  <?xml version="1.0" encoding="UTF-8"?>
<rss xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/" version="2.0">
  <channel>
    <title>Blog</title>
    <link>/posts</link>
    <description>All the latest developer tutorials, AI insights, news and announcements from 乐播传媒视频</description>
    <language>en-gb</language>
    <pubDate>Wed, 26 Aug 2026 08:00:00 GMT</pubDate>
    <dc:date>2026-08-26T08:00:00Z</dc:date>
    <dc:language>en-gb</dc:language>
    <item>
      <title>From Research to Reality: Fitting an 11B VLM on a Phone</title>
      <link>/posts/from-research-to-reality-fitting-an-11b-vlm-on-a-phone</link>
      <description>&lt;div class="hs-featured-image-wrapper"&gt; 
 &lt;a href="/posts/from-research-to-reality-fitting-an-11b-vlm-on-a-phone" title="" class="hs-featured-image-link"&gt; &lt;img src="/hubfs/llama-vlm-blog-featured.jpg" alt="From Research to Reality: Fitting an 11B VLM on a Phone" class="hs-featured-image" style="width:auto !important; max-width:50%; float:left; margin:0 15px 15px 0;"&gt; &lt;/a&gt; 
&lt;/div&gt; 
&lt;p&gt;It鈥檚 great to have a vision-language model (VLM) in your pocket. The combination of textual and visual intelligence is a perfect fit for personal devices like phones, laptops and wearables. However, such models can be memory-intensive and challenging to deploy on low-power devices. While demand continues to grow for large frontier models running in the cloud, AI for power constrained mobile devices and robotics is also growing fast.&amp;nbsp;&lt;br&gt;&lt;br&gt;A team of 乐播传媒视频 and Arm researchers recently took on a deliberately ambitious challenge: to fit a large 11B VLM, originally requiring more than 21 GB for its weights, into a storage budget of roughly 4 GB, to fit on an Android phone.&amp;nbsp;&lt;br&gt;&lt;br&gt;The result was a 3.7 GB implementation of Llama 3.2 Vision Instruct, which we achieved using a novel format codesigned with Arm CPUs in mind and a quantisation-aware training procedure that samples diverse prompts to retain the capabilities of the original model.&amp;nbsp;&lt;/p&gt;</description>
      <content:encoded>&lt;div class="hs-featured-image-wrapper"&gt; 
 &lt;a href="/posts/from-research-to-reality-fitting-an-11b-vlm-on-a-phone" title="" class="hs-featured-image-link"&gt; &lt;img src="/hubfs/llama-vlm-blog-featured.jpg" alt="From Research to Reality: Fitting an 11B VLM on a Phone" class="hs-featured-image" style="width:auto !important; max-width:50%; float:left; margin:0 15px 15px 0;"&gt; &lt;/a&gt; 
&lt;/div&gt; 
&lt;p&gt;It鈥檚 great to have a vision-language model (VLM) in your pocket. The combination of textual and visual intelligence is a perfect fit for personal devices like phones, laptops and wearables. However, such models can be memory-intensive and challenging to deploy on low-power devices. While demand continues to grow for large frontier models running in the cloud, AI for power constrained mobile devices and robotics is also growing fast.&amp;nbsp;&lt;br&gt;&lt;br&gt;A team of 乐播传媒视频 and Arm researchers recently took on a deliberately ambitious challenge: to fit a large 11B VLM, originally requiring more than 21 GB for its weights, into a storage budget of roughly 4 GB, to fit on an Android phone.&amp;nbsp;&lt;br&gt;&lt;br&gt;The result was a 3.7 GB implementation of Llama 3.2 Vision Instruct, which we achieved using a novel format codesigned with Arm CPUs in mind and a quantisation-aware training procedure that samples diverse prompts to retain the capabilities of the original model.&amp;nbsp;&lt;/p&gt;  
