Radeon RX Vega Discussion Thread

well there could be a bright side here. Nvidia pricing the gtx 1080ti at $699 could mean they expect competition and move as many cards as they can before vega hits the shelf. Smart move by them, I am thinking top end vega might be pretty competitive since I was expecting pricing of 1080ti to be 799 atleast. Nvidia will move them at a good rate before vega hits. Good for them, Raja can enjoy his shirts until then.
 
well there could be a bright side here. Nvidia pricing the gtx 1080ti at $699 could mean they expect competition and move as many cards as they can before vega hits the shelf. Smart move by them, I am thinking top end vega might be pretty competitive since I was expecting pricing of 1080ti to be 799 atleast. Nvidia will move them at a good rate before vega hits. Good for them, Raja can enjoy his shirts until then.
I don't think Vega is going to be competitive. AMD has shown us nothing to think it was only going to be slightly faster then a 1080. The 1080 price cut is the real response to Vega.
 
I don't think Vega is going to be competitive. AMD has shown us nothing to think it was only going to be slightly faster then a 1080. The 1080 price cut is the real response to Vega.

People said the same thing about Ryzen, And well it blew away everyones expectation away.

Now is Vega going to be good? I have no idea, all we have is 1 random benchmark in that AMD benchmark/game. No one knows the clockspeed or anything.
 
People said the same thing about Ryzen, And well it blew away everyones expectation away.

Now is Vega going to be good? I have no idea, all we have is 1 random benchmark in that AMD benchmark/game. No one knows the clockspeed or anything.
The thing is all the leaks and demos have showed Ryzen to be good. Vega has always been "we will be slightly better than 1080 but only in DX12 and Vulkan as we don't have enough engineers to optimize DX11 and OpenGL". There has been no indication from team red they have anything comparable to the 1080Ti.
 
The thing is all the leaks and demos have showed Ryzen to be good. Vega has always been "we will be slightly better than 1080 but only in DX12 and Vulkan as we don't have enough engineers to optimize DX11 and OpenGL". There has been no indication from team red they have anything comparable to the 1080Ti.

True, but there have been absolutely no leaks or even a slight hint at specs or shaders amount or anything about vega. All we had was a doom demo done in a sealed case. So yes there were leaks about ryzen but vega other than AMD demo of doom, there has been absolutely nothing concrete or even a leaked bench. RTG has literally locked up everything about vega.
 
The thing is all the leaks and demos have showed Ryzen to be good. Vega has always been "we will be slightly better than 1080 but only in DX12 and Vulkan as we don't have enough engineers to optimize DX11 and OpenGL". There has been no indication from team red they have anything comparable to the 1080Ti.

The first leaks I saw for Ryzen is was barely faster than Sandybridge-X. Again leaks can be misleading.
 
Why does AMD find this so hard to do?


MGCr8Vx.jpg

Maybe because they are fighting on 2 fronts against 2 companies with multiple times R&D, in a nerd centric market that are as demanding as my 3 year old child - yeah, come the fuck on AMD, give me what i want when i want it or i'll smear your brand all over the internet! And dont give me any fucking explanation about how difficult it is i dont give a shit!
 
  • Like
Reactions: Ronin
like this
This is what I've been worried about. Everyone was jerking off over the leaked VEGA gameplay "beating a 1080". Who cares about beating the 1080 when the 1080 ti will be out by the time VEGA finally launches!?


Also, yall know I called it months ago that the 1080 TI would have nearly identical specs to the TITAN X. same core count, TDP, memory bus.
 
This is what I've been worried about. Everyone was jerking off over the leaked VEGA gameplay "beating a 1080". Who cares about beating the 1080 when the 1080 ti will be out by the time VEGA finally launches!?


Also, yall know I called it months ago that the 1080 TI would have nearly identical specs to the TITAN X. same core count, TDP, memory bus.

Okay sure. Unless you can give me inside confirmation, I assume you were guessing and you got a good guess right.
 
Well if you watched the event there was a big difference between December demonstration and today's with Vega. A 30% difference (one or two games not sure) and HBCC is now working, in December it was just a Fiji driver. So now you have the hardware working more to design. That was a big peak that looks like not too many here picked up. Now how transparent that hardware feature is or how easy to incorporate games into it and if 30% is best case, avg etc. is anyone's guess. So basically AMD is still mysterious about Vega - but I give 50/50 chance of Vega beating 1080Ti hands down.

