Vega/Navi Rumors (Updated)

[IEC]

From my source. No. Still Q2 expected release. That I was told.

A certain poster who posted a thread earlier today claiming that Vega was "indefinitely delayed" has egg on their face now. Especially because Ryan Smith here at AT has a "not Vega" in his hands which would be very odd if launch was not imminent.
https://twitter.com/RyanSmithAT/status/854584751948038144

 

[JDG1980]

A certain poster who posted a thread earlier today claiming that Vega was "indefinitely delayed" has egg on their face now. Especially because Ryan Smith here at AT has a "not Vega" in his hands which would be very odd if launch was not imminent.
https://twitter.com/RyanSmithAT/status/854584751948038144

If these cards are already in the hands of reviewers, it's surprising we haven't seen leaks yet - unless AMD is being extremely selective about who gets them. Could Vega be coming sooner than most people expect?

The R9 300 series rebrands were released on June 18, 2015. Fury X was released on June 24, 2015.
The RX 500 series rebrands were released on April 18, 2017. Vega will be released... ???

 

[Mockingbird]

A certain poster who posted a thread earlier today claiming that Vega was "indefinitely delayed" has egg on their face now. Especially because Ryan Smith here at AT has a "not Vega" in his hands which would be very odd if launch was not imminent.
https://twitter.com/RyanSmithAT/status/854584751948038144

Okay, IEC.

When will Vega launch? (ie. last week of April, first week of May, etc.)

How will it perform? (i.e. big Vega as fast as Geforce GTX 1080 Ti on air, etc.)

How much will it costs? (i.e. $499 for big Vega, etc.)

 

[sandorski]

Okay, IEC.

When will Vega launch? (ie. last week of April, first week of May, etc.)

How will it perform? (i.e. big Vega as fast as Geforce GTX 1080 Ti on air, etc.)

How much will it costs? (i.e. $499 for big Vega, etc.)

You're the one with all the inside info, you tell us.

 

[Mockingbird]

You're the one with all the inside info, you tell us.

Most likely, AMD is going to drop HBM2 like a sack of hot potatoes and go with GDDR5X for the high end to keep cost under control.

 

[sandorski]

Most likely, AMD is going to drop HBM2 like a sack of hot potatoes and go with GDDR5X for the high end to keep cost under control.

No.

 

[unseenmorbidity]

Most likely, AMD is going to drop HBM2 like a sack of hot potatoes and go with GDDR5X for the high end to keep cost under control.

The gpu is already done. They're almost certainly manufacturing now, as the release is a month out.

 

[Mopetar]

Most likely, AMD is going to drop HBM2 like a sack of hot potatoes and go with GDDR5X for the high end to keep cost under control.

This is so far off the mark that not even wccftech would run it as a story.

 

[raghu78]

Most likely, AMD is going to drop HBM2 like a sack of hot potatoes and go with GDDR5X for the high end to keep cost under control.

AMD has demoed Vega. Raja Koduri has held out the Vega chip with HBM2 for everyone to see.

http://www.anandtech.com/show/11002/the-amd-vega-gpu-architecture-teaser/4

Something tells me you are not going to last long here with these kind of posts.

 

[Mockingbird]

AMD has demoed Vega. Raja Koduri has held out the Vega chip with HBM2 for everyone to see.

http://www.anandtech.com/show/11002/the-amd-vega-gpu-architecture-teaser/4

Something tells me you are not going to last long here with these kind of posts.

Plan A is HBM2.

Plan B is GDDR5X.

It's the backup option.

 

[unseenmorbidity]

Plan A is HBM2.

Plan B is GDDR5X.

It's the backup option.

What is your source?

 

[Mockingbird]

What is your source?

Financial analyst expert.

He is not a fanboy at all.

 

[Mopetar]

What is your source?

Financial analyst expert.

He is not a fanboy at all and is analysing AMD strictly from financial views

You sure he's not a proctologist, because his analysis is coming from his posterior. This person is clearly clueless and I wouldn't trust other predictions with this being so far off the mark.

This is a product where AMD has talked about HBM2 for over a year, has shown a chip with HBM2 visible, and is expected to ship within the next 60 days or so.

Vega not having HBM2 is about as likely as getting hit by lightning. If you're going to insist on believing that you at least need to make a cat food eating video when this turns out to be wrong.

 

[Mockingbird]

You sure he's not a proctologist, because his analysis is coming from his posterior. This person is clearly clueless and I wouldn't trust other predictions with this being so far off the mark.

This is a product where AMD has talked about HBM2 for over a year, has shown a chip with HBM2 visible, and is expected to ship within the next 60 days or so.

Vega not having HBM2 is about as likely as getting hit by lightning. If you're going to insist on believing that you at least need to make a cat food eating video when this turns out to be wrong.

He didn't said that (big) Vega won't ship with HBM2, but that AMD has to either make (big) Vega faster or make it cheaper.

If AMD chooses to make (big) Vega cheaper, the first thing that would be on the chopping block would be HBM2.

 

[itsmydamnation]

He didn't said that (big) Vega won't ship with HBM2, but that AMD has to either make (big) Vega faster or make it cheaper.

So what exactly does that have to do with anything?

Your friend doesn't have much of an idea what he is talking about, does he? So whats the cost of HBM vs HBM2? whats the cost of the passive interposer? But more importantly whats the cost of 12 GB of GDDR5X vs 8 GB of HBM2!

