These were all some of the reasons I was 100% sold on 4nm before the motherboard leak happened.
i don't forgot. 20 GB/s is more than enough. Then 100 Gb/s for GPU + 20 GB/s for CPU and the LPDDR5X make sense.
That's an PHY
news.synopsys.com
The problem here is not the 5nm, but the Samsung part.
This article doesn't specifically say that they are bringing LPDDR5X to the 8LPU. You're the one who assumed that.That's an PHY
Also, LPDDR5X isn't proof of anything regarding node. Nvidia can reuse their own controller across multiple nodes if desired.
Synopsys announced an expanded collaboration for ready to use external IP on Samsung Foundry and that included LPDDR5X for proccesses like 8LPU:
Synopsys and Samsung Collaborate to Deliver Broad IP Portfolio Across All Advanced Samsung Foundry Processes
Synopsys Interface & Foundation IP for Samsung 8LPU, SF5(A), SF4(A) and SF3 Processes Accelerate Path to Silicon Success for Advanced SoC Designs Highlights: Synopsys Interface IP for the most...
Thats what i tought too, but maybe they choose the strong components because the node is weak and its custom made with some efficiency gains, we know the prices skyrocketed, look at the ps5 pro 4N i think. I still hope its 5nm, think it will be 7nm and prepared for it being 8....
The ps5pro is 5nm isn't it? Which is basically the same node. Node names are confusing on purpose. Nvidia modified 5nm = 4nm.
Kind of a tough question. I think Rockstar picks 30 FPS because they have certain design goals in mind, and I also don't think their games would be nearly as ambitious if the target was 60 FPS. If GTA5 targeted 60 FPS on 7th gen I can't imagine how compromised the game would have been compared to what we got, and look at the performance even when they targeted 30 (it also had some performance issues on 8th gen consoles too, but not as pronounced):
Could we maybe see a 40FPS mode? Possibly, but I wouldn't count on it
edit: so tl;dr for my answer, the consoles could do 60 but it would be a completely different game
yes 4n is a subnode and ps5 pro is a bigger chip. I think we will be surprised either way, the combination of a custom made chip from the ground up with backported rtx4000 features, infused a with highly optimized Nvidia/Nintendo OS, a powerful t-239 chip with good ram size, speed and no bigger bottleneck + dlss with transformer technology... i think PS4 Pro visuals should be possible, not in raw power, but in todays standarts, thats not necessary anymore. Just look at the ugly unsharp RDR2 ps4 pro version with 1440p checkerboard upscaling, its so damn ugly i couldnt play it for 5 minutes straight and iam absolutely sure switch 2 can do better with dlss 4 technology
Nvidia made the Tegra X1+ that used LPDDR4X, and they made use of that extra power of the SoC and RAM for the 2019 Nvidia Shield TV. It's not a custom design by Nintendo like the T239 is.
fascinatingNo, according to this source who seems to know what they're talking about, it was a separate chip. The first reply here.
"
This is a good question.
Originally they planned to have two separate Cell chips on the PS3; A cell-based CPU, physics & geometry chip(the Cell chip we got), and a Cell-based pixel chip(GPU).
The Cell-GPU would have had 4–6 of the Cell SPUs and then some fixed function graphics hardware, like z-check(ROP) and texturing units (TMU)s.
Having similar architecture on all number crunching cores of the system would probably have given them some nice synergy advantages, same code could execute in either the GPU chip or in the CPU chip, and any improvements they would make to their compilers would benefit both physics/geometry side and the pixel side.
At some point they understood that they cannot get the cell-based GPU working, either in the schedule they had, or at the performance they wanted. This happened when the development of the console was already quite far and there was not very much time to the release.
So they quickly needed a different GPU for PS3. So they went shopping in panic and nVidia offered them a slightly customized geforce 7800. When PS3 was released, the GPU was already obsolete feature-wise, and they also had no synergy advantages between the CPU and GPU because the GPU did not have those cell SPUs anymore."