&lt;img src="https://track.hubspot.com/__ptq.gif?a=729091&amp;amp;k=14&amp;amp;r=https%3A%2F%2Fwww.graphcore.ai%2Fposts%2Ffrom-research-to-reality-fitting-an-11b-vlm-on-a-phone&amp;amp;bu=https%253A%252F%252Fwww.graphcore.ai%252Fposts&amp;amp;bvt=rss" alt="" width="1" height="1" style="min-height:1px!important;width:1px!important;border-width:0!important;margin-top:0!important;margin-bottom:0!important;margin-right:0!important;margin-left:0!important;padding-top:0!important;padding-bottom:0!important;padding-right:0!important;padding-left:0!important; "&gt;</content:encoded>
      <category>AI</category>
      <category>Research</category>
      <pubDate>Wed, 26 Aug 2026 08:00:00 GMT</pubDate>
      <guid>/posts/from-research-to-reality-fitting-an-11b-vlm-on-a-phone</guid>
      <dc:date>2026-08-26T08:00:00Z</dc:date>
      <dc:creator>Douglas Orr and Luke Ribar</dc:creator>
    </item>
    <item>
      <title>Welcome to 乐播传媒视频 Taipei</title>
      <link>/posts/welcome-to-graphcore-taipei</link>
      <description>&lt;div class="hs-featured-image-wrapper"&gt; 
 &lt;a href="/posts/welcome-to-graphcore-taipei" title="" class="hs-featured-image-link"&gt; &lt;img src="/hubfs/Taiwan_social_preview.png" alt="Welcome to 乐播传媒视频 Taipei" class="hs-featured-image" style="width:auto !important; max-width:50%; float:left; margin:0 15px 15px 0;"&gt; &lt;/a&gt; 
&lt;/div&gt; 
&lt;p&gt;&lt;span&gt;On 4 June, we officially opened a stunning new 乐播传媒视频 office and engineering lab in Taipei鈥檚 Nangang district, in addition to our existing facility in Hsinchu. This expansion demonstrates 乐播传媒视频鈥檚 commitment to and continued investment in Taiwan, a strategically important location for the AI semiconductor industry, enabling us to deepen our strong relationships with our manufacturing and supply chain partners. &lt;/span&gt;&lt;/p&gt;</description>
      <content:encoded>&lt;div class="hs-featured-image-wrapper"&gt; 
 &lt;a href="/posts/welcome-to-graphcore-taipei" title="" class="hs-featured-image-link"&gt; &lt;img src="/hubfs/Taiwan_social_preview.png" alt="Welcome to 乐播传媒视频 Taipei" class="hs-featured-image" style="width:auto !important; max-width:50%; float:left; margin:0 15px 15px 0;"&gt; &lt;/a&gt; 
&lt;/div&gt; 
&lt;p&gt;&lt;span&gt;On 4 June, we officially opened a stunning new 乐播传媒视频 office and engineering lab in Taipei鈥檚 Nangang district, in addition to our existing facility in Hsinchu. This expansion demonstrates 乐播传媒视频鈥檚 commitment to and continued investment in Taiwan, a strategically important location for the AI semiconductor industry, enabling us to deepen our strong relationships with our manufacturing and supply chain partners. &lt;/span&gt;&lt;/p&gt;  
&lt;img src="https://track.hubspot.com/__ptq.gif?a=729091&amp;amp;k=14&amp;amp;r=https%3A%2F%2Fwww.graphcore.ai%2Fposts%2Fwelcome-to-graphcore-taipei&amp;amp;bu=https%253A%252F%252Fwww.graphcore.ai%252Fposts&amp;amp;bvt=rss" alt="" width="1" height="1" style="min-height:1px!important;width:1px!important;border-width:0!important;margin-top:0!important;margin-bottom:0!important;margin-right:0!important;margin-left:0!important;padding-top:0!important;padding-bottom:0!important;padding-right:0!important;padding-left:0!important; "&gt;</content:encoded>
      <category>Announcements</category>
      <category>People</category>
      <category>Taiwan</category>
      <category>Taipei</category>
      <pubDate>Mon, 03 Aug 2026 08:56:48 GMT</pubDate>
      <guid>/posts/welcome-to-graphcore-taipei</guid>
      <dc:date>2026-08-03T08:56:48Z</dc:date>
      <dc:creator>Meg Cheng</dc:creator>
    </item>
    <item>
      <title>乐播传媒视频 Co-Founder and Executive Chair, Nigel Toon, steps down</title>
