New Upcoming ATI/AMD GPU's Thread: Leaks, Hopes & Aftermarket GPU's

The die consumes more power, the larger it is. Nvidia isn't really "trading", as they are not on the level of needing to push their hardware higher, unlike AMD who seems to need to run everything they make outside its optimal voltage curve. One is ahead, one is clearly behind.

Nvidia will go with UV, which means that their yields and costs will be completely fine. As that matures, they will be worth the trade off, like in every node transition.
I cannot see higher clocks out of them, but I could see a 6k+ core 3080Ti, on a smaller die size and lower power than the current 2080Ti.

There will be no HBM this time around, I think that AMD is traumatised by it.

 

I bet there will be Big Navi with HBM2 + HBCC* (HBM3 will be too expensive)
Save my post for the future

* high-bandwidth cache controller

 

SK Hynix reveals insanely fast HBM2E memory - Faster than an RTX 2080 on a single chip



SK Hynix has revealed the world's fastest HBM memory product, promising staggering speeds of up to 460GB/s per memory stack.
Each stack can also offer up to 16GB of capacity per stack, offering insane speeds and capacities given its form factor.

For context, Nvidia's RTX 2080 ships with 8GB of GDDR6 memory and delivers 448GB/s of memory bandwidth.
Yes, that means that SK Hynix can offer two times as much memory capacity and more memory performance on a single HBM2E memory chip.

Back in March, Samsung announced HBM2E memory stacks which offer 4.1GB/s of memory bandwidth, placing them behind SK Hynix's latest offerings.
This memory was already 2x as fast as the HBM2 memory used on AMD's RX Vega 56, a feat which makes SK Hynix's HBM2E offerings appear all the more impressive.

SK Hynix targets mass production of its HBM2E memory in 2020. This is when SK Hynix believe markets for HBM2E memory will to open up.
SK Hynix expects customers from the GPU market, creators of machine learning accelerators and other AI chip makers.

Right now, the most bandwidth rich graphics card to use HBM memory is AMD's Radeon VII, which offers 1024GB/s of total memory bandwidth.
Using SK Hynix' new HBM2E memory, a Radeon VII like graphics card would offer 1840GB/s of bandwidth, delivering a performance boost of almost 80%.

Even in the wake of GDDR6, SK Hynix has confirmed that HBM memory has a future.
The company's latest HBM2E memory modules pack more capacity and bandwidth per chip than all but the highest-end GDDR6 memory setups, which is no small feat for such compact, energy-efficient chips.

 

Couldn't hurt but I wonder if the clock speed isn't a primary issue over just bandwidth at least for gaming though it is known AMD's GPU's have been seeing higher than expected gains from memory improvements so there is some bottleneck and it could very well be both memory speed and bandwidth related, improved with VII and then again with Nano but still some sort of bottleneck though HBM is going to be costly and I don't expect E here to change that even if HBM is a nice addition to the GPU hardware but it's problematic from how much it currently costs.

Guess HBM3 is further out too then if they made additional adjustments to HBM2 (Samsung had some 2.99 thing earlier too I believe.) and I heard some low-cost version was also considered and then nothing more although it probably sacrifices cache or something to reduce the cost and complexity of the HBM and interposer circuitry.

Nice to see though whatever this ends up used in.

 

Except that you wrote that I am right. 1st, you even think that nVidia will not be able to increase clock (while I think around 12% is realistic). And that leaves you at gains gained only via transistor count.
And no, 7nm is not cheaper than 12/14/16nm per transistor. Neither is 7nm EUV. Therefore direct die shrink of 2080 Ti to 7nm EUV will cost more to make. And will deliver higher power efficiency. Therefore higher clock, lower power draw or some combination.

And power thing is one big misunderstanding for you. nVidia is ahead, therefore their clock gains from having better power efficiency at same transistor count will be lower than AMD's gains from better power efficiency. That's because nVidia is closer to clock wall of given manufacturing process at same transistor count and power draw.

As for HBM. That remains to be seen. But it is strong option. Interposer is already low complexity device. And while cost is at HBM I think that AMD more of needs that reduced power draw than bandwidth. Price difference between 12GB of GDDR6 and HBM is not going to be crazy. Especially if AMD does not need cutting edge chips.

 

SAPPHIRE PULSE RX 5700 XT Tested

-> https://wccftech.com/review/sapphire-pulse-rx-5700xt-navi-heartbeat-elevated/

and here:

 

Dell Alienware AW5520QF with 120Hz Freesync 4K UHD OLED screen !