So why would Nvidia all of a sudden reduce pricing on cards that have zero competition and launch a Titan XP performing card, excuse me a faster than Titan XP performing card for $699 - $500 less than the Titan XP? Nvidia being gracious and feeling sorry for the little folks?

Looks like AMD is planning on a major jump in performance with Vega (reason why it is taking so long maybe) - more like a Volta jump in performance but before Volta. The problem is the damn wait, by then Nvidia could sell most of their GP 102 stock that is not fully enabled.
 
Maybe because they are fighting on 2 fronts against 2 companies with multiple times R&D, in a nerd centric market that are as demanding as my 3 year old child - yeah, come the fuck on AMD, give me what i want when i want it or i'll smear your brand all over the internet! And dont give me any fucking explanation about how difficult it is i dont give a shit!

lol, you think it's because of development difficulty?
AMD has had the faster card and chips at times. It's a poorly run company.

They get the most simple things wrong. That has nothing to do with Nvidia or Intel.
AMD can't market their product, so they can't expect the customer to understand what they're are trying to do.

Leave the emotion and excuses out of business. Results is all that matters.

but I give 50/50 chance of Vega beating 1080Ti hands down.

You used "50/50" and "hands down" in the same sentence.

So why would Nvidia all of a sudden reduce pricing on cards that have zero competition and launch a Titan XP performing card, excuse me a faster than Titan XP performing card for $699 - $500 less than the Titan XP? Nvidia being gracious and feeling sorry for the little folks?

Looks like AMD is planning on a major jump in performance with Vega (reason why it is taking so long maybe) - more like a Volta jump in performance but before Volta. The problem is the damn wait, by then Nvidia could sell most of their GP 102 stock that is not fully enabled.

What price was the official Ti MSRP before yesterday?

Here's some history:

780Ti $699
980Ti $649
1080Ti $699

There was no price drop. The Ti always has been positioned at that price.
The GTX 1080 is AMDs real concern, Nvidia can drop that down to $399 with rebates by the time Vega is released.

AMD didn't show anything yesterday. No specs, no framerates, no card, no launch date.
AMD and their fans don't get it.
 
Last edited:
Looks like AMD is planning on a major jump in performance with Vega (reason why it is taking so long maybe)

Isn't the only measurement we've got at this point in time is of Vega running Doom Vulkan on 4K @ 70fps, and based on that the speculation was that:
- it appears to fit somewhere between GTX 1080 and Titan XP in terms of performance
- may go higher once driver is optimised

That sounds like Vega could be roughly at the level of Titan XP, or slightly above it - which would mean similiar to GTX 1080 Ti.

But... and here's a long shot - what would happen when you combine Ryzen (6 or 8 core) with Vega, and driver that is fully optimised for DX12/Vulkan?
I'm just wondering whether AMD might be aiming to fully exploit multithreading capability on all fronts. (where did we hear that before? ;))

Just my wild speculation.... and wishful thinking.
 
Summer of Radeon. That's a hint about RX Vega release time ;)
 
  • Like
Reactions: noko
like this
Okay sure. Unless you can give me inside confirmation, I assume you were guessing and you got a good guess right.

He's always right though. "The prophet." I'll take his guesses over anyone else.
 
basically AMD is still mysterious about Vega - but I give 50/50 chance of Vega beating 1080Ti hands down.

Vega beating 1080ti? I say 0% chance.

So why would Nvidia all of a sudden reduce pricing on cards that have zero competition and launch a Titan XP performing card, excuse me a faster than Titan XP performing card for $699 - $500 less than the Titan XP? Nvidia being gracious and feeling sorry for the little folks?

The TI price isn't different from what its been. Just check history.

But see it from the other side, top Vega card may be 399$ now and with a tshirt? :)
 
Maybe because they are fighting on 2 fronts against 2 companies with multiple times R&D, in a nerd centric market that are as demanding as my 3 year old child - yeah, come the fuck on AMD, give me what i want when i want it or i'll smear your brand all over the internet! And dont give me any fucking explanation about how difficult it is i dont give a shit!

This is the risk that AMD/RTG has decided to take on for over a decade now. Why make excuses for a corporation that decided to waste a shit-ton of money for half of that decade by having idiot CEOs that only knew how to party. If AMD/RTG had used their party money to invest in R&D we might have seen nVidia with the 20% marketshare in the GPU market instead of the other way around. AMD pretty much almost fucked themselves out of business due to not knowing how to run a business for quite some time.

