Tuesday, April 2nd 2019
AMD to Simultaneously Launch 3rd Gen Ryzen and Unveil Radeon "Navi" This June
TAITRA, the governing body behind the annual Computex trade-show held in Taipei each June, announced that AMD CEO Dr. Lisa Su will host a keynote address which promises to be as exciting as her CES keynote. It is revealed that Dr. Su will simultaneously launch or unveil at least four product lines. High up the agenda is AMD's highly anticipated 3rd generation Ryzen desktop processors in the socket AM4 package, based on "Zen 2" microarchitecture, and a multi-chip module (MCM) codenamed "Matisse." This launch could be followed up by a major announcement related to the company's 2nd generation EPYC enterprise processors based on the "Rome" MCM.
PC enthusiasts are in for a second major announcement, this time from RTG, with a technical reveal or unveiling of Radeon "Navi," the company's first GPU designed from the ground up for the 7 nm silicon fabrication process. It remains to be seen which market-segment AMD targets with the first "Navi" products, and the question on everyone's minds, whether AMD added DXR acceleration, could be answered. Lastly, the company could announce more variants of its Radeon Instinct DNN accelerators.
PC enthusiasts are in for a second major announcement, this time from RTG, with a technical reveal or unveiling of Radeon "Navi," the company's first GPU designed from the ground up for the 7 nm silicon fabrication process. It remains to be seen which market-segment AMD targets with the first "Navi" products, and the question on everyone's minds, whether AMD added DXR acceleration, could be answered. Lastly, the company could announce more variants of its Radeon Instinct DNN accelerators.
119 Comments on AMD to Simultaneously Launch 3rd Gen Ryzen and Unveil Radeon "Navi" This June
Radeon VII * 20/64 is still over 100W peak. How much more effective would Navi have to be?
Either a miracle or a serious underclock (and there goes your GTX1060 equivalent ;-))
What about die size?
Radeon VII * 20/64 is still 103mm2 - over 50% more than GPU die in 2400G.
Moreover, the chiplet in Zen2 is ~75mm2.
I'm not saying it's not possible, but very difficult for sure.
And what about the lost art of "making sense"? :-)
RX550-RX570 still sell pretty well. So even if all those customers got an APU, it would have to cost $200 to fill the revenue gap. I don't think that's the price people expect.
Clearly, I'm not a fan of how AMD does their business, but even I don't think they are that stupid.
Or are they?Honestly, why would anyone watercool a 1600X? :o
If you have to buy the cooler, it makes no economical sense at all.
Even if you already have a cooler that fits, you'll still likely be better off selling it and getting a faster CPU.
And it's not like there's some amazing OC potential in that CPU anyway.
Polaris is old, they are happy to get rid of it. Furthermore APU's are actually relatively cheap to make since they don't need their own PCB, GPU cooler, or usually built-on ram. It's just a cpu die and a gpu die.
On top of it being at best same clock +200Mhz.
On the other hand, #moarcoars.
Oh, and I'm pretty sure Intel will sell its chips in droves, even if the wildest dreams about Ryzen 3000 come true, no worries.
We know that Navi was never destined for any use as a Radeon Professional GPU. There's been different bits-n-pieces since 2016, that one of things RGT and Raja Koduri had been tasked with after Polaris, was to strip out much of the unused functions and features that basically only supported HPC and professional workloads. While this alone will not make GCN entirely efficient, that along with efficacies from Variable Rate Shading, GDDR6, while this striping "down" might actually "for once" might free refinement/fixes to the geometry engine and memory for gaming throughput. That along with 7nm might finally provide what is more the first "gaming centric" GCN architecture. If they can get between Vega 56 and 1070 Ti in 1440p, and do it say at 160W TDP they have a chance. It's not going to be all that earth shattering verse the completion, but if they can get it done for $280, it will be enough to hold out for Next-Gen "macro-architecture" sometime 2020-21.
On the consumer side they are supposedly willing to increase the TDP to 125-135w, thus opening up room for the greater clockspeeds. But of course they are only worrying about efficiency in servers.All the time if you are a content creator. All. The. Time.
You and others seem to be stuck in a logic loop about quad-cores in general. It's not that "Nothing needs more than a quad core," it's that "Nothing use to be programmed for more because there were only quad-cores." If 8 cores were the standard, games would make good use of them.
Rendering itself can't scale to arbitrary number of cores, even with the best effort splitting the rendering up into more than 2-3 queues is rarely beneficial, since rendering is a pipelined synchronous workflow. While it is technically possible to have multiple threads work on assembling a single queue, it is and will remain a bad idea, since the thread synchronization would kill performance outright. The future of rendering will continue to remove the CPU as a bottleneck for rendering, not by splitting the queue assembly over more cores, but by moving more of the heavy lifting to the GPU itself.
The only benefit games have from more than 4-5 cores is for non-rendering tasks, e.g. streaming. This usually don't impact gaming performance a lot, but sometimes there are some minor gains to frame time consistency.
The lack of creativity some people have.Then jog on, your input isn't needed here. Some people would like to not be held back by people who think PC's are only used for Fortnight and linear shooters with the same dumb AI that has been pigeon-walking around for 10 years.
It's like some people here don't want innovation, or maybe you have forgotten it can happen? Not sure. Or actually - I seriously think some people just want to be able to keep acting like their old Quad-Cores are high-end.
With no real improvements they start lacking more and more with any new generation.
Why do you think that is ?
A lot of companies make CPUs and GPUs, you know that I'm talking about PC parts.
Intel has the resources, but their GPU business will be anything but fire and forget.
When you have quad-core CPUs that are faster than the competition, you can sell them for years and people buy them.
Though I think there may be an issue with the ZEN2 Bios Updates for all Motherboard Manufacturers. There's talk about some sort of performance regression. A Regression is ~ 5% only.
Whatever the issue, AMD needs to fix this issue or current ZEN owners may suffer a little more latency issues.
Amd/comments/b8xc3cWhat Adored TV stated is 3000 is ready, but the x570 chipset may require a little more time for mobo manufacturers to implement. Or they are ironing out bios updates for stability and performance improvement.
This is the key to Navi if sources are correct, and Raja actually fixed Navi properly. Vega had nothing to do with Raja, and was pretty much
targeted to the compute crowd, though it does pretty well in gaming, it probably could have done much much better.
Also speculation based on a couple other sources I remember reading about. Not sure if this is true, but interesting nevertheless. Picture Below: