Vega Rumors

[razor1]

You should not buy a card from AMD or Nvidia based on their marketing. Wait until proper reviews come out. Anybody who buys a card based on Marketing slides from either company is an idiot.

You and I might not, can't say that about many other people, was on facebook the other day, and guy bought a FX series CPU thinking it was Ryzen.

See the the difference between us and others?

Marketing works on the weak, uneducated minds, and look at AMD's async marketing, still have so many people thinking that way. That is bad marketing, it uses peoples weakness because their products are crap, that is the only way they can sell their products right now, out right fooling people into believing BS.

Marketing is used to hide the fact the products are inferior. When a product is good, there is no need to market, just show that its better and in this market you are set.

There are no need for graphs that are skewed or cherry picked.

You can't say anything other then BS that's why you just made it personal, can't use facts in your statements shouldn't post man. Simple.

nV's marketing is way ahead of AMD because their products are that much better.

If you can't accept that, then go buy a Vega FE right now and save the trouble of waiting for RX. Oh who needs marketing? So far what AMD has shown seems to be real. 1080 gaming performance and better productivity performance than Titan X right? Its real there is nothing fake about it. Its good marketing. How many people are still saying "it's not a gaming card though!" That's AMD's marketing working for them. Its bad though, cause its going to back fire when RX is released and the reality is RX is the same as FE for the most part.

Marketing 101, you don't make promises you can't keep. AMD promised RX will be faster than FE in games, it will be, but not by much and being so little, its going to really cause major disappointment.

Marketing's JOB is to set expectations and to sell products. Not to bloat their own ego and create disappointment at the end.

How many fuckin times have we seen AMD do this? Count them for me I'll give you till midnight tonight. Then if you don't, I'll write them out for you. Just go through ever single launch they have ever done, you will see it started after A64 and never stopped, almost ever single one of their products came out the same way when it came to marketing. Lots of hot air and promises that never materialize.

Just go through ever single product lunch. If you want then go through nV's product launches, then go through Intel's. You will see a huge pattern with AMD, they do BS ever single launch doesn't matter if the product is good or bad. nV and Intel, only do that when they fuck up.

 

[THUMPer]

You and I might not, can't say that about many other people, was on facebook the other day, and guy bought a FX series CPU thinking it was Ryzen.

See the the difference between us and others?

Marketing works on the weak, uneducated minds, and look at AMD's async marketing, still have so many people thinking that way. That is bad marketing, it uses peoples weakness because their products are crap, that is the only way they can sell their products right now, out right fooling people into believing BS.

Marketing is used to hide the fact the products are inferior. When a product is good, there is no need to market, just show that its better and in this market you are set.

There are no need for graphs that are skewed or cherry picked.

You can't say anything other then BS that's why you just made it personal, can't use facts in your statements shouldn't post man. Simple.

nV's marketing is way ahead of AMD because their products are that much better.

If you can't accept that, then go buy a Vega FE right now and save the trouble of waiting for RX. Oh who needs marketing? So far what AMD has shown seems to be real. 1080 gaming performance and better productivity performance than Titan X right? Its real there is nothing fake about it. Its good marketing. How many people are still saying "it's not a gaming card though!" That's AMD's marketing working for them. Its bad though, cause its going to back fire when RX is released and the reality is RX is the same as FE for the most part.

Marketing 101, you don't make promises you can't keep. AMD promised RX will be faster than FE in games, it will be, but not by much and being so little, its going to really cause major disappointment.

Marketing's JOB is to set expectations and to sell products. Not to bloat their own ego and create disappointment at the end.

How many fuckin times have we seen AMD do this? Count them for me I'll give you till midnight tonight. Then if you don't, I'll write them out for you. Just go through ever single launch they have ever done, you will see it started after A64 and never stopped, almost ever single one of their products came out the same way when it came to marketing. Lots of hot air and promises that never materialize.

Man I can't wait for vega!

 

[razor1]

https://www.inc.com/josh-linkner/stop-under-promising-try-this-instead.html

On the topic of marketing.

https://www.livechatinc.com/blog/avoid-overpromising/

This is what AMD does ALL the TIME. Not by themselves, but by the message they send out from marketing, then others make it into a big ass pile of shit. But that is bad marketing, its AMD's marketer's JOB to set expectations, if they can't do it right, they shouldn't be there.