I welcome their return to competitiveness but I won't excuse them for fighting on two fronts as that is the choice AMD made when they bought ATI in 2006.
 
Vega/Ryzen bundles is where things can get interesting IMO, taking a reduction in margins to move product.

Intel and nvidia both generally have higher price/performance, and will hold the performance crown for a while longer.

If they can offer the performance of i5-6600k/gtx1070 at a lower price point, they lock in people to a new platform for 3+ years. Lots of strategery possibilities :). Sure I'd like them to compete with i7-6700k/gtx1080 but they have to take some baby steps
 
I think HBM is the future, I'm just not sure that future has come yet. Right now, GDDR5X is providing similar bandwidth, for a cheaper price, and no VRAM capacity constraints.

I do think HBM's time will come, but IMO it doesn't make the most logical sense, yet.
]
Agreed, the biggest "mistake" Nvidia made was in letting HBM slip from one gen to the next, and then next again(at least for consumer desktop). GDDR5X is rocking enough for now, and the memory compression algorithms are improving as well. Wish AMD would have had the $ or will to make GDDR5X variants of their Fury/Vega. I want AMD to succeed so bad, but there's a limit to my loyalty. I may be getting the 1080ti just b/c it's going to be here and not vaporware. Further, I'm a bit learly of AMD launch drivers. Would be nice if AMD learned from their many many delays/mistakes and had a "one up" card waiting in the wings. Wishful thinking.
 
The conference showed a big announcement and that was the HBCC working.
Deus Ex 100% minimum frame rate improvement which logically means a 50% increase in average with a proper bin distribution of framerate values.
Sooo, if the RX 480 had HBM and HBCC, the card would perform 50% better on average with frames never dropping below 2x the current minimum framerate. 2304 SP. Now if we take into account 4096 SP of Vega 10, the increase in performance is huge. We are talking up to 2.5x times or more (due to new architecture, not only improvement mem. hierarchy and extra SP) against the RX 480.

And that is without software optimization, driver enablement only.

They showed halving the available memory did not affect performance because the allocated size is not the used size, so the cache would allow for bigger data sets to go into the game.
 
Well if you watched the event there was a big difference between December demonstration and today's with Vega. A 30% difference (one or two games not sure) and HBCC is now working, in December it was just a Fiji driver. So now you have the hardware working more to design. That was a big peak that looks like not too many here picked up. Now how transparent that hardware feature is or how easy to incorporate games into it and if 30% is best case, avg etc. is anyone's guess. So basically AMD is still mysterious about Vega - but I give 50/50 chance of Vega beating 1080Ti hands down.

So why would Nvidia all of a sudden reduce pricing on cards that have zero competition and launch a Titan XP performing card, excuse me a faster than Titan XP performing card for $699 - $500 less than the Titan XP? Nvidia being gracious and feeling sorry for the little folks?

Looks like AMD is planning on a major jump in performance with Vega (reason why it is taking so long maybe) - more like a Volta jump in performance but before Volta. The problem is the damn wait, by then Nvidia could sell most of their GP 102 stock that is not fully enabled.


I give it like 0% chance lol, they would have shown Titan Pascal performance if they could and they haven't.

The conference showed a big announcement and that was the HBCC working.
Deus Ex 100% minimum frame rate improvement which logically means a 50% increase in average with a proper bin distribution of framerate values.
Sooo, if the RX 480 had HBM and HBCC, the card would perform 50% better on average with frames never dropping below 2x the current minimum framerate. 2304 SP. Now if we take into account 4096 SP of Vega 10, the increase in performance is huge. We are talking up to 2.5x times or more (due to new architecture, not only improvement mem. hierarchy and extra SP) against the RX 480.

And that is without software optimization, driver enablement only.

They showed halving the available memory did not affect performance because the allocated size is not the used size, so the cache would allow for bigger data sets to go into the game.

The HBCC nada, they limited Vega to use 2 GB of vram to show HBCC working, yep you won't see that type of performance increase when ya have 8gb to begin with. They are showing the capability of going over the frame buffer size, which we all know the performance hit that will be taken by graphics cards when that happen.


Back to Noko's quote.

