As someone who does a good deal of editing, that is impressive. I dare say it will be a bit out of my price range though
Indeed. Even if you used an M.2 SSD as a scratch disk it wouldn't perform anywhere near what this GPU can do.
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
AMD announce Radeon Pro SSG
- Thread starter SiDeards73
- Start date
Indeed. Even if you used an M.2 SSD as a scratch disk it wouldn't perform anywhere near what this GPU can do.
Soldato
It's nice to see AMD bringing in new things that help them get back in the game.
Soldato
It was mainly Orangey posting a complaint about how AMD fans always say the same tired lines followed by a succession of people posting praise for the product with excessively fancy word choices. Fairly light stuff, I thought - but I guess it was read as trash talk or something.
EDIT: Wow. I didn't get a chance to watch the actual video before. That is amazing. I love the way Raja says "buddery smooth 30Hz".
Seriously - he is awesome and a true engineer. It seemed so casual when he just called over to the person running the demo: "hey, why don't you show them the scrubbing". He's really in his element with something like this.
Last edited:
Associate
- Joined
- 24 May 2011
- Posts
- 1,541
As someone who does a good deal of editing, that is impressive. I dare say it will be a bit out of my price range though
You do a great deal of 8k editing do you? No didn't think so.... This card with 10k real time decode and playback is aimed at the Avid, Resolve, Final Cut X devs not editors... and when it is finally out of beta it would still not be aimed at some guy that edits his personal YouTube videos for a living. It's aimed at pro editors that make money out of their craft.
Heck I get paid for editing 4.6k material and it still wouldn't be aimed at me!
Last edited:
Soldato
+1. Mind = blown.
Caporegime
LMAO, you really have a problem with anything I say and go out of your way to rip apart and twist anything. You need to chill down a bit fella and don't get so wound up. I don't do any 8K editing as it goes for clarification.
Associate
- Joined
- 24 May 2011
- Posts
- 1,541
Hey easy man was simply stating you don't do 8k editing and don't need the card so there is no need to comment on the fact you can't afford it. I mean obviously you can't as its aimed a big companies and not a lone youtuber... I just don't see the issue?
I would also love to know what you edit on?
PC per has updated their article on this story with the following update.
Interestingly it turns out it is a Fiji GPU using the same bridge chip the Pro duo uses to access a pair of Samsung 950pro SSD's in a raid config.
Overall a very impressive idea and the prototype seems to work well. A bit of a shame that software needs to be coded for this sort of setup, but I'm sure if AMD go ahead with it, NVidia will also use a similar setup and the software would then quickly follow.
UPDATE (July 27th, 1am ET): More information on the Radeon Pro SSG has surfaced since the original article. According to AnandTech, the prototype graphics card actually uses an AMD Fiji GPU. The Fiji GPU is paired onboard PCI-E based storage using the same PEX8747 bridge chip used in the Radeon Pro Duo. Storage is handled by two PCI-E 3.0 x4 M.2 slots that can accommodate up to 1TB of NAND flash storage. As I mentioned below, having the storage on board the graphics card vastly reduces latency by reducing the number of hops and not having to send requests out to the rest of the system. AMD had more numbers to share following their demo, however.
From the 8K video editing demo, the dual Samsung 950 Pro PCI-E SSDs (in RAID 0) on board the Radeon Pro SSG hit 4GB/s while scrubbing through the video. That same video source stored on a Samsung 950 Pro attached to the motherboard had throughput of only 900MB/s. In theory, reaching out to system RAM still has raw throughput advantages (with DDR4 @ 3200 MHz on a Haswell-E platform theroretically capable of 62 GB/s reads and 47 GB/s writes though that would be bottlenecked by the graphics card having to go over the PCI-E 3.0 x16 link and it's maximum of 15.754 GB/s.). Of course if you can hold it in (much smaller) GDDR5 (300+GB/s depending on clocks and memory bus width) or HBM (1TB/s) and not have to go out to any other storage tier that's ideal but not always feasible especially in the HPC world.
