Vega II Logo Trademarked by AMD

[btarunr]

AMD late November filed a trademark application with the USPTO for a new logo, for its second generation "Vega" graphics architecture, built around the 7 nm silicon fabrication process. The logo looks similar to the original Vega "V," with two bands marking out the Roman numeral II (2). This logo could appear on the product and marketing on a series of new Radeon Pro and Radeon Instinct (and possibly even gaming-grade Radeon RX?) graphics cards based on AMD's new "Vega 20" multi-chip module. This chip features a doubling in memory bandwidth thanks to its 4096-bit wide HBM2 interface, and the optical shrink of the GPU die to the 7 nm node could enable AMD to dial up engine clocks significantly.



View at TechPowerUp Main Site

 

[Aldain]

Slav Vega

 

[qubit]

I didn't like the first version and this one is no better.

 

[ArbitraryAffection]

16GB HBM2 1TB/s Vega 20-based RX card with 1800 MHz boost and RTX 2080 level performance for < £450 pls. :3

 

[SL2]

Mm yeah I wonder why they went with VII and not V2...

 

[ZoneDymo]

I didn't like the first version and this one is no better.

why not? (on both statements)

 

[qubit]

why not? (on both statements)

Just don't like it, looks ugly. It's a matter of preference. The new one isn't sufficiently different for to change my opinion.

If you happen to like it, then enjoy, it doesn't bother me.

 

[craigo]

 

[dinmaster]

Would be awesome to see the gpu go the way cpu' s from amd are going with its infinity fabric I think it's called. Put 2 or 4 gpus in one die. From what I hear memory bandwidth is the problem but I think with hbm2 and it's big bandwidth interface it could be a reality..

 

[craigo]

Would be awesome to see the gpu go the way cpu' s from amd are going with its infinity fabric I think it's called. Put 2 or 4 gpus in one die. From what I hear memory bandwidth is the problem but I think with hbm2 and it's big bandwidth interface it could be a reality..

Be gone with your optimistic outlooks on actual technology, This is a patented marketing thread..

sorry, i just miss Ruby.

 

[Ruru]

I didn't like the first version and this one is no better.

Didn't like the first, but this looks better.

 

[the54thvoid]

View attachment 112268

At first, I saw what I thought I saw, then realised it was eyes and a nose.

However, I revert back to my original visual stimulus and say that you have added a 'cock and balls'.

 

[Hardware Geek]

I sincerely hope to they introduce a graphics card with similar performance to the 2080 at around half the price and do to the GPU market what they're trying to do in the CPU market.

 

[moproblems99]

I sincerely hope to they introduce a graphics card with similar performance to the 2080 at around half the price and do to the GPU market what they're trying to do in the CPU market.

Yeah because hey, why should AMD get any of those phat margins. Then I remembered, even at half of the 2080, it is still $400. They might be able to get a margin out of that - though not with HBM. If they can manage 2080 levels, it will not be less than $500 until NV drops prices.

At first, I saw what I thought I saw, then realised it was eyes and a nose.

However, I revert back to my original visual stimulus and say that you have added a 'cock and balls'.

It is very clearly what you thought you saw, I mean really, what else would it be?

 

[Hardware Geek]

Yeah because hey, why should AMD get any of those phat margins. Then I remembered, even at half of the 2080, it is still $400. They might be able to get a margin out of that - though not with HBM. If they can manage 2080 levels, it will not be less than $500 until NV drops prices.



It is very clearly what you thought you saw, I mean really, what else would it be?

Oh, don't get me wrong, I want them to make money. Even though their CPUs are significantly less expensive than their intel counterparts, they still are making a profit. I'm hoping they can pull off a similar move with their GPUs and still make a profit. Half was just a quick comment off the top of my head, if they can get similar performance levels and even undercut Nvidia by 10-20% they will have a winning product and gain much needed market share as well as provide competition to drive down prices on Nvidia cards as well.

 

[AsRock]

WTF, i do wounder how much time time it took for them to come up with pretty much, errm well nothing.

EDIT Still cracking me up of all the wasted time.

 

[RealNeil]

I sincerely hope to they introduce a graphics card with similar performance to the 2080 at around half the price and do to the GPU market what they're trying to do in the CPU market.

Agreed

NVIDIA needs the same slap and tickle that Intel got.
(anything to wake them up from the profit coma that they're in)

 

[efikkan]

16GB HBM2 1TB/s Vega 20-based RX card with 1800 MHz boost and RTX 2080 level performance for < £450 pls. :3

I believe 7 nm is about twice as expensive to make, HBM is still expensive, and a chip full of fp64 units is a waste for gaming.

A consumer Vega 20 is not planned. Plans can change, but availability will be low, the price will be high, and it's still going to be Vega, which even at 7 nm will not be doing well against Turing.

 

[HTC]

Agreed

NVIDIA needs the same slap and tickle that Intel got.
(anything to wake them up from the profit coma that they're in)

Agreed.

I'm not expecting performance of upcoming AMD flagship to be higher performer than around RTX 2070 (without the RT part): what i am expecting is much more power efficiency and hoping for much better pricing VS similar performing nVidia cards.

Will this happen? Dunno, really.

 

[RealNeil]

At the low end, there will supposedly be a Radeon RX 3060 graphics card, based on the small Navi 12 design. It will be paired with 4 GB of GDDR6, cost $129.99, have a 75 W TDP, which means it will not need any additional power than the PCIe slot is specified to provide. Its performance is supposed to be approximately that of the RX 580. Next up is the RX 3070, still using Navi 12 but now with 8 GB of GDDR6, a 120 W TDP, $199.99 price, and matching the performance of the Vega 56. Lastly we have the RX 3080 using the larger Navi 10 chip, with 8 GB of GDDR6 and a 150 W TDP, priced at $249.99 with performance about 15% above the RX Vega 64.

Source: https://www.overclockersclub.com/news/42145/

 

[FordGT90Concept]

Vega II
VII
7
7nm

 

[dan80331]

Well,that's it.

 

[NC37]

At the low end, there will supposedly be a Radeon RX 3060 graphics card, based on the small Navi 12 design. It will be paired with 4 GB of GDDR6, cost $129.99, have a 75 W TDP, which means it will not need any additional power than the PCIe slot is specified to provide. Its performance is supposed to be approximately that of the RX 580. Next up is the RX 3070, still using Navi 12 but now with 8 GB of GDDR6, a 120 W TDP, $199.99 price, and matching the performance of the Vega 56. Lastly we have the RX 3080 using the larger Navi 10 chip, with 8 GB of GDDR6 and a 150 W TDP, priced at $249.99 with performance about 15% above the RX Vega 64.

Source: https://www.overclockersclub.com/news/42145/

Aggressive pricing would be like a return of the 3870 days when ATI couldn't make a competitive GPU and all they could do was sell for dirt cheap. nVidia rode the 8800 series and later the G92 forever. Pricing that cheap would mean Navi will flop. Which is kinda expected. At this point AMD practically needs to hire the GPU version of Jim Keller. They can probably ride node shrinks for a bit more before it'll be clear their designs need to change radically.

 

[efikkan]

At the low end, there will supposedly be a Radeon RX 3060 graphics card, based on the small Navi 12 design. It will be paired with 4 GB of GDDR6, cost $129.99, have a 75 W TDP, which means it will not need any additional power than the PCIe slot is specified to provide.

I really hope that any card with a 75W TDP will rely on additional power connectors. Cards tend to have peak consumption above their TDP rating, and releasing a card relying on the motherboard to supply out-of-spec power is not a good idea.