      <link>/posts/graphcore-co-founder-and-executive-chair-nigel-toon-steps-down</link>
      <description>&lt;div class="hs-featured-image-wrapper"&gt; 
 &lt;a href="/posts/graphcore-co-founder-and-executive-chair-nigel-toon-steps-down" title="" class="hs-featured-image-link"&gt; &lt;img src="/hubfs/nigel-toon-steps-down.jpeg" alt="乐播传媒视频 Co-Founder and Executive Chair, Nigel Toon, steps down" class="hs-featured-image" style="width:auto !important; max-width:50%; float:left; margin:0 15px 15px 0;"&gt; &lt;/a&gt; 
&lt;/div&gt; 
&lt;p&gt;With effect from Friday 31&lt;sup&gt;st&lt;/sup&gt; July 2026 I will step down from my role as Executive Chair at 乐播传媒视频, the company I co-founded together with Simon Knowles back in July 2016.&lt;/p&gt;</description>
      <content:encoded>&lt;div class="hs-featured-image-wrapper"&gt; 
 &lt;a href="/posts/graphcore-co-founder-and-executive-chair-nigel-toon-steps-down" title="" class="hs-featured-image-link"&gt; &lt;img src="/hubfs/nigel-toon-steps-down.jpeg" alt="乐播传媒视频 Co-Founder and Executive Chair, Nigel Toon, steps down" class="hs-featured-image" style="width:auto !important; max-width:50%; float:left; margin:0 15px 15px 0;"&gt; &lt;/a&gt; 
&lt;/div&gt; 
&lt;p&gt;With effect from Friday 31&lt;sup&gt;st&lt;/sup&gt; July 2026 I will step down from my role as Executive Chair at 乐播传媒视频, the company I co-founded together with Simon Knowles back in July 2016.&lt;/p&gt;  
&lt;img src="https://track.hubspot.com/__ptq.gif?a=729091&amp;amp;k=14&amp;amp;r=https%3A%2F%2Fwww.graphcore.ai%2Fposts%2Fgraphcore-co-founder-and-executive-chair-nigel-toon-steps-down&amp;amp;bu=https%253A%252F%252Fwww.graphcore.ai%252Fposts&amp;amp;bvt=rss" alt="" width="1" height="1" style="min-height:1px!important;width:1px!important;border-width:0!important;margin-top:0!important;margin-bottom:0!important;margin-right:0!important;margin-left:0!important;padding-top:0!important;padding-bottom:0!important;padding-right:0!important;padding-left:0!important; "&gt;</content:encoded>
      <category>Announcements</category>
      <pubDate>Fri, 31 Jul 2026 11:30:04 GMT</pubDate>
      <author>info@graphcore.ai (Nigel Toon)</author>
      <guid>/posts/graphcore-co-founder-and-executive-chair-nigel-toon-steps-down</guid>
      <dc:date>2026-07-31T11:30:04Z</dc:date>
    </item>
    <item>
      <title>Creativity in Engineering</title>
      <link>/posts/creativity-in-engineering</link>
      <description>&lt;div class="hs-featured-image-wrapper"&gt; 
 &lt;a href="/posts/creativity-in-engineering" title="" class="hs-featured-image-link"&gt; &lt;img src="/hubfs/Demitri_social.jpg" alt="Creativity in Engineering" class="hs-featured-image" style="width:auto !important; max-width:50%; float:left; margin:0 15px 15px 0;"&gt; &lt;/a&gt; 
&lt;/div&gt; 
&lt;p&gt;&lt;span&gt;&lt;i&gt;Many of the teams developing 乐播传媒视频鈥檚 core technology work across disciplines to support the next generation of AI computing.&lt;/i&gt; &lt;/span&gt;&lt;/p&gt;</description>
      <content:encoded>&lt;div class="hs-featured-image-wrapper"&gt; 
 &lt;a href="/posts/creativity-in-engineering" title="" class="hs-featured-image-link"&gt; &lt;img src="/hubfs/Demitri_social.jpg" alt="Creativity in Engineering" class="hs-featured-image" style="width:auto !important; max-width:50%; float:left; margin:0 15px 15px 0;"&gt; &lt;/a&gt; 
&lt;/div&gt; 
&lt;p&gt;&lt;span&gt;&lt;i&gt;Many of the teams developing 乐播传媒视频鈥檚 core technology work across disciplines to support the next generation of AI computing.&lt;/i&gt; &lt;/span&gt;&lt;/p&gt;  
