←back to thread

AMD unveils Ryzen Pro 8000-series processors

(www.tomshardware.com)
172 points marban | 4 comments | 16 Apr 24 13:38 UTC | HN request time: 0s | source
Show context
bearjaws ◴[16 Apr 24 14:12 UTC] No.40052158[source]▶
The focus on TOPS seems a bit out of line with reality for LLMs. TOPs doesn't matter for LLMs if your memory bandwidth can't keep up. Since it doesn't have quad channel memory mentioned I guess it's still dual channel?

Even top of the line DDR5 is around 128GB/s vs a M1 at 400GB/s.

At the end of the day, it still seems like AI in consumer chips is chasing a buzzword, what is the killer feature?

On mobile there are image processing benefits and voice to text, translation... but on desktop those are no where near common use cases.

replies(3): >>40052204 #>>40052260 #>>40052353 #
futureshock ◴[16 Apr 24 14:26 UTC] No.40052353[source]▶
Upscaling for gaming or video.

Local context aware search

Offline Speech to text and TTS

Offline generation of clip art or stock images for document editing

Offline LLM that can work with your documents as context and access application and OS APIs

Improved enemy AI in gaming

Webcam effects like background removal or filters.

Audio upscaling and interpolation like for bad video call connections.

replies(2): >>40052491 #>>40052497 #
kanbankaren ◴[16 Apr 24 14:35 UTC] No.40052497[source]▶
All of this(except upscaling) is possible with iGPU/CPU without breaking a sweat?
replies(1): >>40052666 #
bayindirh ◴[16 Apr 24 14:47 UTC] No.40052666[source]▶
The things which doesn't make GPU to break a sweat has its own specialized (or semi-specialized) processing blocks on the GPU, too.
replies(1): >>40052781 #
1. kanbankaren ◴[16 Apr 24 14:55 UTC] No.40052781{3}[source]▶
I meant the current generation of GPUs that don't have any AI acceleration blocks.
replies(1): >>40052810 #
2. bayindirh ◴[16 Apr 24 14:57 UTC] No.40052810[source]▶
They are MATMUL machines by design already. They do not need to "accelerate" AI to begin with.

Their cores/shaders can be programmed to do that.

Also, name a current gen GPU which doesn't have video encoding/decoding capabilities/facilities in silicon, even ones which do not allow shaders to be used in this process for post-processing. It's impossible (to not to have these capabilities) at this point in time.

replies(1): >>40053026 #
3. kanbankaren ◴[16 Apr 24 15:13 UTC] No.40053026[source]▶
I was talking about AI blocks and you moved the goal post to video codec blocks.
replies(1): >>40053078 #
4. bayindirh ◴[16 Apr 24 15:17 UTC] No.40053078{3}[source]▶
No. I didn't move anything.

I said that the core (3D rendering hardware) of a GPU with shaders is the AI block already, and said that other tasks like video encoders have their own blocks, but still pull capabilities from the "core" to improve things.