But this is tech, its based on math, AMD is well aware of their products and what they are capable of before they start manufacturing, added to that if they already know what their competition is up to well they should never need to over promise anything. There is nothing to push to get better, everything is known before hand.

For anyone to say AMD's marketing has been equal to nV or Intel, in the past decade and half has not been paying attention to what has been going on.

Look AMD doesn't need to go for the top dog to make compelling products, we saw that with Polaris right? Its not anywhere as good as the gtx 1060 when it comes to power usage, but its performance is, and its selling as good as the entire line up of AMD's previous graphics cards top to bottom.

now managing expectations

http://blog.cx-iq.com/strategic-marketing-of-managing-expectations/

Has anyone heard this before, if lets say the Fed says markets are going to go down for the next couple of months for reasons XYZ, no one will panic when they go down. But if they don't say anything and the markets start falling, then there is panic?

That is managing expectations. What ever anyone else says shouldn't matter, cause the guys with the balls stated its going to happen anyways.

AMD doesn't have the guys with the balls, they can't manage their "fan boys" expectations.

But even with Polaris, they didn't manage expectations, weeks before Polaris launch people were still saying its going to match a 1070.... really, a card that functions like a 970 was blown up to a 1070? That's kinda blown out of proportion like 100% out of proportion?

Vega, just a couple of months ago, weeks prior to Vega FE launch people were thinking 225 watts for 1080 level performance, and I can list out the people that were thinking that too lol, cause I remember having the discussions with them for close to 6 months. Do we know where that rumor came from, sure we do, Linus tech tips, taking a photo of Vega's power cables at a show. Why would AMD let that go, cause just the previous event, they hid the power cables to the card entirely lol. But it was + news for them so they let it go, and it was Linus, the biggest youtuber for tech. Great press for AMD, but its not the truth.

Where is the management of expectations?

They showed Doom and BF1 and Sniper Elite 4 and Star Wars Battlefront, everyone was saying Vega is faster than the 1080 and they still had a few months left to optimize drivers....

Same thing

Where was the management of expectations, just because something looks good to show and gives great press doesn't mean when other vocal parties start mushrooming those thoughts out of proportion means good marketing. The marketing will back fire.

 

[SighTurtle]

https://www.inc.com/josh-linkner/stop-under-promising-try-this-instead.html

Where is the management of expectations?

Its pretty clear AMD regards any hype, any talk about them to be good talk. Or a disconnect from marketing and engineering. RTG was supposed to solve this, but if Raja Koduri is going on stage and presenting a name as a big surprise (Capsaicin?), I'm not sure marketing is to blame in total. Raja runs the whole RTG doesn't he? I guess hes fine with the hype. Still, with Polaris AMD pulled out the "mid-range" targeted marketing, and it worked out ok, but clearly the market expected Vega to compete at the high-end, Im wondering at what point RTG realized that was not gonna happen. Honestly, they should have angled away from the high-end again, but I guess AMD is still gonna try to compete with 1080 instead of 1070.

 

[razor1]

Its pretty clear AMD regards any hype, any talk about them to be good talk. Or a disconnect from marketing and engineering. RTG was supposed to solve this, but if Raja Koduri is going on stage and presenting a name as a big surprise (Capsaicin?), I'm not sure marketing is to blame in total. Raja runs the whole RTG doesn't he? I guess hes fine with the hype. Still, with Polaris AMD pulled out the "mid-range" targeted marketing, and it worked out ok, but clearly the market expected Vega to compete at the high-end, Im wondering at what point RTG realized that was not gonna happen. Honestly, they should have angled away from the high-end again, but I guess AMD is still gonna try to compete with 1080 instead of 1070.

The last Raja presentation seemed to be on the money though, he didn't hype anything, I even stated this, think I was talking with Anarachist at the time too and pointed it out several times.

But yes I agree it has to start from the top of AMD. Doesn't matter if its the small guys talking or hyping shit up cause it reflects on the entire company.

