• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

NVIDIA to Acquire Arm for $40 Billion, Creating World's Premier Computing Company for the Age of AI

It's not like they can't afford it. In fact Intel might welcome the influx of money.

Nope. Intel has been doing everthing in their power so x86 remain on existing player. To the point they wiling to sour the initial cross licensing deal they made with nvidia in 2004 just because they saw some glimpse about nvidia interest in making their own x86 based cpu. If intel wiling to license x86 to nvidia they then the settlemen in 2011 would not mention "nvidia parmenently barred from any form of x86 licensing".
 
Son looked at his failure of WeWork and probably has been forced to go through this, giving up ARM in a competitive mobile era is bad decision esp Japan lost Toshiba's Memory to Kioxia, probably Sony is one top still left but they are not a tech corporation. Nvidia will gain a lot through this, now they can create a full chip for their AI industry plus a win for licensing costs from Apple and others. Huang did a smart move.

I feel like AMD is about to find itself in a difficult position. They are a small company compared to intel, and before AMD only had to contend with intel. Now they have to contend with both intel and Nvidia in the cpu market, which are both much larger. I think you'll soon see more inroads in desktop and laptop and servers with arm. I can see AMD being squeezed out more and they don't have other acquisitions to diversify with like intel and NV has (buying other companies over the years and branching out in other markets)

This ARM is not going to replace the Server market. It's not possible, the Software ecosystem is huge that cannot change anytime soon, AMD EPYC is very successful product and with new Milan it's going to wipe the floor with Xeon, So x86 is not going anywhere esp looking at Intel having already prepared for the AMD onslaught on the consumer DIY space and DC markets plus the most important thin and light bga trash market as well. Qualcomm's 8cx is not going to achieve anything over a U series Ryzen or Intel's new Tiger Lake. Plus Qualcomm abandoned their ARM server dreams and Marvell is going custom instead of the off shelf ARM parts for the markets and there are no other big players, Altera has to prove it and you have Nuvia which is all magical smoke for now & AWS Graviton 2 won't replace their EPYC and XEON, it's that usual Amazon's business of getting into everything (Lumberyard engine for games, Twitch, Amazon Originals etc) and also have a cheaper options for low end EC2 HW & Apple is only catering for their own loyal customer base where Mac is also devolved into looking like a mobile UX and will be a hybrid OS plus now the SoC from A series which is not going to impact the MS Windows / Linux or Datacenter RHEL. And imho ARM is always custom B.S one cannot simply run any software or OS just like that unlike on an x86 hardware, the OS needs blobs and proprietary drivers (at-least on Android). So AMD is not going to feel any heat from ARM or Nvidia on the CPU side, nor Intel is going to feel, they must innovate on the Lithography now to keep up that's all. Both x86 giants will be going the Big Little too soon, Intel already has it, AMD will also venture as per patents and that will take thin and light garbage in the future.
 
The Cortex-X1 has nothing to do with NVIDIA’s Tegra X1. The Cortex-X1 is a recent high performance quad core architecture design to license from ARM. The Tegra X1 is still using ARM’s Cortex-A57 cores.
 
It's important to note that AMD Ryzen latest gen CPU's have 0 security issues to my knowledge, ARM, Intel, were both plagued by Meltdown and other security issues. So if I were a business, small or big, I would still be banking on AMD CPU's personally.
They have issues aswell. Google it

p.p. You have nice imaginary PC build btw.
 
Last edited:
I'm just wondering if this will become Nvidia's "AMD buying ATi" kinda moment.
 
The transaction does not include Arm's IoT Services Group.
It's starting already...
Nvidia isn't interested in a lot of what ARM has been doing, so I wonder what else they'll either sell off, be forced to part with due to government regulation for this deal to close and what they'll simply just stop developing due to cost related issues.

ARM will never be the same after this, that's for sure. I'm expecting Adreno to be gone first and foremost, as why would Nvidia want to use an AMD derived graphics IP in their products? They want to license their own graphics IP, despite the fact that they've proven it's inferior in terms of power usage if nothing else.

Then we'll see what else they throw out, most likely the video codecs will go as well. I hope they're sensible enough to keep the MCU cores, even though the focus here seems to be in "AI", Nvidia ought to be smart enough to realise they need those for the peripheral chipsets, or maybe they'll just consider that a job for someone else.

If you really start looking, you'll see there are ARM based cores in so many things. Even AMD's chipsets has an ARM core built into them...

It's definitely not a bright future ahead and Apple ought to take a big long look at the direction they're heading in.
 
I'm just wondering if this will become Nvidia's "AMD buying ATi" kinda moment.

Nvidia's CEO did propose a merger with AMD back in 2005. He was a former CPU engineer working for AMD at one point in his career, before he cofounded Nvidia.

Of course, the AMD CEO (at that time) rejected that offer and decided to sink much of AMD's money into acquiring ATi above their (then-current) value. IMO, this dragged both down for almost a decade and allowed Intel to just gouge the market with increasingly expensive quad cores and a new motherboard socket every two years. Sometimes, this action was blatantly transparent (1151 two "incompatible" versions in particular). Nvidia and ATi were a bit more competitive, but it's not hard to see that fight as largely one-sided, too. Just not as bad.
 
Nvidia's CEO did propose a merger with AMD back in 2005. He was a former CPU engineer working for AMD at one point in his career, before he cofounded Nvidia.

Of course, the AMD CEO (at that time) rejected that offer and decided to sink much of AMD's money into acquiring ATi above their (then-current) value. IMO, this dragged both down for almost a decade and allowed Intel to just gouge the market with increasingly expensive quad cores and a new motherboard socket every two years. Sometimes, this action was blatantly transparent (1151 two "incompatible" versions in particular). Nvidia and ATi were a bit more competitive, but it's not hard to see that fight as largely one-sided, too. Just not as bad.

