cheapchips
Veteran
So what could an exceedingly unlikely Intel gen 9 soc look like?
It'll be on 7nm or 5nm given the timeframe. I don't know where that leads.
It'll be on 7nm or 5nm given the timeframe. I don't know where that leads.
Predict: Next gen console tech (9th iteration and 10th iteration edition) [2014 - 2017]
- Thread starter Shifty Geezer
- Start date
- Status
I would still bet on a 10nm even if it were Intel. Although, what Intel calls 10nm may be denser and better performing than what TSMC/Samsung deliver as 10nm. I guess 10nm is supposed to be out in late 2016/early 2017 and give it another year or so for refinement and maturation so that would put it around 2018 for a mature 10nm process.
Getting OT of course, maybe sort of on topic since these are key technology companies in the console space. but if it's any consolation not surprisingly Intel has been caught pretty flat footed in the mobile market which is arguably the future if not the present, and dominated by ARM. They are still trying to get in but losing money, and of course there are hardly any Intel chips in a smartphone at all.
I guess it's kind of scary that Intel eventually they can buy their way to success in the mobile market as well, but their business model is so fundamentally different than ARM's I have by doubts. I guess it is kind of like Microsoft, working to buy there way into all these markets they aren't core at, like Bing or One Drive, to varying degrees of success. Or even Google for that matter, that tried to take on Facebook with something like Google+ that failed quite miserably. A lot of these huge companies try to take over every market and it's just never going to be possible to be the best at everything, there is always some company that this is the one thing they do and they're better at it and native at it. Netflix is a good example. It wasn't some giant multimedia conglomerate that took over that lucrative streaming market as you'd have expected, it was an upstart that lived there from the beginning and focused on it and delivered a great product while the huge corps were probably still having clueless meetings about it.
Getting back a bit more to consoles tech speculation, is AMD even going to be around soon? My hope would be Samsung buys them, but keeps them as they are now, just infusing them with plenty of cash. It boggles my mind AMD Market cap is something under 2 billion. That's absolutely NOTHING to these these large corporations, how have they not been snapped up yet? What did Facebook pay for Instagram? That's absurd, the patents alone AMD are sitting on have got to be worth 10X that.
I guess it's kind of scary that Intel eventually they can buy their way to success in the mobile market as well, but their business model is so fundamentally different than ARM's I have by doubts. I guess it is kind of like Microsoft, working to buy there way into all these markets they aren't core at, like Bing or One Drive, to varying degrees of success. Or even Google for that matter, that tried to take on Facebook with something like Google+ that failed quite miserably. A lot of these huge companies try to take over every market and it's just never going to be possible to be the best at everything, there is always some company that this is the one thing they do and they're better at it and native at it. Netflix is a good example. It wasn't some giant multimedia conglomerate that took over that lucrative streaming market as you'd have expected, it was an upstart that lived there from the beginning and focused on it and delivered a great product while the huge corps were probably still having clueless meetings about it.
Getting back a bit more to consoles tech speculation, is AMD even going to be around soon? My hope would be Samsung buys them, but keeps them as they are now, just infusing them with plenty of cash. It boggles my mind AMD Market cap is something under 2 billion. That's absolutely NOTHING to these these large corporations, how have they not been snapped up yet? What did Facebook pay for Instagram? That's absurd, the patents alone AMD are sitting on have got to be worth 10X that.
Instagram was a $1Bn cheque. Whatsapp was $19Bn, which is mental. Anyone could buy AMD, but the real question is why? Anyone interested in buying them would need to think very hard about what they get out of them and what return they will make in the long run.
Patents are only worth something if they give you competitive advantage and/or you can license them. Owning lots of patents that provide neither benefit isn't a particularly positive thing. It would appear that these patents aren't particularly valuable to AMD because they aren't managing to make crazy money from them and they aren't stopping Intel et al from out-competing them.
Everybody understands that Intels current strategy of essentially giving away their product to mobile vendors is unsustainable. What they are doing is putting financial pressure on Qualcomm, Mediatek, et al, giving them narrower margins with which to develop competitive products. The more squeezed their competitors are, the better Intels position. Plus if Windows on x86 makes sufficient inroads, they will also have cornered a relatively juicy part of the mobile market. The problem, as you note, is that as long as ARM based chips provide an alternative platform, this segment will never provide the margins that Intel enjoys in their virtual monopoly market segments. And the two major sources of revenue and profit in mobile, Apple and Samsung, have no interest in being dependent on Intel in this space. The best Intel can get out of their mobile effort is extra production volume, so that they don't have to cut back fab capacity too much. Similar could be said about the comparatively tiny console market.
