AMD Alveo V70 AI inference accelerator card

This is inference only. AMD should invest into the full AI stack starting from training. For this they need a product comparable to NVIDIA 4090, so that entry level researchers could use their hardware. Honestly, I don't know why AMD aren't doing that already, they are best positioned to do that in the industry landscape.

AMD XDNA – Versal AI Core &#x2F; 2nd-gen AIE-ML tiles<p>Are these programmable by the end-user? The &quot;software programmability&quot; section describes &quot;Vitis AI&quot; frameworks supported. But can we write our own software on these?<p>Is this card FPGA-based?<p>EDIT: [1] more info on the AI-engine tiles: scalar cores + &quot;adaptable hardware (FPGA?)&quot; + {AI+DSP}.<p>[1] <a href="https:&#x2F;&#x2F;www.xilinx.com&#x2F;products&#x2F;technology&#x2F;ai-engine.html" rel="nofollow">https:&#x2F;&#x2F;www.xilinx.com&#x2F;products&#x2F;technology&#x2F;ai-engine.html</a>

If this is based on fpga tech (xilinx) I don&#x27;t think it will have a cost&#x2F;benefit edge over asics. Why not do their own TPU like Google did? Nowadays even embedded MCUs come with AI accelerators (last I heard was 5TOPS in a banana pi-cm4 board - that is sufficient for object detection stuff and perhaps even more).

No price is listed on their site so I&#x27;m assuming its gonna be stupid expensive, but if anyone knows would you mind posting?

&gt; High-Density Video Decoder**: 96 channels of 1920x1080p<p>&gt; [...]<p>&gt; **: @10 fps, H.264&#x2F;H.265<p>Is 10 fps a standard measure for this kind of thing?

Every time I hear about an AI accelerator, I get really excited, then it turns out to be inference only.

Did I miss something did AMD buy Xilinx? Makes sense I suppose after Intel bought Altera.<p>Who owns lattice?

How much memory does it have?

What TOPS means exactly in &quot;... TOPS*|(INT8) 404 ...&quot; ?

Sad to see amds ROCm efforts essentially abandoned. They were close to universal interop for cudnn and cuda on amd (and other!) Architectures.<p>Hopefully Intel takes a stab at it with their ARC line out now.

Douglas Adams said we&#x27;d have robots to watch TV for us. That seems to be the designed use case for this.<p>16gb RAM &#x2F; 96 video channels ... I haven&#x27;t done any of that work but it feels like they expect that &quot;96&quot; not to be fully used in practice.

I won&#x27;t even take a look at the numbers unless they show a PyTorch model running on it, the problem is the big disconnect between HW and SW, realistically, have you ever seen any off-the shelf model running on something other than NVidia?