 

[chenw]

Look AMD doesn't need to go for the top dog to make compelling products, we saw that with Polaris right? Its not anywhere as good as the gtx 1060 when it comes to power usage, but its performance is, and its selling as good as the entire line up of AMD's previous graphics cards top to bottom.

The problem with this sentiment is that AMD needs to stop just doing "as good" as previous generations, and they NEED to start doing BETTER than previous generations if they want to win back any sort of market share. Ryzen is a huge step in the right direction (at least it's not a die shrunk Bulldozer), but Vega falls far short, or perhaps a step backwards.

Sure, RX 480 and 1060 are similar in performance, but power consumption, and by extension, power requirement, is still different. We already saw what a lower consumption with similar performance to competition did for nV with Maxwell, so even if power usage is irrelevant in majority of cases (barring cases where it could mean requiring a much higher power supply), it is still a great marketting tool when you have very little else to compare with competition.

 

[SighTurtle]

The last Raja presentation seemed to be on the money though, he didn't hype anything, I even stated this, think I was talking with Anarachist at the time too and pointed it out several times.

But yes I agree it has to start from the top of AMD. Doesn't matter if its the small guys talking or hyping shit up cause it reflects on the entire company.

I don't remember the last Raja presentation aside from the Analyst Day cause I do remember the next event AMD had, he didn't show up and AMD presented something about Vega and competing with 1080 performance there. But my memory is faulty, I defer to you. When that Raja presentation came up with the no hype, when was that?

 

[razor1]

I don't remember the last Raja presentation aside from the Analyst Day cause I do remember the next event AMD had, he didn't show up and AMD presented something about Vega and competing with 1080 performance there. But my memory is faulty, I defer to you. When that Raja presentation came up with the no hype, when was that?

One before analyst day. He didn't hype anything about Vega, they showed things about it, like HBCC and primitive shaders and FP16 for better throughput. But when they showed those demos, its was obvious Vega's front end didn't change much from Polaris, but still people argued with me. All we have to do is look at what AMD has been showing, and its easy to see what the end results will be. Take everything as "best case", not overall.

Even on Analyst day he didn't hype anything either if I remember correctly they only talked about pro apps, and DL, and HBCC, even with DL benchs he stated its not about beating anyone, they just want to be in the talk about it. And pretty much he stated those benchmarks aren't realistic results by saying that.

 

[SighTurtle]

The problem with this sentiment is that AMD needs to stop just doing "as good" as previous generations, and they NEED to start doing BETTER than previous generations if they want to win back any sort of market share. Ryzen is a huge step in the right direction (at least it's not a die shrunk Bulldozer), but Vega falls far short, or perhaps a step backwards.

Sure, RX 480 and 1060 are similar in performance, but power consumption, and by extension, power requirement, is still different. We already saw what a lower consumption with similar performance to competition did for nV with Maxwell, so even if power usage is irrelevant in majority of cases (barring cases where it could mean requiring a much higher power supply), it is still a great marketting tool when you have very little else to compare with competition.

I think everyone knows AMD needs to do better to gain (including AMD), at this point, I think what the most realistic scenario is for AMD to hold onto the mid-range and try to retain marketshare it has until Navi or future comes about. Its pretty clear AMD decided to focus on Zen/holding off bankruptcy. Still, RTG represents a acknowledgement that AMD is losing the GPU war and it represents a regroup to try and bring it back. But that takes time. I'll buy a 1080 in the meanwhile.

One before analyst day. He didn't hype anything about Vega, they showed things about it, like HBCC and primitive shaders and FP16 for better throughput. But when they showed those demos, its was obvious Vega's front end didn't change much from Polaris, but still people argued with me. All we have to do is look at what AMD has been showing, and its easy to see what the end results will be. Take everything as "best case", not overall.

Even on Analyst day he didn't hype anything either if I remember correctly they only talked about pro apps, and DL, and HBCC, even with DL benchs he stated its not about beating anyone, they just want to be in the talk about it. And pretty much he stated those benchmarks are realistic results by saying that.

Yeah, he was very subdued in regards to Vega there. Hmm, seems like it applied across the board.