Still, sustaining excess fab capacity seems a rather weak motivation. Trying to find and foster a new market (and locking it in to your product) seems like a much better proposition. Of course, that requires foresight and/or creativity.
I really wonder now with AMD's troubles (again) if that jeopardizes x86's future in the next consoles. The possibility of AMD going bankrupt or being bought out is significantly higher than it was three years ago. Would Sony or MS be willing to take a gamble on their x86 offerings now?
I'm going to predict that the next consoles will make another switch and this time to ARM, Nintendo won't go x86 either.
I'm going to predict that the next consoles will make another switch and this time to ARM, Nintendo won't go x86 either.
AMD patents value when it comes to x86 and intel isn't based on the amount of revenue they generate nor giving AMD an advantage in terms of competition. Their value lies in giving AMD access to technology that lies behind Intel's patents. If AMD didn't have those patents there would nothing to offer Intel for a cross licensing agreement. Intel would be extracting licensing fees from AMD which AMD would have to pass that to the customers.
Last edited:
D
Deleted member 11852
Guest
AMD need to exist if only to push Intel and Nvidia but sometimes I think this is the only reason they exist. They've just pulled out of the micro server market losing a lot of money in the process and I wonder why they thought that was going anywhere in the first place
It's ironic, I don't want AMD to go away and their CPUs and GPUs seem to represent good value for money but they have nothing in their current CPU and GPU lineup that would appeal to me if I were building a PC. I've only owned three AMD-based products, the PlayStation 4, my retired 2009 iMac with Radeon HD4850 and over a decade back a I briefly toywed with an Athlon XP based PC before moving to a Pentium4 system.
In which case their patents are nigh worthless to anyone not wanting to go head-to-head against Intel with x86 chips?
Its not just patents related to x86, the cross patent licensing encompasses all intel and amd patents. So how much do you think Intel would charge for access to its entire patent portfolio? Thats a round about way to determine how much AMD's patents are worth.
Plus, AMD being the sole competitor in the x86 space has nothing to do with no one wanting to produce x86 chips and compete with intel. Intel restricts everyone outside of AMD. Intel had problems with GF producing x86 chips and won't allow AMD to license out x86 designs. Its why AMD sells the actual console apus to MS and Sony versus licensing out the design and letting MS or Sony source production.
So if you want to produce x86 chips or not, there is definitely a lot of value in AMD patents. Owning them would make entering the x86, gpu and cpu market easier in general. Simply acting as leverage to enter into cross patent licenses with others. If you are already operating in that space you can extract more favorable terms for cross patent licenses. Or you could simply charge others license fees for using the technology based on your patents.
Last edited:
A worthy question with uncertain answers.
The way things are structured, you can't do both. A bought AMD or one getting a metric ton of free cash isn't the same AMD in a lot of legal clauses.
It would be paying 2 billion or more for the privilege of getting 2+ billion in debt, operations that lose tens to hundreds of millions of dollars a year, for a customer base that is either served by Intel/Nvidia or is not part of the market AMD addressed.
It would also be an automatic entry into a protracted legal tangle with a multibillion dollar Intel, who--if the buyer has some kind of electronics presence--is likely a supplier or licenses something else they already depend on.
They're worth as much as buyers are willing to pay. For a company that doesn't care about AMD's niche, it's limited and it's quite possible to get them cheaper should AMD be liquidated.
What troll is going to have the cash to outbid the giant companies that make products, who might want to buy them (or just Intel for just that reason)? And what company with actual operations and products is going to risk getting nuked by Intel if they try something like that?
I was making an argument against Shifty assertion that AMD's patents aren't worth much because they don't make crazy cash for owning them. Plus being a giant company doesn't eliminate you from acting as a troll like the Rockstar Consortium.
Anyone with designs on operating in this space or already operating in this space would love to own AMD patents. Owning them has the potential to get more privilege access to Intel portfolio as well as make it easy to navigate the chip design and semiconductor market. Nvidia would probably love to own AMD patents as their cross patent agreement with Intel doesn't encompass the rights to produce x86 chips. Owning AMD patents would provide additional leverage in any attempt to change that reality and ensure a continued presence in producing x86 if Nvidia could change the current reality. If MS made a serious foray into a semiconductor then owning AMD patent would go a long way in forming cross patent agreement with other semis.
