Soldato
It's funny because the ones that are butt hurt about the joke, are the ones that play games on their computers. 
-
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 VEGA confirmed for 2017 H1
- Thread starter Doom112
- Start date
- Status
The dude in the Q&A literally asks about Volta and for Raja to tell him a joke.
People really getting sand in their bits over it lol. Some thinking it's a dig at Volta or something, to say the cards will need more Voltage to over clock.
Last edited:
Some just seem like they are panting while waiting for the opportunity for an AMD employee to slip up so they can pounce on them. lol.
First world problems
Doom will be along soon, should be entertaining
Soldato
I'm getting quite impatient now, it's been a long time waiting, and whilst I've always said in this thread that Vega will not be available until after July, the realisation of that is kind of a big bummer. I know it's not the 31st yet, I know nothing concrete has been said, but you can just read between the lines quite clearly on the gaming Vega.
AMD really do make it difficult for us, a little frustrating to say the least, a 1080ti purchase is really gnawing away at me here, Vega needs to hurry the hell up, I'm tired of waiting. Will wait just a little longer, but if it turns out gaming Vega is still 3 or so months away, that's going to be a killer.
AMD really do make it difficult for us, a little frustrating to say the least, a 1080ti purchase is really gnawing away at me here, Vega needs to hurry the hell up, I'm tired of waiting. Will wait just a little longer, but if it turns out gaming Vega is still 3 or so months away, that's going to be a killer.
Why not turn that animosity towards their HBM2 suppliers. Remember they promised mass production by Q3 2016.
Back then Vega looked like it matched a highly over clocked GTX 1080.
Many would have been happy if that Vega launched in Q4 2016, yet since then the suppliers have reduced their HBM2 catalogue, and moved up the date of the slowest modules to Q2 2017 for mass production.
You really don't know that. There's no info to suggest that, and all they've said is 1H 2017, which is until the end of June. We've seen AMD launch cards with "will be shipping to next week".
We've waited this long, might as well wait till Computex and find out what is going on, it's only a couple of weeks away. I suspect Vega is much closer than you think, and you should wait because even if you go green, Vega will cause Nvidia to cut prices.
Soldato
I wouldn't call it animosity mate, more so just frustration always playing the waiting game.
Yeah I'm still waiting, and like I said I'll wait until we have definite word on what's what, I like many others have been waiting and waiting and waiting but as the finish line gets closer it almost becomes harder. The thing that worries me more than the release date, is the amount of stock. I imagine it's going to be FuryX levels of ordering frustration all over but double the wait.
Last edited:
This is AMD's product and end of the day its their issue and customers waiting if it turns up late.
One of those things imo if HBM2 was new and not fully available maybe they should have also looked at GDDR5 alternatives.
i will say iirc it was always due in 2017 H1 and that times not up yet so case of wait and see before people assume its late
Hardly AMDs fault if their partners have failed to deliver for what is nearly a full year.
They certainly don't have the money to suddenly create a new memory controller to use standard GDDR5 and replace the HBCC within that time.
At least it looks as though they've used that time to improve Vega as much as possible.
Originally matching a 1080,and now showing snippets at 4K games where its around overclocked 1080ti performance.
It's annoying to see they couldn't get the 512GB/s memory bandwidth, as a result of their partners either. Instead having to settling for lower speed modules.
it will be a similar situation to the titan vs the rest of the geforce range. Frontier will have more grunt but slightly worse pcb and components than the custom rx vega cards.
i am specuating that frontier will lack the voltage range that RX vega has and that's why raja is dropping hints. frontier is probably bios locked at a lower range than the RX vega whilst frontier has the lower cloking chip but more functionality enabled for better FP performance.
Soldato
I'm not calling it AMDs fault, I'm not actually placing blame on any one party, I'm just calling it frustrating waiting.
IMO worst case scenario in 8 weeks time you will have a Vega GPU in your machine, but I am willing to bet it will be in 4-6 weeks time.