 

[razor1]

The problem with this sentiment is that AMD needs to stop just doing "as good" as previous generations, and they NEED to start doing BETTER than previous generations if they want to win back any sort of market share. Ryzen is a huge step in the right direction (at least it's not a die shrunk Bulldozer), but Vega falls far short, or perhaps a step backwards.

Sure, RX 480 and 1060 are similar in performance, but power consumption, and by extension, power requirement, is still different. We already saw what a lower consumption with similar performance to competition did for nV with Maxwell, so even if power usage is irrelevant in majority of cases (barring cases where it could mean requiring a much higher power supply), it is still a great marketting tool when you have very little else to compare with competition.

Its always good to be equal or better than the competition in every metric, but when a product doesn't make that mark in any metric, all the marketing in the world won't save it. The best thing to do, is eat it lol or lower the target levels. Shit Vega is pushing 300 watts for air cooled, that is going to be its nail in the coffin against a gtx 1080. Just ridiculous amounts of power, we aren't talking about 30 watts difference here from the 1060 to Polaris (rx 480, 580 is more like 80 watts).

Why start a marketing campaign for Vega a year before its launch when its going to fall really short?

Look at the FX launch for nV, they didn't talk about their product a year before launch. They actually tried to avoid the talk as much as possible once the 9700 was out, only as a last resort they talked about it. If I remember correctly the first real "show" of the fx was Kyle on a video, where Kyle stated this thing is really hot (temps wise lol). In presentations they showed things like UT2003 or 4, but never showed frame rates and this was weeks before launch.

As a Dev I didn't receive a FX at launch nV didn't send these cards to devs til about 6 months after launch. And we seriously needed those cards, cause nV had majority of the marketshare and we have to do special optimizations for them pretty much use the older DX8 path, so shaders had to be rewritten completely (no programability with shaders in DX8), pretty much we were already 1 year into dev of a project, can't do much when you don't have time. That's how much nV cared to push the FX series, they didn't care, cause they knew its not going to make a difference. FX was also the start of the TWIMTBP program by Roy Taylor, it fell flat on its face for the FX line, nV still lost marketshare by the droves.

So lessons learned from past mistakes from themselves or competitors? Why spend millions of tens or hundreds of millions of dollars marketing or hiring marketing folks without fixing the problem to begin with.

 

[c3k]

I buy my video cards based on whichever maker has the hottest cartoon chick avatar.

 

[Chimpee]

I buy my video cards based on whichever maker has the hottest cartoon chick avatar.

Lol free Ruby avatar with future dlc nude patch with purchase of RX Vega!

 

[Brackle]

Lol free Ruby avatar with future dlc nude patch with purchase of RX Vega!

lol RTG might need that to sell those Vega cards if the rumors are true.

 

[Chris_B]

If a game didn't support SLI and nVidia rolled out a driver that enabled SLI, then what's wrong with showing the gains?

AMD advertise the best case gains for their drivers all the time.

Because it was misleading, iirc the slide didn't even mention sli, just some random waffle about up to 70% gains. Was only later that people figured out where the magical performance gains were coming from.

And if you read my post properly you'd see I mentioned both companies are guilty of benching very specific scenarios in games and putting out numbers like the whole game plays that way.

 

[OutOfPhase]

Because it was misleading, iirc the slide didn't even mention sli, just some random waffle about up to 70% gains. Was only later that people figured out where the magical performance gains were coming from.

And if you read my post properly you'd see I mentioned both companies are guilty of benching very specific scenarios in games and putting out numbers like the whole game plays that way.

Following video card technology shows that if there's a gain beyond a few percent, it really means the driver (or sometimes game, and driver is working around it) was broken to begin with, and the "optimization" is just getting you to where you should have been all along. Seeing a claimed gain of 70% is a very bad thing - not something to boast about.

That aside, one could put forth an argument that overstating gains in a driver update is a different beast than overstating a card's basic performance level. The latter is used to mislead a buyer into a purchase, the former is not, at least not beyond general hype.

 

[reaper12]

More all AMD does is bad, all Nvidia does is good rubbish from you.

What is wrong with you? Did you really write all those long posts, detailing every time AMD were wrong? Are you like 12 or something? AMD lied 12 times, Nvidia only lied 5 times, na na na na.

