SoylentNews is people
SoylentNews
AMD, Nvidia Have Launched the Least-Appealing GPU Upgrades in History
Yesterday, AMD launched the Radeon VII, the first 7nm GPU. The card is intended to compete with Nvidia's RTX family of Turing-class GPUs, and it does, broadly matching the RTX 2080. It also matches the RTX 2080 on price, at $700. Because this card began life as a professional GPU intended for scientific computing and AI/ML workloads, it's unlikely that we'll see lower-end variants. That section of AMD's product stack will be filled by 7nm Navi, which arrives later this year.
Navi will be AMD's first new 7nm GPU architecture and will offer a chance to hit 'reset' on what has been, to date, the least compelling suite of GPU launches AMD and Nvidia have ever collectively kicked out the door. Nvidia has relentlessly moved its stack pricing higher while holding performance per dollar mostly constant. With the RTX 2060 and GTX 1070 Ti fairly evenly matched across a wide suite of games, the question of whether the RTX 2060 is better priced largely hinges on whether you stick to formal launch pricing for both cards or check historical data for actual price shifts.
Such comparisons are increasingly incidental, given that Pascal GPU prices are rising and cards are getting harder to find, but they aren't meaningless for people who either bought a Pascal GPU already or are willing to consider a used card. If you're an Nvidia fan already sitting on top of a high-end Pascal card, Turing doesn't offer you a great deal of performance improvement.
AMD has not covered itself in glory, either. The Radeon VII is, at least, unreservedly faster than the Vega 64. There's no equivalent last-generation GPU in AMD's stack to match it. But it also duplicates the Vega 64's overall power and noise profile, limiting the overall appeal, and it matches the RTX 2080's bad price. A 1.75x increase in price for a 1.32x increase in 4K performance isn't a great ratio even by the standards of ultra-high-end GPUs, where performance typically comes with a price penalty.
Rumors and leaks have suggested that Nvidia will release a Turing-based GPU called the GTX 1660 Ti (which has also been referred to as "1160"), with a lower price but missing the dedicated ray-tracing cores of the RTX 2000-series. AMD is expected to release "7nm" Navi GPUs sometime during 2019.
Radeon VII launch coverage also at AnandTech, Tom's Hardware.
Related: AMD Returns to the Datacenter, Set to Launch "7nm" Radeon Instinct GPUs for Machine Learning in 2018
Nvidia Announces RTX 2080 Ti, 2080, and 2070 GPUs, Claims 25x Increase in Ray-Tracing Performance
AMD Announces "7nm" Vega GPUs for the Enterprise Market
Nvidia Announces RTX 2060 GPU
AMD Announces Radeon VII GPU, Teases Third-Generation Ryzen CPU
AMD Responds to Radeon VII Short Supply Rumors
(Score: 1, Interesting) by Anonymous Coward on Saturday February 09 2019, @05:07AM (3 children)
Let's all love Lain!
(Score: 0) by Anonymous Coward on Saturday February 09 2019, @05:27AM (2 children)
Hey! Listen!
(Score: 1, Funny) by Anonymous Coward on Saturday February 09 2019, @06:32AM (1 child)
Sorry, can't. Running Systemd and Pulseaudio. What?
(Score: 0) by Anonymous Coward on Saturday February 09 2019, @03:07PM
You'll never gonna lennart.
(Score: 1, Insightful) by Anonymous Coward on Saturday February 09 2019, @05:22AM (4 children)
We should not be surprised as the manufacturing at 7nm must be far more complex and costly than 10nm or 14nm. The price reflects that. In time, as usual, it will come down. But for now the early adopters pay for the bulk of the R&D. Same pattern with any technology.
(Score: 0) by Anonymous Coward on Saturday February 09 2019, @08:00AM (3 children)
Complete and utter bs.
They just want more money for a similar product
(Score: 3, Touché) by NateMich on Saturday February 09 2019, @08:30AM (2 children)
Actually, that's complete and utter truth.
But I'm sure you're right about them wanting more money.
(Score: 2, Insightful) by Anonymous Coward on Saturday February 09 2019, @03:04PM (1 child)
There is no truth in capitalism, only lies and high prices.
(Score: 3, Insightful) by acid andy on Sunday February 10 2019, @12:40AM
FTFY.
If a cat has kittens, does a rat have rittens, a bat bittens and a mat mittens?
(Score: 5, Interesting) by black6host on Saturday February 09 2019, @05:34AM (2 children)
I've been chasing the video card manufacturers since the days of Hercules cards. Nowadays, I'm way behind the curve. A lowly GTX 960. And, it plays everything I want so fuck all that spending! I'll let the rich folks pay the crazy prices and wait for, as always happens, the tides to turn.
(Score: 2) by richtopia on Saturday February 09 2019, @06:54PM (1 child)
I am surprised that the GPU manufacturers aren't pursuing the budget market more. Both companies have new GPU architectures and access to new manufacturing lines, so why are there no sub $200 cards younger than two years old? AMD has demonstrated Vega can in APU configurations with less than 56 compute units, but there are no dedicated GPU solutions fitting that market segment.
(Score: 2) by takyon on Saturday February 09 2019, @07:17PM
GTX 1050 Ti, 1050, and GT 1030 are a bit older than 2 years, but they definitely address the segment.
RX 560 and RX 550 [anandtech.com] actually meet your requirements (less than 2 years old). They launched mid-2017 at $79-$99.
