|
27 | 27 |
|
28 | 28 |
|
29 | 29 |
|
30 | | - <title>Supporting Intel Gaudi accelerators - dstack</title> |
| 30 | + <title>Supporting Intel Gaudi AI accelerators - dstack</title> |
31 | 31 |
|
32 | 32 |
|
33 | 33 |
|
|
87 | 87 |
|
88 | 88 |
|
89 | 89 | <meta property="og:type" content="website" /> |
90 | | -<meta property="og:title" content="Supporting Intel Gaudi accelerators - dstack" /> |
| 90 | +<meta property="og:title" content="Supporting Intel Gaudi AI accelerators - dstack" /> |
91 | 91 | <meta property="og:description" content="dstack now supports Intel Gaudi accelerators with SSH fleets, simplifying container orchestration across private clouds and data centers." /> |
92 | 92 | <meta property="og:image" content=" |
93 | 93 | https://github.com/dstackai/static-assets/blob/main/static-assets/images/dstack-intel-gaudi-and-intel-tiber-cloud.png-v2?raw=true" /> |
|
96 | 96 | <meta property="og:image:height" content="630" /> |
97 | 97 | <meta property="og:url" content="https://dstack.ai/blog/intel-gaudi/" /> |
98 | 98 | <meta property="twitter:card" content="summary_large_image" /> |
99 | | -<meta property="twitter.title" content="Supporting Intel Gaudi accelerators - dstack" /> |
| 99 | +<meta property="twitter.title" content="Supporting Intel Gaudi AI accelerators - dstack" /> |
100 | 100 | <meta property="twitter:description" content="dstack now supports Intel Gaudi accelerators with SSH fleets, simplifying container orchestration across private clouds and data centers." /> |
101 | 101 | <meta property="twitter:image" content=" |
102 | 102 | https://github.com/dstackai/static-assets/blob/main/static-assets/images/dstack-intel-gaudi-and-intel-tiber-cloud.png-v2?raw=true" /> |
|
119 | 119 | <div data-md-component="skip"> |
120 | 120 |
|
121 | 121 |
|
122 | | - <a href="#supporting-intel-gaudi-accelerators" class="md-skip"> |
| 122 | + <a href="#supporting-intel-gaudi-ai-accelerators" class="md-skip"> |
123 | 123 | Skip to content |
124 | 124 | </a> |
125 | 125 |
|
|
173 | 173 | <div class="md-header__topic" data-md-component="header-topic"> |
174 | 174 | <span class="md-ellipsis"> |
175 | 175 |
|
176 | | - Supporting Intel Gaudi accelerators |
| 176 | + Supporting Intel Gaudi AI accelerators |
177 | 177 |
|
178 | 178 | </span> |
179 | 179 | </div> |
|
3473 | 3473 |
|
3474 | 3474 |
|
3475 | 3475 |
|
3476 | | -<h1 id="supporting-intel-gaudi-accelerators">Supporting Intel Gaudi accelerators<a class="headerlink" href="#supporting-intel-gaudi-accelerators" title="Permanent link">¶</a></h1> |
| 3476 | +<h1 id="supporting-intel-gaudi-ai-accelerators">Supporting Intel Gaudi AI accelerators<a class="headerlink" href="#supporting-intel-gaudi-ai-accelerators" title="Permanent link">¶</a></h1> |
3477 | 3477 | <p>At <code>dstack</code>, our goal is to make AI container orchestration simpler and fully vendor-agnostic. That’s why we support not |
3478 | 3478 | just leading cloud providers and on-prem environments but also a wide range of accelerators.</p> |
3479 | 3479 | <p>With our latest release, we’re adding support |
3480 | | -for <a href="https://www.intel.com/content/www/us/en/products/details/processors/ai-accelerators/gaudi.html" target="_blank">Intel Gaudi <span class="twemoji external"><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"><path d="m11.93 5 2.83 2.83L5 17.59 6.42 19l9.76-9.75L19 12.07V5z"/></svg></span></a> and |
3481 | | -launching a new partnership with Intel.</p> |
| 3480 | +for Intel Gaudi AI Accelerator and launching a new partnership with Intel.</p> |
3482 | 3481 | <p><img src="https://github.com/dstackai/static-assets/blob/main/static-assets/images/dstack-intel-gaudi-and-intel-tiber-cloud-v2.png?raw=true" width="630"/></p> |
3483 | 3482 | <!-- more --> |
3484 | 3483 |
|
3485 | 3484 | <h2 id="about-intel-gaudi">About Intel Gaudi<a class="headerlink" href="#about-intel-gaudi" title="Permanent link">¶</a></h2> |
3486 | | -<p>Intel Gaudi is a series of accelerators built to handle AI tasks. Powered by Intel’s Habana architecture, Gaudi is |
| 3485 | +<p>Intel Gaudi AI Accelerator is a series of accelerators built to handle AI tasks. Powered by Intel’s Habana architecture, Gaudi is |
3487 | 3486 | tailored for high-performance AI inference and training, offering high throughput and efficiency. It has a scalable |
3488 | 3487 | design with numerous cores and ample memory bandwidth, enabling better performance per watt.</p> |
3489 | 3488 | <p>Here's a brief spec for Gaudi 2 and Gaudi 3:</p> |
@@ -3640,8 +3639,17 @@ <h2 id="apply-a-configuration">Apply a configuration<a class="headerlink" href=" |
3640 | 3639 | </div> |
3641 | 3640 |
|
3642 | 3641 | <p><code>dstack</code> will automatically create containers according to the run configuration and execute them across the fleet.</p> |
3643 | | -<p>!!! info "Examples": |
3644 | | - Explore our <a href="../../examples/accelerators/intel/">examples</a> to learn how to train and deploy large models on Intel Gaudi.</p> |
| 3642 | +<blockquote> |
| 3643 | +<p>Explore our <a href="../../examples/accelerators/intel/">examples</a> to learn how to train and deploy large models on |
| 3644 | +Intel Gaudi AI Accelerator.</p> |
| 3645 | +</blockquote> |
| 3646 | +<div class="admonition info"> |
| 3647 | +<p class="admonition-title">Intel Tiber AI Cloud</p> |
| 3648 | +<p>At <code>dstack</code>, we’re grateful to be part of the Intel Liftoff program, which allowed us to access Intel Gaudi AI |
| 3649 | +accelerators via <a href="https://www.intel.com/content/www/us/en/developer/tools/tiber/ai-cloud.html" target="_blank">Intel Tiber AI Cloud <span class="twemoji external"><svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"><path d="m11.93 5 2.83 2.83L5 17.59 6.42 19l9.76-9.75L19 12.07V5z"/></svg></span></a>. |
| 3650 | +You can sign up if you’d like to access Intel Gaudi AI accelerators via the cloud.</p> |
| 3651 | +<p>Native integration with Intel Tiber AI Cloud is also coming soon to <code>dstack</code>.</p> |
| 3652 | +</div> |
3645 | 3653 | <div class="admonition info"> |
3646 | 3654 | <p class="admonition-title">What's next?</p> |
3647 | 3655 | <ol> |
|
0 commit comments