And for all your BS, it still doesn't change anything, Nobody should buy a GPU based on Marketing, they should wait for reviews. Your very first line, proves the point that I was making, that only an idiot buys anything based on Marketing slides.

Every company out there stretches the truths to the limit when trying to Market products, look at cars for prime example, specifically their fuel consumption figures, figures that you won't get near in real life.

Wait, you are telling me that I need to supply facts in my statements? What exactly is there to prove with the statements I made? That people should wait for reviews before buying a card and not believe everything on market slides. I would call that good advice, no matter what you are buying.

Your complete Nvidia Bias is leading to you believe that my posts supports AMD. They don't, I apply the same rules to them, I don't believe AMD marketing, I don't believe Nvidia's marketing. And I give the same advice to anyone purchasing a card, wait for proper reviews.

But, you obviously think that's bad advice and is a pro-AMD, Anti Nvidia statement that I should back up with facts.

 

[Dan]

Hey guys?! Vega here yet? no? im leaving

 

[razor1]

What is wrong with you? Did you really write all those long posts, detailing every time AMD were wrong? Are you like 12 or something? AMD lied 12 times, Nvidia only lied 5 times, na na na na.

And for all your BS, it still doesn't change anything, Nobody should buy a GPU based on Marketing, they should wait for reviews. Your very first line, proves the point that I was making, that only an idiot buys anything based on Marketing slides.

Every company out there stretches the truths to the limit when trying to Market products, look at cars for prime example, specifically their fuel consumption figures, figures that you won't get near in real life.

Wait, you are telling me that I need to supply facts in my statements? What exactly is there to prove with the statements I made? That people should wait for reviews before buying a card and not believe everything on market slides. I would call that good advice, no matter what you are buying.

Your complete Nvidia Bias is leading to you believe that my posts supports AMD. They don't, I apply the same rules to them, I don't believe AMD marketing, I don't believe Nvidia's marketing. And I give the same advice to anyone purchasing a card, wait for proper reviews.

But, you obviously think that's bad advice and is a pro-AMD, Anti Nvidia statement that I should back up with facts.

Please put facts in your statements, and we can discuss, otherwise I will continue doing the same thing.

PROVE to me you know what you are posting about

PROVE to me you want to discuss about something, instead of trolling posts that have nothing factual!

Simple right

Its easy to say this

AMD has crappy marketing nV has great marketing

But why?

Its because of the reasons I have stated previous

Give examples and enumerate by explanation

Generalization doesn't do anyone good.

You generalized and now you are getting personal again.

You have no interest in upholding what you stated, you just like to flame and talk about other peoples preferences.

Yes how much did you spend on AMD hardware this year?

You think that I only buy nV hardware?

Do you want me to show you pics of my mining rigs?

I have over 20 now 23 to be exact, half of them are AMD the other half nV. I can prove it by pictures if you like or you can ask NKD, as we have been both talking about mining over PM about what is best for each type of mining rig.

Its what ever I can get my hands on. I probably spent more money in the past 3 months on AMD hardware then you did in a LIFE TIME.

So don't tell me what I am or what I am not.

Funny enough with the new claymore miners, I get over 35 mhs dual mining on my 1070's, and my 580's are only getting 29 mhs. If your assumptions about me were correct I would have just stuck with nV cards only and would be making 3% more then what I am now + the power draw diff, which is 1%, 11 systems at 4% is oh 500 bucks extra a month, a graphics card a month!. Yes that is how I would have looked at it back then, but that is ok, when I get my hands on more 1070's, I will be getting rid of the 580 cards at a premium. Satisfied? I look at things for what they are, not what I want them to be. If AMD can't come up with good products for gaming, Screw em. If they have to start marketing a year before release, and doing crap marketing to get sales, screw em. nV has not done anything of the sort in the past 7 generations or so, so how do you equalize them?

Tell me why you think AMD marketing and nV marketing is the same and give me examples of when and how they are the same.

Did nV ever have crappy marketing to the likes of Poor Volta? Have they fudged their TDP levels when they go out of pci-e specs? Have they had their director of server marketing ever go to a forum and lie about an up coming product (AMD with BD)? Do we have on stage presentation after presentation that just shows the same FUD over and over again for years to come? Have they locked frame rates and shown power usage to give us ideas of perf/watt, which never comes out true when the frame rate locks are not on? Do they mislead their fan base with performance figures without a base line? You tell me how in the world you equal them when nV doesn't do these types of things?