Regardless if you aren't making a tons of cash from licensing your patents, the value of your portfolio can be calculated by the amount of additional fees you would have to pay on an annual basis if you own no patents. Of course, that wouldn't apply if your company had the ingenuity to freely operate by coming up with performant workarounds for all the patented technologies that are commonly employed in the space you operate in.
Last edited:
There's pretty much no company that could try to pull a Rockstar that doesn't base a lot of their business on hardware provided by Intel. Intel could make things painful enough for a long time, and I thought the idea was to make money after paying 2 billion+ for AMD's 2 billion+ in debt and tens to hundreds of millions of dollars of annual losses. Given Intel's vast reach into other fields, they would have a lot of their own IP to bear in a countersuit.
Given how badly AMD has shrunken in scope and given away IP, and the aging of its most significant IP, the clock is winding down on a project that would need years to even get started.
If they are operating in this space, they are already exposed. You have to be an actual troll that produces nothing so that there's nothing to counter with.
There's a fair amount of leeway on this. Being part of an already existing cross-licensing agreement may make an injunction less likely, and if Intel must it might just litigate to the point of patent expiration and cut a check if there is a finding of infringement. Intel has done this before, and that is the upside to buying AMD with its debt and money-losing operations, or somehow outbidding Intel at the fire-sale.
Patent portfolio are like tactical nukes, so I have no problem with your assertion. But Intel would find it just as difficult with its monopolistic control over the x86 markets to try to hurt someone by depraving them of x86 hardware. Intel would easily run into antitrust issues by using its monopolistic powers against someone for just trying to get Intel to pay licensing fees for patents that company owns.
Plus, how you go about gaining AMD's patents doesn't absolutely require you purchasing AMD as a company nor its debt in all cases. If AMD decided to divest its x86 business, it would at most have to let its agreement with Intel lapse. If AMD went into bankruptcy, its agreement with Intel would probably be voided and could possibly be sold off in an effort to pay off its debt.
The use of trolls have been employed by giants as a way to mitigate exposure. You hand over control of part of your patent portfolio and then take a cut of the licensing fees. There hasn't been a case I know of that has set a precedence in an attempt to challenge this strategy. I've havent heard of a case of a company suing another company's shareholder due to a patent dispute and because that shareholder operates in the same space as the plantiff.
If AMD was ever in the mood to sell off its patent portfolio related to its cpu business it would in most cases mean that AMD has left the cpu business especially if we are talking x86 based patents. This creates a circumstance where AMD is no longer a competitor in the x86 market and gives Intel even more control. In all likelihood Intel wouldn't be allowed to participate because even in bankruptcy asset buyers are still subject to antitrust liabilities. No one can stop AMD from leaving the market but Intel can be forbidden from buying up portions of AMD.
Last edited:
AMD's Gpus are sometimes better than NVidia's and often times are extremely competitive. It was only recently with the 9x0 series that NVidia really pulled away and even then its only at certain resolutions.
I think AMD needs to simply get out of the CPU game and be done with it. That or focus only on mobile chips like NVidia is doing. Their desktop apu line up just gets slaughtered by intel but if they focused on making fast arm chips tied to their gpu tech they could succeed in mobile quite easily.
With global Xbox One sales sinking like a Star Wars fan's boner the when Ja Ja Binks appears, MS are going to be looking at a strategic correction requiring new hardware at the earliest sensible opportunity. Which is probably late 2017.
Options are limited. AMD are .... ah balls it's sad but AMD are in trouble. Power and bandwidth are increasingly issues facing even high performance kit. Could we see a next gen Power VR based console in 2017?
Options are limited. AMD are .... ah balls it's sad but AMD are in trouble. Power and bandwidth are increasingly issues facing even high performance kit. Could we see a next gen Power VR based console in 2017?
Sure let me count how many phone soc wins intel has
Five.
Nope. AMD tech is already killing it in the mobile space. Problem is that AMD's mobile division was sold to Qualcomm (radeon + adreno = anagram) in 2009. And the reason we haven't seen AMD make a run at smartphones is probably because Qualcomm asked for and got a non compete clause in that space.
- Status