I dont think people like you realize how cheap it is to manufacture CPU's/GPU's ( you can bin/salvage them so well) and how expensive a massive amount of a extremely boutique memory type is, GDDRX5 is used in far fewer products then HBM2.

But consider this Vega is smaller then Fiji, with a smaller interposer and fewer memory stacks using a technology that has a vastly larger amount being manufactured. So even if Vega performed as poorly per as fiji it is still not going to be a worse product financially.

This is because Vega already supports GDDR5X and switching from HBM2 to GDDR5X would cause the least delay.

You really have no idea, so what your saying is that AMD has wasted something like 15-20% of its die space to package a 384/512bit GDDR memory interface because you certianly cant use the HBM interface! Or your other option is you go all the way back to RTL and change the memory subsystem. So whats the cost of that, consider opportunity, development and the impact that has on future development?

Then we move onto they delay, what delay exactly? You realize your stupid HBM to GDDRX swap is not a 3 month delay, more like 9 to 12.

But what your financial expert friend has really missed is the part where AMD has said they have used there priority access to prepurchased millions of dollars of HBM2 up front.

But what i find even funnier is, that GCN is still more advanced then pascal in terms of compute, Vega is bringing big updates in the area's that NV has clear advantage ( ROP's, ROP cache, frontend, increased clock). But then on top of that they are bring the follow on ALU architecture to GCN.

Go and read any of the GPU ALU paitents from the last 2-3 years from AMD, if any of it appears is NCU then GCN's advantage extends.

You friends logic doesn't stand up to even the simplest of scrutiny.

 

[beginner99]

He didn't said that (big) Vega won't ship with HBM2, but that AMD has to either make (big) Vega faster or make it cheaper.

If AMD chooses to make (big) Vega cheaper, the first thing that would be on the chopping block would be HBM2.

This is because Vega already supports GDDR5X and switching from HBM2 to GDDR5X would cause the least delay.

Well I can agree that maybe small Vega has GDDR5x but I still doubt it. Big Vega will be HBM2 as we have seen the chip already.

 

[HumblePie]

I can't believe such a big company allows its stuff to give production samples to someone who doesn't work, not just in AMD, but even on that specific project. Especially if it is such a big and important project. Totally unprofessional.

I can't wait to find out more about Vega, but I think you shouldn't reveal any details here.

I'm not just anyone off the street. Nor am I revealing anything either that I am not given permission to do so for, or aren't generalities. After I test it a bit and if asked if it looks to be a good card, I could probably answer that with a highly subjective opinion with only slightly more grounding than the average person. That would be about it without permission. Which is basically nothing more than what their current marketing has put out I guess.

 

[jpiniero]

Well I can agree that maybe small Vega has GDDR5x but I still doubt it. Big Vega will be HBM2 as we have seen the chip already.

True, but they don't have to release any Big Vega consumer models if the 1080 Ti killed their pricing.

 

[krumme]

True, but they don't have to release any Big Vega consumer models if the 1080 Ti killed their pricing.

Vega only runs using hbm2. Its in the dna as i understands it.
The problem for amd is the amount of ram possible to offer. It will probably be 4 or 8 gb. Its a huge marketing disadvantage. 4gb for 350 usd anyone?

I dont care. Its all about perf, but amd surely have a major challenge as they use to sell by more ram in a market where nv brand is extreme strong.

 

[SpaceBeer]

It seems 4GB will be more than enough, especially for models used in 1080p and probably 1440p gaming
https://www.youtube.com/watch?v=85ProuqAof0
However, it will be hard to convince people to buy $500 GPU with only 4GB of memory. Since it can efficiently use system RAM dut to HBCC, my advice to them would be to advertise Vega saying it has "4GB HBM2 and up to* 20 GB of memory"

 

[Sven_eng]

Vega can use any type of memory.

 

[mohit9206]

It seems 4GB will be more than enough, especially for models used in 1080p and probably 1440p gaming
https://www.youtube.com/watch?v=85ProuqAof0
However, it will be hard to convince people to buy $500 GPU with only 4GB of memory. Since it can efficiently use system RAM dut to HBCC, my advice to them would be to advertise Vega saying it has "4GB HBM2 and up to* 20 GB of memory"

It won't have 4gb. Releasing a 4gb card would be worse than 780Ti 3GB.

Vega can use any type of memory.

Even GDDR3?

 

[SpaceBeer]

AMD confirmed there will be 4 GB configurations
https://videocardz.com/67642/amd-radeon-rx-vega-is-just-around-the-corner
And according to demos we've seen, it should be enough due to HBCC. The only issue is marketing

 

[bononos]

Maybe releasing a 4Gb version shows that AMD has more than a little confidence in the performance of the design.

 

[SpaceBeer]

In their Deus Ex demo with 2 GB of memory, it was mentioned average FPS can be increased up to 50% with HBCC on (min FPS up to 100%). The difference in avg FPS between RX 460 2 GB and 4 GB is ~20%.
Having in mind that Fury (X) works quite well even without HBCC, I have no doubts this will be more than enough.

I suppose they will use 2x2GB stacks, and overall production cost should be similar to Fury, if not lower.

What bothers me is, Raja confirmed it will be called RX Vega. It means no numbers, and therefore not so many different configurations. We know there is 64 CU chip, but are there any 44 CU or 56 CU parts? Does that mean they'll only release one or two SKUs?