Shit if Vega was at 225 watts, I wouldn't even fricken hesitate to buy 20 of them when they are getting 55mhs. I wouldn't even need to think about it or wait for under-clocking results, its that simple. But that isn't the case.

 

[_mockingbird]

Vega is going to be an EPYC disaster.

In fact, this would be the biggest disaster in the history of NVIDIA and ATI.

 

[Briguyatx]

Please put facts in your statements, and we can discuss, otherwise I will continue doing the same thing.


Tell me why you think AMD marketing and nV marketing is the same and give me examples of when and how they are the same.

Did nV ever have crappy marketing to the likes of Poor Volta? Have they fudged their TDP levels when they go out of pci-e specs? Have they had their director of server marketing ever go to a forum and lie about an up coming product (AMD with BD)? Do we have on stage presentation after presentation that just shows the same FUD over and over again for years to come? Have they locked frame rates and shown power usage to give us ideas of perf/watt, which never comes out true when the frame rate locks are not on? Do they mislead their fan base with performance figures without a base line? You tell me how in the world you equal them when nV doesn't do these types of things?

How much ram does a 970 have again? I don't know about you, but I would also hesitate to buy an Nvidia card that utilized wood screws.

I'm not going to spend time right now finding more, but I would be willing to bet there are more out there. All companies lie - that's basically what advertising is, saying your product is the best at everything and unicorns and rainbows will result from a purchase.

With that said, I don't disagree with some of your comments.

 

[THUMPer]

How much ram does a 970 have again? I don't know about you, but I would also hesitate to buy an Nvidia card that utilized wood screws.

I'm not going to spend time right now finding more, but I would be willing to bet there are more out there. All companies lie - that's basically what advertising is, saying your product is the best at everything and unicorns and rainbows will result from a purchase.

With that said, I don't disagree with some of your comments.

I always keep this handy

 

[razor1]

How much ram does a 970 have again? I don't know about you, but I would also hesitate to buy an Nvidia card that utilized wood screws.

I'm not going to spend time right now finding more, but I would be willing to bet there are more out there. All companies lie - that's basically what advertising is, saying your product is the best at everything and unicorns and rainbows will result from a purchase.

With that said, I don't disagree with some of your comments.

The card still performed the way it did with Day 1 reviews, and it still had 4gb ram, the only thing that screwed nV was the ROP amounts. It became false adverting because the ROP amounts were right there on the box, which were not in the card.

Again nV never stated things they didn't deliver. The ROP amounts, yeah they got sued for it and they are paying for it. I signed the petition for it. They deserved to get sued.

Showing off a mock board when they were 6 months late, yeah when they fucked up, what did I say? They don't do things like that when they don't screw up. Even with that they still gave the market a card that held the performance crown albeit higher then desired power consumption. But there marketing was spot on, its the fastest damn card money could buy. They showed the wood screw Fermi, and then 6 months later they launched, they never had all these presentations what one presentation every 2 or 3 months, just to talk about different features and give performance projections based on those features? Did they do that? No.

Please find more, because when it comes to marketing, nV doesn't flat out do the same shit AMD does. They don't over promise. They don't pick obscure metrics without baselines. They just don't do those things because there is no need to, when the main metrics are just so much better.

Now for everyone saying it takes years to make drivers and drivers must be made after the silicon comes back.

Fermi wood screw edition lol, shown off, at that time they had no silicon back, but in 6 months they got the silicon back got the drivers up and going and lunched and took the performance crown. OK now where is VEGA now? What does that 1 month of driver revisions do?

Anyone that takes or talked about Fermi wood screw edition, should think twice before they bring up drivers for Vega. Apparently nV is so much better at making drivers they were able to get drivers done in 6 months with a COMPLETELY new architecture and AMD with 1 year can't do the same thing with a modified architecture! Yes there is heavy sarcasm in that last sentence.

 

[KazeoHin]

If you have a massive dick, you just whip it out and get laid. If you have a tiny dick, you talk big game and say how good you are and wait for the lights to go out before going full commando.