You're missing some 50% of this story, as AMD at the time screwed up a lot and didn't quite understand what they were buying. As such, they sold off profitable ATI business units to try and recuperate some of their investments in GPUs. ATI was actually a lot more than just a GPU company at the time. There was also quite a lot of management friction and some other issues that caused further problems, but my point is that it's not as simple as you're making it out to be either.
 
I expect to have a family and a career in next couple of years. My gaming time is going to basically vanish, so I imagine this rtx 3080 and zen 3 build will be my last. I probably will never get to experience ARM cpu's

You don't have a cellphone?

While they are using their own designs, the base instruction set is still ARM based and Apple is required to license it. NVidia might increase the pricing for Apple or deny them a license outright. Apple is not in a favorable position ATM.

MIPS is crying in a corner somewhere, if NVIDIA-ARM is mean they always have that option for cheap.
 
After the Wework disaster I guess Softbank didn't get the memo, this is an even bigger eff up :shadedshu: :wtf:
 
MIPS is crying in a corner somewhere, if NVIDIA-ARM is mean they always have that option for cheap.
The POWER ISA is another option if Apple wants to move back.
 
It's starting already...
Nvidia isn't interested in a lot of what ARM has been doing, so I wonder what else they'll either sell off, be forced to part with due to government regulation for this deal to close and what they'll simply just stop developing due to cost related issues.

ARM will never be the same after this, that's for sure. I'm expecting Adreno to be gone first and foremost, as why would Nvidia want to use an AMD derived graphics IP in their products? They want to license their own graphics IP, despite the fact that they've proven it's inferior in terms of power usage if nothing else.

Then we'll see what else they throw out, most likely the video codecs will go as well. I hope they're sensible enough to keep the MCU cores, even though the focus here seems to be in "AI", Nvidia ought to be smart enough to realise they need those for the peripheral chipsets, or maybe they'll just consider that a job for someone else.

If you really start looking, you'll see there are ARM based cores in so many things. Even AMD's chipsets has an ARM core built into them...

It's definitely not a bright future ahead and Apple ought to take a big long look at the direction they're heading in.

Why would adreno be gone? Adreno is owned by Qualcomm not ARM.
 
well, I guess we now know who is pushing for Skynet...
 
The POWER ISA is another option if Apple wants to move back.

Always, of course. But I think apple is still salty they couldn't make a laptop version of the PPC970 AKA G5 processor.

MIPS is a better candidate for lowpower devices like Apple seems to love.
 
This can end up two ways.

Nvidia is throwing money and innovation in the ARM platform without trying to find ways to lock it. The ARM platform becomes the main CPU architecture in 10-15 years (or less?), covering from high efficiency simple cores for IOT, to high performance, but less efficient cores for HPCs, AI and gaming. Nvidia is the new Intel. Nvidia can offer from a simple smartwatch to a supercomputer ALL with Nvidia technologies/hardware in it. Intel (and AMD?) are left in the dust.

Nvidia plays nice in the beginning, but starts doing it's things gain. It keeps the best ARM CPU cores and technologies for itself, makes life for other companies much harder, forcing them to start searching for alternatives like RISC-V. ARM architecture is dead in 10-20 years.
 
The transaction does not include Arm's IoT Services Group.

That's really odd and confirms a lot of concerns people had, it shows they have little interest in ARM's main mobile technology domain and that they are after something else.

It was a good run, ARM showed that mobile computing didn't have to be absolute crap, I don't even want to know what will follow in the next few years.

This can end up two ways.

Nvidia is throwing money and innovation in the ARM platform without trying to find ways to lock it. The ARM platform becomes the main CPU architecture in 10-15 years (or less?), covering from high efficiency simple cores for IOT, to high performance, but less efficient cores for HPCs, AI and gaming. Nvidia is the new Intel. Nvidia can offer from a simple smartwatch to a supercomputer ALL with Nvidia technologies/hardware in it. Intel (and AMD?) are left in the dust.

Nvidia plays nice in the beginning, but starts doing it's things gain. It keeps the best ARM CPU cores and technologies for itself, makes life for other companies much harder, forcing them to start searching for alternatives like RISC-V. ARM architecture is dead in 10-20 years.

Oh I wonder which will be it. Hmm ...
 
Last edited:
Doesn't make sense - didn´t nVidia move to RISC-V for there GPU controller design? I read that somewhere in 2015 or 2016........I think, could be wrong tho.....
 
This will be very bad for the industry. There are some very good reasons why nVidia is known to be hard to work with.

I don't understand why Apple did not buy ARM, as Apple and nVidia do not get on, and nVidia will not make things easy for Apple going forward. I wonder if Apple have some kind of protection from whatever happens to ARM?
 
Nvidia's CEO did propose a merger with AMD back in 2005. He was a former CPU engineer working for AMD at one point in his career, before he cofounded Nvidia.

Of course, the AMD CEO (at that time) rejected that offer and decided to sink much of AMD's money into acquiring ATi above their (then-current) value. IMO, this dragged both down for almost a decade and allowed Intel to just gouge the market with increasingly expensive quad cores and a new motherboard socket every two years. Sometimes, this action was blatantly transparent (1151 two "incompatible" versions in particular). Nvidia and ATi were a bit more competitive, but it's not hard to see that fight as largely one-sided, too. Just not as bad.

If AMD didn't purchase ATi back then they would be gone today. Their custom SoC business (PS4, Xbox One) is what kept the company afloat in the Bulldozer days.
 
Back
Top