They reduced prices and introduced new sku's. The 1080ti is still 700 bucks, that is higher than any Ti version before it. So they aren't reacting with the 1080ti, they are reacting with the price cuts, without knowing where Vega ends up. Vega needs to be at 225 watts with 1080ti performance for any one that wants a Vega now, what they have been showing so far is Vega at 1080 performance at something like ~ 250 watts so lets say 225 for argument purposes. How are they going to price that at 500 bucks when they use more power, they can't price it at 700 cause performance isn't there lol. So your ending up with something less than 500 bucks, for a GPU with HBM 2? each stack of HBM 2 is like 120 buck, so 240 bucks for the vram alone.
 
Last edited:
Vega needs to be at 225 watts with 1080ti performance for any one that wants a Vega now

It needs to be >= 1080 at $500 for me to purchase one. Otherwise I'm not upgrading from the dual 290s I have.

I'm not giving Jen-Hsun a red cent. Looking forward to giving the blue team the finger as well once my 1700X gets here.

Anti-competitive business practices can kick rocks.
 
  • Like
Reactions: Ronin
like this
I give it like 0% chance lol, they would have shown Titan Pascal performance if they could and they haven't.



The HBCC nada, they limited Vega to use 2 GB of vram to show HBCC working, yep you won't see that type of performance increase when ya have 8gb to begin with. They are showing the capability of going over the frame buffer size, which we all know the performance hit that will be taken by graphics cards when that happen.


Back to Noko's quote.

They reduced prices and introduced new sku's. The 1080ti is still 700 bucks, that is higher than any Ti version before it. So they aren't reacting with the 1080ti, they are reacting with the price cuts, without knowing where Vega ends up. Vega needs to be at 225 watts with 1080ti performance for any one that wants a Vega now, what they have been showing so far is Vega at 1080 performance at something like ~ 250 watts so lets say 225 for argument purposes. How are they going to price that at 500 bucks when they use more power, they can't price it at 700 cause performance isn't there lol. So your ending up with something less than 500 bucks, for a GPU with HBM 2? each stack of HBM 2 is like 120 buck, so 240 bucks for the vram alone.
Had to find that part again for HBCC, Dues Ex MD. Yep your right it was limited from the start with 2gb to show effects of limited memory without it I suppose. He pretty much indicated in a around about way 2gb will perform as good as 4gb with HBCC => which if I take that right means really no performance boost if you have the memory. So my conclusion was wrong and no performance increase if you are within the memory amount. So 0% chance is maybe right, my hands are up now (n).

Nvidia still has other cards that could be made as in a full version GP 102, a consumer GP 100 (maybe some of the not fully enabled GP 100 that didn't make the cut). Still they did do some significant price cuts after some hefty price hikes. Now the new pricing is more back to normal in other words - this may just to get the sells back up as well to move inventory. The 1080Ti should do very well for that price range.
 
The conference showed a big announcement and that was the HBCC working.
Deus Ex 100% minimum frame rate improvement which logically means a 50% increase in average with a proper bin distribution of framerate values.
Sooo, if the RX 480 had HBM and HBCC, the card would perform 50% better on average with frames never dropping below 2x the current minimum framerate. 2304 SP. Now if we take into account 4096 SP of Vega 10, the increase in performance is huge. We are talking up to 2.5x times or more (due to new architecture, not only improvement mem. hierarchy and extra SP) against the RX 480.

And that is without software optimization, driver enablement only.

They showed halving the available memory did not affect performance because the allocated size is not the used size, so the cache would allow for bigger data sets to go into the game.
That is how I took it but it was really a show of if the VRAM was not sufficient for the game in how HBCC would give performance about the same if you had enough VRAM. Meaning if you have the VRAM then there would be no performance difference. This was one continuous demo and he clarified at the end that it was limited to 2gb of VRam, one with HBCC on and the other off - not with both using all the VRam onboard and HBCC giving the 50% and 100% (that would be an utter game changer there). I would not call that being misleading just confusing.
 
Had to find that part again for HBCC, Dues Ex MD. Yep your right it was limited from the start with 2gb to show effects of limited memory without it I suppose. He pretty much indicated in a around about way 2gb will perform as good as 4gb with HBCC => which if I take that right means really no performance boost if you have the memory. So my conclusion was wrong and no performance increase if you are within the memory amount. So 0% chance is maybe right, my hands are up now (n).
Even if you have enough memory, data is not always there. Stuttering/popping happens. HBCC eliminates or severely mitigates that. It allows to always have the data there, and to have scenes with more data (bigger textures, bigger scenarios). It is the same effect caches on CPUs with good prefetchers have.
 