Same goes with advertising. If you have the better product, you just release it.

 

[Tup3x]

How much ram does a 970 have again?

4 GB. I've had GTX 970 and games happily used it all.

 

[Brackle]

4 GB. I've had GTX 970 and games happily used it all.

Actually it was 3.5gb usable. That is why Nvidia had to pay everyone $30 for false advertisement.

They ripped us off. (I got $60 back as proof)

P.S. that is why the 1080ti only has 11gb usable unlike the Titan Xp #2

 

[KazeoHin]

On the topic of the 970, I never understood the outrage. It performes exactly as advertised, and despite the Ram partitioning, it actually had 4gb of physical RAM on board.

 

[Brackle]

On the topic of the 970, I never understood the outrage. It performes exactly as advertised, and despite the Ram partitioning, it actually had 4gb of physical RAM on board.

Unless you tried to run any game that used above 3.5GB. At the time there were people buying 2 for SLI, and could not run SLI because of the 3.5GB issue. It was proven many times what the issue was.

Sure it was certain cases, but it was still an issue. Just look up the old thread. I posted in there and so did numerous people who bought them. I think the thread was started by Golden Tiger.

It was such an issue that Nvidia has to pay customers back.....They literally got caught, and had to pay people back. Not like this wasn't an issue. Nvidia has to pay customers back, thats a big fuck up IMO.

https://hardforum.com/threads/offic...-check-thread.1921066/page-10#post-1043121840

 

[THUMPer]

This thread is trash. I love it.

 

[bildad]

So looking at the last few pages it seems even the fanboys have conceded that Vega is a massive failure and now we're just arguing about the proper way to market failure?

 

[GoldenTiger]

So looking at the last few pages it seems even the fanboys have conceded that Vega is a massive failure and now we're just arguing about the proper way to market failure?

That's right.

 

[razor1]

So looking at the last few pages it seems even the fanboys have conceded that Vega is a massive failure and now we're just arguing about the proper way to market failure?

The thing is lets say AMD didn't push Vega to the max with clocks and just went after lets say 1070 performance levels, they might have gotten to that fairly easily with 200 watts. And I think it would do just fine for sales. 1070 is the big money maker for nV, and it has the bulk of all sales (volume) and margins..... That would be the card to go for, not the 1080 or 1080ti, specially when pushing voltage and power usage to the max, its not worth it.

 

[Dayaks]

The thing is lets say AMD didn't push Vega to the max with clocks and just went after lets say 1070 performance levels, they might have gotten to that fairly easily with 200 watts. And I think it would do just fine for sales. 1070 is the big money maker for nV, and it has the bulk of all sales (volume) and margins..... That would be the card to go for, not the 1080 or 1080ti, specially when pushing voltage and power usage to the max, its not worth it.

And people love saying things like, "Oh and you can overclock it to 1080 levels!" Also, when people OC they ignore power. It's still 200W in their minds.

 

[KazeoHin]

That's right.

Unfortunately, I've lost all hope in Vega.

So far Raja has the underwhelming Polaris launch and the absolute trash fire of Vega under his belt. In my mind he has ONE more chance to prove he can run RTG before I start thinking this guy has his head up his ass .

 

[Maddness]

The thing is lets say AMD didn't push Vega to the max with clocks and just went after lets say 1070 performance levels, they might have gotten to that fairly easily with 200 watts. And I think it would do just fine for sales. 1070 is the big money maker for nV, and it has the bulk of all sales (volume) and margins..... That would be the card to go for, not the 1080 or 1080ti, specially when pushing voltage and power usage to the max, its not worth it.

But, by the sounds of it, Vega is an expensive card to manufacture. So they would have competed on performance with the 1070, by more than likely cost the same or more than the 1080. I think they are in between a rock and a hard place when it comes to Vega.

 

[razor1]

And people love saying things like, "Oh and you can overclock it to 1080 levels!" Also, when people OC they ignore power. It's still 200W in their minds.

well 200 watts vs 300 watts, thats a huge difference lol, that is much higher than the 1070 to the 1080 150 to 180 watts, pretty much AMD would get a better benefit from not clocking these cards to the max. But of course price also plays a part into this they might not be able to get prices down to the 1070 range to keep it profitable enough..... cut down parts maybe they can do it.

