like most people who will read this, I'm in my 20s and working on a startup I founded. Almost all the advice I've read and listened to regarding decision-making is consistent:
- follow your instincts
- don't get too emotionally attached to your startup
But in order to hone your instincts, surely you want them to be as in touch with your startup as possible? If this is true, and instincts and emotions were the same thing, then this would lead to a contradiction in the advice.
Therefore, I'm beginning to think that there must be a difference between instinct and emotion. However, I'm struggling to distinguish between the two inside myself.
So - is there a difference between instinct and emotion? If there is, what is it?","title":"Ask PG: The difference between instinct and emotion?","updated_at":"2023-09-06T19:15:19Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"timmb"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"You have to follow your instincts in this business"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"https://timmb.com/you-have-to-follow-your-instincts-in-this-business/"}},"_tags":["story","author_timmb","story_31575329"],"author":"timmb","children":[31602574,31602591,31602803],"created_at":"2022-05-31T21:44:20Z","created_at_i":1654033460,"num_comments":0,"objectID":"31575329","points":16,"story_id":31575329,"title":"You have to follow your instincts in this business","updated_at":"2024-09-20T11:18:00Z","url":"https://timmb.com/you-have-to-follow-your-instincts-in-this-business/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"lermontov"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"The Mattering Instinct"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"https://www.edge.org/conversation/rebecca_newberger_goldstein-the-mattering-instinct"}},"_tags":["story","author_lermontov","story_11319947"],"author":"lermontov","children":[11326994,11327070],"created_at":"2016-03-19T19:07:33Z","created_at_i":1458414453,"num_comments":2,"objectID":"11319947","points":15,"story_id":11319947,"title":"The Mattering Instinct","updated_at":"2024-09-19T22:57:09Z","url":"https://www.edge.org/conversation/rebecca_newberger_goldstein-the-mattering-instinct"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"rafaelc"},"story_text":{"matchLevel":"none","matchedWords":[],"value":""},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"Killer instinct: the key to startup success"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"http://blog.rafaelcorrales.com/2010/03/killer-instinct.html"}},"_tags":["story","author_rafaelc","story_1172135"],"author":"rafaelc","children":[1172250,1172372,1172768],"created_at":"2010-03-06T20:45:13Z","created_at_i":1267908313,"num_comments":4,"objectID":"1172135","points":14,"story_id":1172135,"story_text":"","title":"Killer instinct: the key to startup success","updated_at":"2023-09-06T19:15:47Z","url":"http://blog.rafaelcorrales.com/2010/03/killer-instinct.html"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"stevederico"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"Instinct - The easiest way to learn a musical instrument"},"url":{"matchLevel":"none","matchedWords":[],"value":"http://getinstinct.com/"}},"_tags":["story","author_stevederico","story_3068855"],"author":"stevederico","children":[3069285],"created_at":"2011-10-04T00:45:58Z","created_at_i":1317689158,"num_comments":3,"objectID":"3068855","points":13,"story_id":3068855,"title":"Instinct - The easiest way to learn a musical instrument","updated_at":"2024-09-19T18:03:35Z","url":"http://getinstinct.com/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"bookofjoe"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"The evidence is in: there is no language instinct"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"https://aeon.co/essays/the-evidence-is-in-there-is-no-language-instinct"}},"_tags":["story","author_bookofjoe","story_46168625"],"author":"bookofjoe","children":[46177483],"created_at":"2025-12-05T23:11:05Z","created_at_i":1764976265,"num_comments":1,"objectID":"46168625","points":13,"story_id":46168625,"title":"The evidence is in: there is no language instinct","updated_at":"2026-03-05T23:08:35Z","url":"https://aeon.co/essays/the-evidence-is-in-there-is-no-language-instinct"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"googlechromer"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"Early LLM serving experience and performance results with AMD Instinct MI300X"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"https://blogs.oracle.com/cloud-infrastructure/post/llm-performance-results-amd-instinct-mi300x-gpus"}},"_tags":["story","author_googlechromer","story_40783141"],"author":"googlechromer","children":[40783557],"created_at":"2024-06-25T01:08:05Z","created_at_i":1719277685,"num_comments":1,"objectID":"40783141","points":12,"story_id":40783141,"title":"Early