Why did Sony decide to go with the cell processor for its PS3?
Answer: This is a good question. Originally they planned to have two separate Cell chips on the PS3; A cell-based CPU, physics & geometry chip(the Cell chip we got), and a Cell-based pixel chip(GPU). The Cell-GPU would have had 4–6 of the Cell SPUs and then some fixed function graphics hardware...www.quora.com
No, according to this source who seems to know what they're talking about, it was a separate chip. The first reply here.
"
This is a good question.
Originally they planned to have two separate Cell chips on the PS3; A cell-based CPU, physics & geometry chip(the Cell chip we got), and a Cell-based pixel chip(GPU).
The Cell-GPU would have had 4–6 of the Cell SPUs and then some fixed function graphics hardware, like z-check(ROP) and texturing units (TMU)s.
Having similar architecture on all number crunching cores of the system would probably have given them some nice synergy advantages, same code could execute in either the GPU chip or in the CPU chip, and any improvements they would make to their compilers would benefit both physics/geometry side and the pixel side.
At some point they understood that they cannot get the cell-based GPU working, either in the schedule they had, or at the performance they wanted. This happened when the development of the console was already quite far and there was not very much time to the release.
So they quickly needed a different GPU for PS3. So they went shopping in panic and nVidia offered them a slightly customized geforce 7800. When PS3 was released, the GPU was already obsolete feature-wise, and they also had no synergy advantages between the CPU and GPU because the GPU did not have those cell SPUs anymore."
Why did Sony decide to go with the cell processor for its PS3?
Answer: This is a good question. Originally they planned to have two separate Cell chips on the PS3; A cell-based CPU, physics & geometry chip(the Cell chip we got), and a Cell-based pixel chip(GPU). The Cell-GPU would have had 4–6 of the Cell SPUs and then some fixed function graphics hardware...
What games do they use or have had a stock run with 32GB cartridges?
I think it’s just that we’ve reached the end of the race to the common denominator in the basic form of hardware architecture. That doesn’t necessarily mean you’ve landed on the absolutely ideal architecture but that your odds of making something worse instead of better have gotten too high, too much risk for the possible reward. At that point it makes sense to unify around that common denominator and start focusing elsewhere. I think that is what happened here. The last few times someone tried to go off the common path all they got for it was difficult development, high cost, and poor backward compatibility and preservation of the software. Not worth it, so don’t do it anymore.
The 2019 shield sold a small fraction of the Switch,
From the leaks its going to be big on simulations. Things like weather, water, physics, hair etc
How much visuals will they put on the CPU to assist the GPU? How much physics and complex NPC behavior we will get? I think it is a lot of questions, but I don't think it will be that bad if they programmed it right.
Even if that was the case, we can't compare the Switch 2 using LPDDRX5 just to downclock it like Mariko did using LPDDR4X. The 5X is more or less an overclock without the kind of benefit 4X was able to have, thereby, it makes little sense that Nintendo in a custom design would opt for something they had no intention on using to its full extent.
I would expect a video similar to the Switch reveal. That was about 3:40 seconds. Very much about showing off the hardware and its concepts while teasing game footage. Nintendo notoriously refused to state Skyrim was coming to Switch even though they literally showed it in the teaser for Switch
You're talking about the people's champion Kaze Emanuar! This man use every bit of a byte, fold his polynomials, and give us 9 mb of ram in the 64.
So we must be getting some new software feature if that's the case.
Voice/party chat overlay (Community/Chat) for the C-button. Brings up a UI over the game and pauses it. Kinda like the Xbox menu or PS menu when you press the home buttons.
If you can't love me at my 8nm, you don't deserve me at my 5nm.
Well, I believe people can tell at a glance that it is a second-generation product.
Just to add from what I already mentioned from this quote, when did one product selling more or less than another mean anything in terms of which product the hardware was designed for? The Tegra X1 was "not" designed first and foremost for Switch v1, but Switch v1 very likely sold the most using it.