 

[Dayaks]

well 200 watts vs 300 watts, thats a huge difference lol, that is much higher than the 1070 to the 1080 150 to 180 watts, pretty much AMD would get a better benefit from not clocking these cards to the max. But of course price also plays a part into this they might not be able to get prices down to the 1070 range to keep it profitable enough..... cut down parts maybe they can do it.

Well maybe not backing it off that far, but you pay dearly for that last 10%. IIRC my Titan X Maxwell used 250 watts default around 1400 Mhz, 1550 Mhz used 400 watts. Maybe they don't go to the edge... back it off 5% and save 20% on power or something.

 

[razor1]

Well maybe not backing it off that far, but you pay dearly for that last 10%. IIRC my Titan X Maxwell used 250 watts default around 1400 Mhz, 1550 Mhz used 400 watts. Maybe they don't go to the edge... back it off 5% and save 20% on power or something.

That's what I'm thinking, we have seen AMD not shoot for the top brass, and they certainly did better with marketshare. 38xx didn't shoot for the top and they still had 35% of the market, thats all they have to do to keep things afloat, till they start going to the black with the CPU division. More than likely that is the goal of Navi, I don't see AMD taking the top from nV in the near or mid future, to make these types of chips takes so much R&D its out of their reach.

 

[razor1]

Unfortunately, I've lost all hope in Vega.

So far Raja has the underwhelming Polaris launch and the absolute trash fire of Vega under his belt. In my mind he has ONE more chance to prove he can run RTG before I start thinking this guy has his head up his ass .

Its not fully Raja's fault at least not Polaris, cause he came in while its was already in development, Vega, too, I think his hands were tied to a lesser extent. I mean GF contract forced AMD to use them, then HBM. Those things look like they couldn't be side stepped. Now with Navi, lets see what happens, cause if that too fails, that's 100% Raja, no question about it, cause now they have the flexibility to side step GF and use any other foundry as long as they pay extra, and the HBM exclusivity contract should be over, if they are duel sourcing from Samsung with Vega, they have full filled their obligation from Hynix.

 

[SighTurtle]

Its not fully Raja's fault at least not Polaris, cause he came in while its was already in development, Vega, too, I think his hands were tied to a lesser extent. I mean GF contract forced AMD to use them, then HBM. Those things look like they couldn't be side stepped. Now with Navi, lets see what happens, cause if that too fails, that's 100% Raja, no question about it, cause now they have the flexibility to side step GF and use any other foundry as long as they pay extra, and the HBM exclusivity contract should be over, if they are duel sourcing from Samsung with Vega, they have full filled their obligation from Hynix.

CPU architectures take about 4 years to make right? How about the GPU side? If I'm not incorrect, you have been arguing GCN is basically at its limit. If Navi uses GCN I wouldn't be hopeful at this point, but keeping in mind the 2018-2019 predicted launch date of Navi, they would be nearing completion of any new architecture soon right? (Assuming a 2 year timeperiod for GPU)

I'm pretty ignorant in regards to the specifics of R&D funding so maybe I'm thinking about this wrong, you think AMD managed to get Navi a new arch?

 

[razor1]

3 to 5 years for GPU's too.

Navi should be a new architecture, but still, that entire time frame, Navi's R&D budget is still much smaller than Volta's. By a factor of 50% at least. Without the engineers at the design phase of a new GPU is going to hurt, and Raja made that pretty clear when he complained he didn't have the proper engineers in place just 7 months ago. That means the people he would have wanted were not there for Navi and won't be there cause their R&D expenditure as of now hasn't changed.

Also going scale-able might not be a fully new architecture either, so its up in the air what they are doing.

There are good parts to GCN, somethings need to be fixed. First off GCN's shader core is just fine, its potent, but to get the full potential out of it, the front end has to be completely redone, no modifications, completely redone, the entire cache architecture completely redone and so on. Transistor layouts, hand done, that is what nV has been doing for a good portion of their chips, AMD needs to do the same to stay with nV when it comes to power consumption, but this will cost A LOT of money, so don't know if they can do it.