4k USD



The monitor uses a 55” OLED (Organic Light Emitting Diode) panel with 3840 x 2160 (‘4K’ UHD) resolution, 120Hz refresh rate and support for Adaptive-Sync.
This includes AMD FreeSync support, although the variable refresh rate range is currently unknown. The screen surface is glossy, with an anti-reflective finish.
Such a large screen size is significantly larger than your typical monitor, but perhaps surprisingly some users do use such screens at a desk. Naturally, being able to sit a bit further back than with your typical monitor will help. The ‘4K’ resolution spread across a 55” screen, the pixel density is 80.11 PPI (Pixels Per Inch), so similar to a 27” 1920 x 1080 (Full HD) monitor.
A static contrast ratio of 130,000:1 is specified – vastly more impressive than the typical 1000:1 for an LCD monitor (or ~2500:1 – 3000:1 for a typical VA monitor).
A colour gamut of 98.5% DCI-P3 is specified, indicating significant extension beyond sRGB. The monitor does not offer any HDR support, however, and offers a typical maximum luminance of 400 cd/m².
A grey to grey response time of 0.5ms is specified, which unlike on an LCD is probably a fair reflection of what can be achieved without excessive grey to grey acceleration and the overshoot that comes with it.
This is evident from the pursuit photograph below, a still from this video taken by our friend Ziggy Orzeszek at Gamescom.
A very clean 120Hz performance indeed – for an explanation of what is being shown and some comparisons refer to this section of our article on responsiveness.

The rear of the screen uses matte black plastic towards the top and has a matte silver area lower down. This is removable, revealing the port area and some VESA holes for alternative mounting and the ports.
Separating the two areas is a large elongated oval and central Alienware logo RGB LED lighting feature. This can be customised via the OSD, but doesn’t offer any sort of substantial light output so is only designed to be admired from the rear.
The ports of the monitor, located in hexagonal recesses beneath the removable cover, include; DP 1.4, 3 HDMI 2.0 ports, 4 USB ports, S/PDIF audio line-out and a 3.5mm headphone jack. 2 x 14W speakers are also included – significantly more powerful than your typical monitor speakers, something the large screen size makes room for.
The full capabilities of the monitor (including 3840 x 2160 @120Hz and Adaptive-Sync) can therefore be harnessed using DisplayPort 1.4. HDMI 2.1 does not feature, although given the lack of support on current GPUs and the widespread use of DP on PCs that perhaps makes sense.

 

I can only imagine the price tag.

 

4k usd

 

Sony Toys with the Idea of Porting PlayStation-exclusives to PC

Sony Interactive Entertainment (SIE) is opening up to the idea of porting PlayStation-exclusive games to the PC, which is growing in popularity on the backs of titles that play best on this platform, such as Fortnite and Minecraft.
At the moment, Shawn Layden, who heads SIE, is interested in putting some its first-party titles (developed and published entirely by SIE) on the PC platform. Compared to arch-rival Microsoft, Sony's garden has had taller walls.
Microsoft has interests in both its Xbox console, and the Windows PC ecosystem, particularly the Microsoft Store, and has shepherded many of its ecosystem partners to co-develop for both platforms.
Its latest move is the Xbox for PC GamePass, that brings about some coherence between the PC and Xbox.

Sony is gearing up for the next round of the Console Wars against Microsoft, in which its PlayStation 5 console will compete with Microsoft's codenamed "Project Scarlett" console.
Both consoles will be designed for 4K Ultra HD televisions, with preparation for higher resolutions, such as 8K. Speaking on the development, Layden said "We must support the PlayStation platform — that is nonnegotiable.
That said, you will see in the future some titles coming out of my collection of studios which may need to lean into a wider installed base."

 

XFX RAW II

-> http://xfxforce.com/en-gb/Products/product-category/amd-radeon-rx-5700xt-8gb-raw-ii-rx-57xt8off6

 

AMD CEO Lisa Su: "CrossFire Isn't a Significant Focus"

AMD CEO Lisa Su at the Hot Chips conference answered some questions from the attending press.
One of these regarded AMD's stance on CrossFire and whether or not it remains a focus for the company. Once the poster child for a scalable consumer graphics future, with AMD even going as far as enabling mixed-GPU support (with debatable merits).
Lisa Su came out and said what we all have been seeing happening in the background: "To be honest, the software is going faster than the hardware, I would say that CrossFire isn't a significant focus".

There isn't anything really new here; we've all seen the consumer GPU trends as of late, with CrossFire barely being deserving of mention (and the NVIDIA camp does the same for their SLI technology, which has been cut from all but the higher-tier graphics cards).
Support seems to be enabled as more of an afterthought than a "focus", and that's just the way things are.
It seems that the old, old practice of buying a lower-tier GPU at launch and then buying an additional graphics processor further down the line to leapfrog performance of higher-performance,
single GPU solutions is going the way of the proverbial dodo - at least until an MCM (Multi-Chip-Module) approach sees the light of day, paired with a hardware syncing solution that does away with the software side of things.
A true, integrated, software-blind multi-GPU solution comprised of two or more smaller dies than a single monolithic solution seems to be the way to go. We'll see.