Last edited:
That is how I took it but it was really a show of if the VRAM was not sufficient for the game in how HBCC would give performance about the same if you had enough VRAM. Meaning if you have the VRAM then there would be no performance difference. This was one continuous demo and he clarified at the end that it was limited to 2gb of VRam, one with HBCC on and the other off - not with both using all the VRam onboard and HBCC giving the 50% and 100% (that would be an utter game changer there). I would not call that being misleading just confusing.

Even if you have enough memory, data is not always there. Stuttering happens. HBCC eliminates stuttering. It allows to always have the data there, and have scenes with more data (bigger textures, bigger scenarios). It is the same effect caches on CPUs have with good prefetchers


If it needs to be fetched from the system ram, and that will only happen on rare occasions, Fiji had no major problems with this either till it was over like 1.5 gb, like when pushing past the vram ram limit by more than 2 fold, it would have been nice to see what frame rates they were getting to begin with, but I would think is less than 30 fps it looked to be in the teens, which by no means is good. So a 50 or even 100% increase is just ok, still hitting that 30FPS base line.

They also mentioned this was already on consoles, so I am thinking Polaris has this capability yet we haven't seen it?
 
Last edited:
The conference showed a big announcement and that was the HBCC working.
Deus Ex 100% minimum frame rate improvement which logically means a 50% increase in average with a proper bin distribution of framerate values.
Sooo, if the RX 480 had HBM and HBCC, the card would perform 50% better on average with frames never dropping below 2x the current minimum framerate. 2304 SP. Now if we take into account 4096 SP of Vega 10, the increase in performance is huge. We are talking up to 2.5x times or more (due to new architecture, not only improvement mem. hierarchy and extra SP) against the RX 480.

And that is without software optimization, driver enablement only.

They showed halving the available memory did not affect performance because the allocated size is not the used size, so the cache would allow for bigger data sets to go into the game.

Keep in mind the card was artificially bottlenecked to 2GB of physical VRAM.
Will it really show that much difference at 4GB, or 8GB of physical VRAM? Yet to be seen.
 
All I can say is Nvidia is doing what they are doing because of market share. They don't have to do anything, they aren't doing much of this because of Vega. They are doing this because they want people to upgrade from their older Nvidia hardware. The initial price tag was for people like myself who do not care the cost and just want the fastest card, now they are targeting the group that are more budget minded. If Nvidia was doing this because of Vega, they wouldn't be doing this Titan X BS for a couple years now.
 
People said the same thing about Ryzen, And well it blew away everyones expectation away.

Now is Vega going to be good? I have no idea, all we have is 1 random benchmark in that AMD benchmark/game. No one knows the clockspeed or anything.
Difference with Ryzen to Intel is that Intel kept increasing the performance by >5%. NVIDIA on the other hand keeps outdoing their own cards by <25%.
 
Who cares about beating the 1080 when the 1080 ti will be out by the time VEGA finally launches!?

Whoah there. I care. I expect a lot of gamers don't have $700+ to spend on a video card. But if Vega parts can outperform the 1080 and 1070 for same price, or match it for lower price, I think a lot of people will be interested in those.

Vega needs to be at 225 watts with 1080ti performance

Well, what it actually needs to be is a good price/performance ratio. If their top end doesn't touch the 1080ti, it just means they don't have the top end. Right now they're only a good value sub $200, which is a good chunk of the market, but if they can become a good value sub $500, that's basically the entire market.

It needs to be >= 1080 at $500 for me to purchase one.

yeah, match or exceed 1070/1080 performance and match/beat 1070/1080 price is going to be key for Vega parts. If they can do that they'll be in decent shape.
 
It needs to be >= 1080 at $500 for me to purchase one. Otherwise I'm not upgrading from the dual 290s I have.

I'm not giving Jen-Hsun a red cent. Looking forward to giving the blue team the finger as well once my 1700X gets here.

Anti-competitive business practices can kick rocks.
Just seems like little boy anger. Because if anything nVidia is guilty of being TOO competitive.

Time and again Jen Hsun gets up on a stage somewhere and pulls a monster out of his pocket that will be available next week, while at AMD it's just "wait, wait, wait. Wait for Fiji. Wait for Vega.."
 