However, having onboard storage on the same board as the GPU only a single "hop" away vastly reduces latency and offers much more total storage space than most systems have in DDR3 or DDR4. In essence, the solid state storage on the graphics card (which developers will need to specifically code for) acts as a massive cache for streaming in assets for data sets and workloads that are highly impacted by latency. This storage is not the fastest, but is the next best thing for holding active data outside of GDDR5/x or HBM. For throughput intensive workloads reaching out to system RAM will be better Finally, reaching out to system attached storage should be the last resort as it will be the slowest and most latent. Several commentors mentioned using a PCI-E based SSD in a second slot on the motherboard accessed much like GPUs in CrossFire communicate now (DMA over the PCI-E bus) which is an interesting idea that I had not considered.
Per my understanding of the situation, I think that the on board SSG storage would still be slightly more beneficial than this setup but it would get you close (I am assuming the GPU would be able to directly interact and request data from the SSD controller and not have to rely on the system CPU to do this work but I may well be mistaken. I will have to look into this further and ask the experts heh). On the prototype Radeon Pro SSG the M.2 slots are actually able to be seen as drives by the system and OS so it is essentially acting as if there was a PCI-E adapter card in a slot on the motherboard holding those drives but that may not be the case should this product actually hit the market. I do question their choice to go with Fiji rather than Polaris, but it sounds like they built the prototype off of the Radeon Pro Duo platform so I suppose it would make sense there.
Hopefully the final versions in 2017 or beyond use at least Vega though.
Interestingly it turns out it is a Fiji GPU using the same bridge chip the Pro duo uses to access a pair of Samsung 950pro SSD's in a raid config.
Overall a very impressive idea and the prototype seems to work well. A bit of a shame that software needs to be coded for this sort of setup, but I'm sure if AMD go ahead with it, NVidia will also use a similar setup and the software would then quickly follow.
Caporegime
I have no time for your pedantic posting and no wish to answer any of your questions. I praised up the card and made a comment about it being out of my price range and that is that.
Associate
This explains Nvidia's announcement on their new pro cards as quickly as they did - this for a few industries is a game changer....
Associate
- Joined
- 24 May 2011
- Posts
- 1,541
Wow easy man was only asking a question no need to be rude!
Soldato
That came across a little strong, but since we talking about it being in infancy and the fact that if you had the dollar you could record and edit in 8k at home now (camera being about £5k) it will become more relevant in a few years.
Things start somewhere and we are discussing tech. The principle is you could still get something that works similar for editing 1080p to 4k and still speed up the process and that would just come down to return of investment.
We render 4k walk through of buildings for large clients and these can take a few days in the office to render 20 mins of footage so these would also be applicable for us as 2 days of a machine in downtime and then having to redo to adjust is expensive.
Soldato
Interesting. Though it is a devkit. There's a good chance, I would have thought, that you'd be able to use this to develop against and then if they release a Vega-based version for production, it will have served well enough to develop against.
Yup this one is just a prototype, for proof of concept. Now the idea is there, they can take it forward with newer chips, it would probably be a while before any consumer product is ready anyway, not that it is aimed at us anyway.
It is just a great idea.
Caporegime
I have a felling we may see a technology like this in gaming.
Its not a big step from here to stick a 30GB high speed SSD directly on to the GPU's PCB for Texture storage.
Texture pre caching has a significant performance advantage and clearly AMD have the IP to make that happen.
Interesting times ahead.
Its not a big step from here to stick a 30GB high speed SSD directly on to the GPU's PCB for Texture storage.
Texture pre caching has a significant performance advantage and clearly AMD have the IP to make that happen.
Interesting times ahead.
That is certainly a possibility and with 128GB NVME drives coming in at about £60, 32GB of flash shouldn't add too much to the cost of a graphics card.
Definitely interesting times ahead.
Soldato
Maybe AMD is already cooking something up? Might explain why they not jumping onto the highend GPU market this year. Would love to see this used on Desktop GPUs can see it helping out in high demanding situations.