 

Looks slick! Wish it had a bit more white. Costs €437.

 

AMD Navi 14 GPU Shows Up In Compubench, Features 24 CUs – Entry-Tier Radeon RX To Be On Par With Radeon RX 570

From the looks of it, the Navi 14 GPU that has leaked out is aimed at entry-level Radeon RX cards with a focus on maximizing performance per dollar and efficiency.
The GPU in question seems to be configured with 24 CUs which rounds up to 1536 stream processors. It is codenamed GFX1012 which is the internal name for the Navi 14 GPU and has the “AMD 7340:CF” device ID.
The chip is clocked at 1.9 GHz and is coupled with 3 GB of VRAM. There were also the previous entries which reported 8 GB and 4 GB so it looks like the cards will come in different VRAM configurations.

 

24 CUs = 12 WGPs. Clock comparable to RX 5700 XT which has 20 WGPs. I say it will be good 15~20% faster than RX-580 in games. It just have limited compute capability.

 

1660ti and Vega56 performance for 200$?

 

I think it will land just under them. (Depending on clock it will boost to = Power limits.)

 

AMD Overtakes NVIDIA In GPU Market Share For The First Time Since Q3 2014 – Jon Peddie Research

The quarterly report from Jon Peddie Research (JPR) is out and turns out it has been a great quarter for AMD with global GPU shipments up by 9.8%.
NVIDIA’s shipments were flat for the quarter while Intel’s declined by 1.4%. Overall GPU shipments increased 0.6% on a sequential basis but were down 10.4% compared to the year-ago quarter.
Considering the overall PC market is up for the year, the reason why the dGPU market is down can only be attributed to the crypto-hangover and inventory-overstocking that was plaguing the markets. That said, it looks like things are finally starting to clear up.

PC market increases by 3.07% year over year as AMD overtakes NVIDIA in graphics market share for the first time since Q3 2014, according to Jon Peddie Research

Another interesting tidbit hidden in the charts is that AMD has overtaken NVIDIA in terms of market share.
This is absolutely great news for the company and something that hasn’t happened for well over 5 years. In fact, the last time AMD was on top was back in the third quarter of 2014.
While the complete details are hidden behind a paywall, we can see that AMD has overtaken NVIDIA in the total Graphics Market Share chart given in their press release below:



A total of 76.7 million units were shipped in Q2’19 which decreased by -8.93 million units from the same quarter a year ago indicating the market is negative. on a year-to-year basis.

Quick highlights

• AMD’s overall unit shipments increased 9.85% quarter-to-quarter, Intel’s total shipments decreased -1.44% from last quarter, and Nvidia’s decreased -0.04%.
• The attach rate of GPUs (includes integrated and discrete GPUs) to PCs for the quarter was 120% which was down -10.38% from last quarter.
• Discrete GPUs were in 26.95% of PCs,down -1.99% from last quarter.
• The overall PC market increased 9.25% quarter-to-quarter, and increased 3.07% year-to-year.
• Desktop graphics add-in boards (AIBs) that use discrete GPUs decreased -16.62% from last quarter.
• Q2’19 saw a decrease in tablet shipments from last quarter.

It is worth mentioning that the 0.6% increase in the GPU market quarter over quarter is a positive indicator because seasonally, shipments are usually down 2% this quarter.
The report further mentions that while there is uncertainty in the notebook supply chain there have been a ton of announcements in the same space with improvements in design philosophy and internal specifications. Considering the arrival of Zen and Intel’s brand new 10th generation processors things are expected to heat up again soon enough. As the dGPU becomes stronger and more power-efficient for the mobility platform, the numbers should go up again.

GPUs are usually a leading indicator for the PC market and considering they were down 10.4% might be taken by some as a forewarning.
Fortunately, however, the demise of the PC market has been greatly exaggerated and the naysayers have been talking about it since quite a few years. Considering the crypto hangover is just about over, I personally believe we are now entering natural organic growth once again. With Intel poised to join the GPU market in 2020 and AMD giving NVIDIA ample reason to compete once again, I have a feeling the next few quarters are going to see quite an influx of GPU shipments as the three major players kick into action.

Another interesting observation is that considering NVIDIA launched SUPER variants during this quarter and their shipments were mostly flat, it would seem the new lineup has not been well received even though it made the company more competitive from a price/perf perspective. Considering Jensen loves being the top dog around, I can only assume that this will give them ample impetus to stop lazing around and start a price/tech war.
Which is good news for the consumer because that’s when they win.

 

I would truly love it if that graph was discrete GPU sales.

 

On other hand, what new AMD's products were available in 2019Q1/Q2? I think that this is due to some deals made on corporate level.
And intel's much more rapid drop there means CPUs. Which in return means bigger CPU sales for AMD since computer market did not decline.
In general, it is good for AMD. Neutral for nVidia. And "nobody" cares about intel.