GDDR5X was a surprise leap forward. Before than HBM looked much better.
some advantages still remain, tighter packaging, lower power consumption (allegedly)
Sure, but in the future AMD should take precautions not to put every egg in the same basket; they could have added a GDDR5X backup version in case of longer than expected lead times and released those as the "low-end" SKUs to compete with the gtx 1070 and kept HBM2 for GTX1080+ performance and up.
Soldato
It's quite clear that AMD don't have the vast resources as Nvidia and Intel. So they put all the eggs in to single baskets. Those baskets are what they think are going to be of most benefit in the future.
For example, they put all their time in to Bulldozer and Steamroller... at the time the market didn't need 8 core processors and the low IPC on the chips meant they where poor in comparison to Intel. But it set them up to build on it and now we have Ryzen which has pushed the envelop for consumer grade multi core cpu's. Hopefully will mean now that multi cores is more accessible to your average consumer games will follow too.
Same goes for why AMD put so much effort into DX12 and Vulkan. The 1000 series do not focus on DX12. Because for the most part we are not in the DX12 era yet. Lot of work to do. But AMD have already taken a march on Nvidia and are more ready than they are in both software and hardware support.
HBM2. Wasnt needed on the Fury cards, Nvidia went GDDR5X, but again it has laid the ground work for Vega.
For example, they put all their time in to Bulldozer and Steamroller... at the time the market didn't need 8 core processors and the low IPC on the chips meant they where poor in comparison to Intel. But it set them up to build on it and now we have Ryzen which has pushed the envelop for consumer grade multi core cpu's. Hopefully will mean now that multi cores is more accessible to your average consumer games will follow too.
Same goes for why AMD put so much effort into DX12 and Vulkan. The 1000 series do not focus on DX12. Because for the most part we are not in the DX12 era yet. Lot of work to do. But AMD have already taken a march on Nvidia and are more ready than they are in both software and hardware support.
HBM2. Wasnt needed on the Fury cards, Nvidia went GDDR5X, but again it has laid the ground work for Vega.
Soldato
That's the hope right there.
I don't think it was a pure business decision; Raja talks quite passionately about bringing the best of the best to the consumer, and I am sure the decision was made partly because of his vision of what the perfect GPU should be. Doubt they would have waited this long if it wouldn't pay off in the end (maybe they paid for the order upfront and couldn't cancel and switch to something else
). Wish AMD had been more open from the beginning - I really don't why they feel the need to keep things under wraps constantly (seems to be hurting them more than anything).
Soldato
- Joined
- 1 Jun 2014
- Posts
- 5,109
So can anyone tell me, in layman's terms, what Infinity Fabric will bring to GPUs?
Well, if GDDR5X would have allowed the card to be launched much quicker, then surely that would have been better for AMD in the long term. nvidia must have looked at the tech and thought it wasnt quite ready yet.
I do like the idea of smaller cards, as some cards have been getting pretty big. I doubt there is much difference in power-usage, as this would depend more on the gpu.
This all kind-of reminds me what happened with the 980ti vs the fury x. HBM vs GDDR5. However this time at-least we not looking at 4gb for a high-end card (AMD was trying to tell us that was more than enough for 4k - which it wasnt). What we wanted was a card to push nvidia line. If HBM2 forces the card to be a higher price, with little benefits then its going to be a shame, and nvidia will just keep the same pricing - maybe offer more performance for less money. I guess we have to wait to see what happens, but hopefully it wont be another repeat. Id like to see VEGA do well.
Navi, multiple gpu cores, seen as a single large GPU all sharing their HBM2 as High-Bandwidth Cache; and allowing for improved SoC of CPU+GPU.
Just like Ryzen is multiple CCX linked via Infinity Fabric, AMD's goal by Navi is the same for GPUs. Raja spoke about in in 2016 where their end goal is for Multi-GPU cores to be as common as multi-CPU.
Infinity Fabric runs at 30-50GB/s on Ryzen, but goes all the way up to 512GB/s on Naples/Epyc and GPUs.
https://youtu.be/4qJj1ViyyPY?t=4m4s
Last edited:
- Status