LLM serving experience and performance results with AMD Instinct MI300X","updated_at":"2024-09-20T17:20:21Z","url":"https://blogs.oracle.com/cloud-infrastructure/post/llm-performance-results-amd-instinct-mi300x-gpus"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"acsillag"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"The Art of Instinct"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"http://www.nybooks.com/daily/2015/12/30/art-of-instinct/"}},"_tags":["story","author_acsillag","story_10831623"],"author":"acsillag","created_at":"2016-01-03T18:36:28Z","created_at_i":1451846188,"num_comments":0,"objectID":"10831623","points":12,"story_id":10831623,"title":"The Art of Instinct","updated_at":"2024-09-19T22:40:38Z","url":"http://www.nybooks.com/daily/2015/12/30/art-of-instinct/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"arcanus"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"The AMD Instinct MI200 Series Accelerator"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.youtube.com/watch?v=Bm2r4Z7qlcs"}},"_tags":["story","author_arcanus","story_29151198"],"author":"arcanus","children":[29151229,29152400],"created_at":"2021-11-08T16:50:35Z","created_at_i":1636390235,"num_comments":1,"objectID":"29151198","points":11,"story_id":29151198,"title":"The AMD Instinct MI200 Series Accelerator","updated_at":"2024-09-20T09:48:15Z","url":"https://www.youtube.com/watch?v=Bm2r4Z7qlcs"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"mindcrime"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"AMD Confirms Next-Gen Instinct MI400 Series AI Accelerators Already in the Works"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"https://wccftech.com/amd-confirms-next-gen-instinct-mi400-series-ai-accelerators-already-in-the-works/"}},"_tags":["story","author_mindcrime","story_37028617"],"author":"mindcrime","created_at":"2023-08-07T00:50:22Z","created_at_i":1691369422,"num_comments":0,"objectID":"37028617","points":10,"story_id":37028617,"title":"AMD Confirms Next-Gen Instinct MI400 Series AI Accelerators Already in the Works","updated_at":"2024-09-20T14:47:47Z","url":"https://wccftech.com/amd-confirms-next-gen-instinct-mi400-series-ai-accelerators-already-in-the-works/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"2OEH8eoCRo0"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"When horror hits China, the first instinct is shut it down"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.bbc.com/news/articles/cje0z2xppdlo"}},"_tags":["story","author_2OEH8eoCRo0","story_42121228"],"author":"2OEH8eoCRo0","children":[42121337,42121436],"created_at":"2024-11-12T23:38:56Z","created_at_i":1731454736,"num_comments":6,"objectID":"42121228","points":9,"story_id":42121228,"title":"When horror hits China, the first instinct is shut it down","updated_at":"2024-11-14T08:48:35Z","url":"https://www.bbc.com/news/articles/cje0z2xppdlo"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"stevefan1999"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"Day 0 Kimi-K3 Inference Deployment with Atom on AMD Instinct MI355X GPUs"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"https://www.amd.com/en/developer/resources/technical-articles/2026/kimi-k3-on-amd-instinct-gpus.html"}},"_tags":["story","author_stevefan1999","story_49079363"],"author":"stevefan1999","created_at":"2026-07-28T04:29:17Z","created_at_i":1785212957,"num_comments":0,"objectID":"49079363","points":9,"story_id":49079363,"title":"Day 0 Kimi-K3 Inference Deployment with Atom on AMD Instinct MI355X GPUs","updated_at":"2026-07-28T18:02:22Z","url":"https://www.amd.com/en/developer/resources/technical-articles/2026/kimi-k3-on-amd-instinct-gpus.html"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"hellacious"},"story_text":{"matchLevel":"none","matchedWords":[],"value":""},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"VCs' Strange, Instinctual Need to Replace Founders"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["instinct"],"value":"http://blogs.hbr.org/cs/2012/07/unpacking_vcs_strange_instinct.html"}},"_tags":["story","author_hellacious","story_4290772"],"author":"hellacious","created_at":"2012-07-25T13:24:06Z","created_at_i":1343222646,"num_comments":0,"objectID":"4290772","points":9,"story_id":4290772,"story_text":"","title":"VCs' Strange, Instinctual Need to Replace Founders","updated_at":"2024-09-19T18:49:31Z","url":"http://blogs.hbr.org/cs/2012/07/unpacking_vcs_strange_instinct.html"}],"hitsPerPage":50,"nbHits":7677,"nbPages":20,"page":0,"params":"query=Instinct&tags=story&hitsPerPage=50&advancedSyntax=true&analyticsTags=backend","processingTimeMS":15,"processingTimingsMS":{"_request":{"roundTrip":21},"afterFetch":{"format":{"highlighting":1,"total":1}},"fetch":{"query":7,"scanning":5,"total":13},"total":15},"query":"Instinct","serverTimeMS":17}