&lt;img src="https://track.hubspot.com/__ptq.gif?a=729091&amp;amp;k=14&amp;amp;r=https%3A%2F%2Fwww.graphcore.ai%2Fposts%2Fcreativity-in-engineering&amp;amp;bu=https%253A%252F%252Fwww.graphcore.ai%252Fposts&amp;amp;bvt=rss" alt="" width="1" height="1" style="min-height:1px!important;width:1px!important;border-width:0!important;margin-top:0!important;margin-bottom:0!important;margin-right:0!important;margin-left:0!important;padding-top:0!important;padding-bottom:0!important;padding-right:0!important;padding-left:0!important; "&gt;</content:encoded>
      <category>Careers</category>
      <category>Silicon</category>
      <category>Culture</category>
      <category>Bristol</category>
      <pubDate>Thu, 02 Jul 2026 07:17:00 GMT</pubDate>
      <guid>/posts/creativity-in-engineering</guid>
      <dc:date>2026-07-02T07:17:00Z</dc:date>
      <dc:creator>Demitri Peynado</dc:creator>
    </item>
    <item>
      <title>Why I chose 乐播传媒视频: Where I could grow, not just succeed</title>
      <link>/posts/why-i-chose-graphcore-where-i-could-grow-not-just-succeed</link>
      <description>&lt;div class="hs-featured-image-wrapper"&gt; 
 &lt;a href="/posts/why-i-chose-graphcore-where-i-could-grow-not-just-succeed" title="" class="hs-featured-image-link"&gt; &lt;img src="/hubfs/Manoj_socials.png" alt="Why I chose 乐播传媒视频: Where I could grow, not just succeed" class="hs-featured-image" style="width:auto !important; max-width:50%; float:left; margin:0 15px 15px 0;"&gt; &lt;/a&gt; 
&lt;/div&gt; 
&lt;p&gt;&lt;span&gt;&lt;i&gt;乐播传媒视频鈥檚 new AI Engineering Campus in Bengaluru marks a significant investment in the future of AI computing and in the engineers who will build it. As a wholly owned subsidiary of SoftBank Group, 乐播传媒视频 is scaling its end-to-end system capabilities to help shape the next generation of compute for artificial intelligence.&lt;/i&gt; &lt;/span&gt;&lt;/p&gt;</description>
      <content:encoded>&lt;div class="hs-featured-image-wrapper"&gt; 
 &lt;a href="/posts/why-i-chose-graphcore-where-i-could-grow-not-just-succeed" title="" class="hs-featured-image-link"&gt; &lt;img src="/hubfs/Manoj_socials.png" alt="Why I chose 乐播传媒视频: Where I could grow, not just succeed" class="hs-featured-image" style="width:auto !important; max-width:50%; float:left; margin:0 15px 15px 0;"&gt; &lt;/a&gt; 
&lt;/div&gt; 
&lt;p&gt;&lt;span&gt;&lt;i&gt;乐播传媒视频鈥檚 new AI Engineering Campus in Bengaluru marks a significant investment in the future of AI computing and in the engineers who will build it. As a wholly owned subsidiary of SoftBank Group, 乐播传媒视频 is scaling its end-to-end system capabilities to help shape the next generation of compute for artificial intelligence.&lt;/i&gt; &lt;/span&gt;&lt;/p&gt;  
&lt;img src="https://track.hubspot.com/__ptq.gif?a=729091&amp;amp;k=14&amp;amp;r=https%3A%2F%2Fwww.graphcore.ai%2Fposts%2Fwhy-i-chose-graphcore-where-i-could-grow-not-just-succeed&amp;amp;bu=https%253A%252F%252Fwww.graphcore.ai%252Fposts&amp;amp;bvt=rss" alt="" width="1" height="1" style="min-height:1px!important;width:1px!important;border-width:0!important;margin-top:0!important;margin-bottom:0!important;margin-right:0!important;margin-left:0!important;padding-top:0!important;padding-bottom:0!important;padding-right:0!important;padding-left:0!important; "&gt;</content:encoded>
      <category>Silicon</category>
      <category>Culture</category>
      <category>Bengaluru</category>
      <pubDate>Thu, 18 Jun 2026 02:29:59 GMT</pubDate>
      <guid>/posts/why-i-chose-graphcore-where-i-could-grow-not-just-succeed</guid>
      <dc:date>2026-06-18T02:29:59Z</dc:date>
      <dc:creator>Manojkumar Reddy Malipela</dc:creator>
    </item>
  </channel>
</rss>