As for why there is no budget Vega, it probably comes down to factors related to the architecture and economics. AMD could address the budget segment with Navi instead within the next few months.
https://wccftech.com/four-amd-navi-gpu-variants-leaked-rumored-for-july-2019-launch/ [wccftech.com]
RX 3070 at $199, RX 3060 at $129 (according to leak/rumor).
https://wccftech.com/amd-navi-7nm-radeon-rx-gpus-delayed-q4-2019/ [wccftech.com]
Rumored delay to October 2019.
https://wccftech.com/exclusive-first-amd-navi-gpu-will-have-40-cus-and-is-codenamed-navi-12/ [wccftech.com]
That November article says that "Vega 7nm will not be coming to gamers". While you could say that's not true, you could also say that Radeon VII is a very niche part apparently being sold in limited quantities, which may have been released to fill in a gap between the earlier Vega release and possibly-delayed Navi release.
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 3, Interesting) by bzipitidoo on Saturday February 09 2019, @05:47AM (3 children)
If die sizes still mean anything, haven't been totally distorted by marketing, how is it that CPUs have only just moved from 14 nm to 10 nm or maybe 12 nm, while GPUs are already at 7 nm? What did I miss?
(Score: 3, Informative) by takyon on Saturday February 09 2019, @06:40AM
The only "7nm" process in use right now is TSMC's. And it's been used to make Apple's SoCs and some other products:
https://en.wikichip.org/wiki/apple/ax/a12 [wikichip.org]
AMD decided to release a GPU on it, seemingly available only in very limited quantities [tomshardware.com], before launching "7nm" CPUs later this year (probably in May, no earlier than March, or no later than July).
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 0) by Anonymous Coward on Saturday February 09 2019, @08:33AM (1 child)
The last couple of years of semiconductor news, apparently.
(Score: 1, Interesting) by Anonymous Coward on Saturday February 09 2019, @03:13PM
Also memory usually moves first to most new techs as it is 'easy and regular'. Most of a GPU is memory.
(Score: 2, Interesting) by Anonymous Coward on Saturday February 09 2019, @12:42PM (4 children)
As long as I can play Senran Kagura: Peach Beach Splash, plus all those lewd games by Illusion, I'm probably good with my graphics card (Radeon RX 470). These fancy lithographies do nothing to make the tits look any better, so why would I buy them? Where's the part on the spec sheet where they brag about how cleanly they can render a variety of nipple puffinesses?
(Score: 0) by Anonymous Coward on Saturday February 09 2019, @03:06PM
VR pom?
(Score: 2) by takyon on Saturday February 09 2019, @05:42PM
Given the amount of Japanese media (novels, manga, anime) about post-Singularity VR titties and VRMMOs, you will probably be pushed into getting a better GPU at some point.
But if you are planning on skipping several generations, the absolute low-end of GPUs might become more interesting to you than the high end. Imagine something in the price lane of AMD RX 460, Nvidia GTX 1050 or even the GT 1030, except a few lithography nodes down the line. You could imagine cards like that outperforming RX 470 at a lower cost, lower power consumption, with more than 4 GB VRAM, on something like the TSMC "5nm" node.
If we end up giving Moore the finger and getting orders of magnitude more GPU performance, than we can talk about absurd scene complexity, 8-16K resolution, 240 Hz, and real-time raytracing... all in a standalone VR headset that doesn't warm your face.
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 0) by Anonymous Coward on Sunday February 10 2019, @01:27PM
Just think about this for a moment. One day, hentai titles like that may be indistinguishable from real life porn. Just imagine. It's still going to be generated actors, just deepfaked well enough to look real.
Can't wait.
(Score: 0) by Anonymous Coward on Sunday February 10 2019, @01:44PM
I think I am in love [peachbeachsplash.com].
(Score: 0) by Anonymous Coward on Saturday February 09 2019, @05:07PM (1 child)
VR + PhysX, droll!
there once existed a dedicated pci(e?) slot card just to do the physX stuff ...
ofc all the nintendo console are so appealing because everything runs smooth at framerates ...
the PC is a frankenstein and has gazillion component-combos possibilities ...
sheesh .. i don't know what to say. so the next generation of GPU isn't as ...uhm...errr... game changing as anticipated? oh well.
anyways, personally i would prefer more physics then more graphic-realism in the next game iteration.
(Score: 5, Interesting) by takyon on Saturday February 09 2019, @06:05PM
I think it's more about the game engines and game developers than GPU hardware, although the move to 8-16 core CPUs might help.
There's research being done that could improve graphic-realism and physics at the same time. A lot of it involves machine learning:
https://www.youtube.com/watch?v=R5t74AC6I0A [youtube.com]
https://www.youtube.com/watch?v=aAsejHZC5EE [youtube.com]
https://www.youtube.com/watch?v=ZEjUqZU1hNQ [youtube.com]
https://www.youtube.com/watch?v=MCHw6fUyLMY [youtube.com]
https://www.youtube.com/watch?v=uGhyOBSzdTs [youtube.com]
https://www.youtube.com/watch?v=Mu0ew2F-SSA [youtube.com]
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 1, Insightful) by Anonymous Coward on Sunday February 10 2019, @04:32AM (1 child)
"A 1.75x increase in price for a 1.32x increase in 4K performance"
You assume that the relationship between price and performance is linear. That's almost never true. The price difference between a car with a max speed of 200 MPH vs one with a max speed of 100 MPH is proportionately larger than the price difference between a car with a max speed of 100 MPH vs one with a max speed of 50 MPH. What about a car with a max speed of 400 MPH vs one with a max speed of 200 MPH. Eventually you reach a limit where the price difference between a car with a max speed of X vs one with a max speed of X+1 approaches infinity.
(Score: 0) by Anonymous Coward on Sunday February 10 2019, @02:38PM
They didn't claim that there is a perfectly linear relationship between price and performance at the high end.
"A 1.75x increase in price for a 1.32x increase in 4K performance isn't a great ratio even by the standards of ultra-high-end GPUs, where performance typically comes with a price penalty."