Whoah there. I care. I expect a lot of gamers don't have $700+ to spend on a video card. But if Vega parts can outperform the 1080 and 1070 for same price, or match it for lower price, I think a lot of people will be interested in those.

Well, what it actually needs to be is a good price/performance ratio. If their top end doesn't touch the 1080ti, it just means they don't have the top end. Right now they're only a good value sub $200, which is a good chunk of the market, but if they can become a good value sub $500, that's basically the entire market.


How do you expect this, HBM 2 just one stack of 4 gb (which Vega has 2 stacks) is around 120 bucks, this board is going to cost just to make around 350 bucks, they will not have enough margins to sustain that kind of business.


Then you factor in power consumption which is a big thing, may be not to you, but over all when people see that a 225 watt card competing against a 180 watt card or less that is a big difference, 20%.

So that's two strikes against it right off the bat. And if performance is in the realm of the 1070 and 1080, then we got one walk and two strikes.

We saw what power consumption did with Fermi sales, we saw what it did with x1800, x1900 sales.

Performance isn't everything.
 
My Newegg CC awaits for either Vega or the new 1080ti. Come on [H] with the reviews....
 
How do you expect this, HBM 2 just one stack of 4 gb (which Vega has 2 stacks) is around 120 bucks, this board is going to cost just to make around 350 bucks, they will not have enough margins to sustain that kind of business.


Then you factor in power consumption which is a big thing, may be not to you, but over all when people see that a 225 watt card competing against a 180 watt card or less that is a big difference, 20%.

So that's two strikes against it right off the bat. And if performance is in the realm of the 1070 and 1080, then we got one walk and two strikes.

We saw what power consumption did with Fermi sales, we saw what it did with x1800, x1900 sales.

Performance isn't everything.


Meh doesn't make sense. that seems like a insanely high cost to make a card and get it out the door. I would assume the card to cost 200-250 max for amd to make any reasonable profits. Or we are looking at one hbm2 card and rest probably with gddr5 or 5x
 
Meh doesn't make sense. that seems like a insanely high cost to make a card and get it out the door. I would assume the card to cost 200-250 max for amd to make any reasonable profits. Or we are looking at one hbm2 card and rest probably with gddr5 or 5x


The chip alone is around 100 bucks, the memory 240 for both stacks, that's 340 bucks man. Then start factoring in the cost of the interposer, labor, and other components. Kinda stuck at a rock and hard place to get it below 400, so Top end Vega at 500, is still at the 30% margin range....which is not enough.
 
Even if Vega is faster than 1080, but not as good as 1080ti, Nvidia still will have the upper hand with the faster release and adoption rate. People aren't going to wait for the slower card even if it's cheaper. Not in a world of instant gratification. If they wait too long to release Vega it'll be dead in the water before the boat is launched.
 
razor1 We haven't gotten details of any Vega parts yet, so this is all guesswork on my part. I guess that wasn't clear? I'm just going based on the typical back-and-forth between AMD and NVIDIA over the last few years, factoring in that AMD hasn't had the performance lead for a while and is targeting more of the mainstream now, and making an educated guess.

And if "performance isn't everything", are you saying you'd really spend an extra $50 on a slower card just because it was rated for a lower TDP?

There's a handy power calculator here that I've used before to try and figure this out, but let's say we've got a hypothetical Radeon Vega 570 which is exactly the same speed as a GTX 1070 but used 45W more power and costs $50 less. At 12c kWhr, for 8 hours of gaming 365 days/yr that means the Vega would cost you an extra $15 a year, assuming it's even at peak load that entire time.

I would buy the cheaper card, in this instance, because while "performance isn't everything", I didn't say that. I said a good price/performance ratio is what they need.

(Also I'm curious to know where you're getting your manufacturing costs from.)

Anyways, I wouldn't be too surprised at this point to see them release something with GDDR5X if it would keep costs down, and save HBM2 for the top end, sort of like what they did with Fury when it launched - we got the Fiji with HBM and everything else was using normal memory (granted it was also old tech).

Maybe Vega 11 won't use HBM and Vega 10 will? We don't know! That's what has been so maddening about this whole thing. AMD could take five minutes and answer a lot of these questions and it would help a whole lot.

[edit] - fixed typo at the end where I mixed up Vega 11 and Vega 10.
 
Last edited by a moderator:
Back
Top