Is Tegra 4 confirmed to be compatible with third-party modems?
- Status
Is Tegra 4 confirmed to be compatible with third-party modems?
I don't know. But if not, your supposition about the i500 readiness can't be true at the same time (though, I guess availability could be an issue). Since they have stated (also by the same picture you posted above) that Tegra 4 phones will go into production during the first half of this year.
The Icera i500 4G LTE baseband modem is already production ready. The issue is that carrier certifications need to be performed on the actual 4G LTE-equipped end user devices. This adds at least six months to the development cycle of these 4G LTE-equipped devices. Nevertheless, Wifi-only tablets are still the most cost-effective and most widely purchased tablets (even though the market for 4G LTE-equipped tablets will continue to grow over the next few years). Having Tegra 4 + i500 powered tablets available by the holiday season in 2013 is not bad and within the realm of expectations.
We've known for a while from Anandtech that the first Tegra 4i-powered devices would be available no sooner than end of 2013, with the majority of Tegra 4i-powered devices not being available until Q1 2014 at the earliest. Qualcomm has gone on the record at MWC 2013 as saying that S800-powered devices will be available in the second half of 2013 (and not before middle of the year). So we've known for a while that S800 would come to market many months before Tegra 4i. This is not the end of the world for NVIDIA, because Tegra 4i silicon is reportedly half the cost of S800 silicon, and Tegra 4i is aimed at much more affordable devices than S800.
Making an SOC compatible with an external modem is just a matter of adding an interface. SPI, I2C, some USB variant, PCIe, take you pick.
How would it not be compatible?
Since the lack of LTE modem was often cited as a major problem for NVIDIA, I figured there might be issues with the use of third-party modems, maybe extra validation work or something.
It's true that WiFi tablets represent most of the market, but it's good to have a cellular option on top of it. Last year, most of the successful tablets had a 3G option (iPad, Nexus 7, Galaxy Tab, etc). This year, the same should hold true for LTE. But I suppose a third-party modem would work just as well, or OEMs could just introduce the WiFi model first and the LTE variant as soon as it's available.
Yes, I was just under the impression that the gap would be smaller. But it's true that they don't really compete against each other, anyway. I guess Qualcomm will probably offer dual-Kraits in this market segment.
I think the issue is that many IHVs make both modems and soc's and they sell them together as a package, with drivers and stuff ironed out. Since SoC's are so similar, not offering a complete package can often swing a deal the other way just because using a package means not having to validate this stuff.
This is standard practice is the general semiconductor business. Not uncommon to see 2 chips together in a bundle be almost as cheap as each chip individually.
E.g. Ethernet switch + fiber PON transceiver bundled is $10, but PON transceiver alone is $9. (Prices are a complete invention.)
It's a great way for an incumbent to prevent challengers from entering their market, thus forcing the challenger to provide a full solution (which they usually don't have.)
I've always assumed there must be some anti-competitive rules being violated this way, but that's apparently not the case because everybody who can seems to do it.
My understanding of anti-competitive rules is that they only apply when one company is in a clearly dominant position. I suppose Qualcomm may be nearing that position by now.
NathansFortune
Regular
Tegra 4i is also a piece of shit based on 2011/12 A9 architecture rather than A15. S800 and T4i are not comparable, S800 will destroy T4i in just about every benchmark (and looking at the supposed leaked LG/Sony ones, Adreno 330 even bests T4 in the GPU stakes).
If T4i with it's quad A9 is going to be on sale in Q1 2014 then what hope does T5i have? Is T4i going to have a super short lifecycle or something because Nvidia need to get T5i with Kepler based ALUs and quad A15 with an LTE modem sampling by Q2 2014 for H2 devices.
If Nvidia want to hold onto the gains they made with T3 then they need much, much better execution. They beat ATi and 3Dfx with raw power and marketing. Qualcomm are a very different prospect and Qualcomm haven't failed to deliver like ATi and 3Dfx did over and over again to allow Nvidia into the market. S800 is going to push almost every other SoC maker to the sidelines as it is power efficient, powerful, has a decent GPU and readily available. Nvidia don't have anything like that to push T4/T4i so they have to eat into their margins. Look at where that has left AMD after years of pushing down prices.
That is simply not true. The CPU used in Tegra 4i is a rearchitected quad-core R4 Cortex A9 that shares many of the same features and performance enhancements as the Cortex A15: http://www.nvidia.com/docs/IO/116757/NVIDIA_Quad_a15_whitepaper_FINALv2.pdf
Right. I wonder how in the world you could know that when neither of these SoC's has been benchmarked yet, and when both of these SoC's are using rearchitected CPU/GPU hardware.
Maybe you are confused here, because there are no leaked benchmarks showing that Adreno 330 outperforms ULP Geforce in Tegra 4.
It seems clearer and clearer to me that Nvidia future is bound to how successful their CPU are which is quiet ironic.
With consoles setting a ceiling on the PC realm for the upcoming years, it should not be that tough even for Intel to ship in the upcoming years APUs that have perfs in the ball park of next generation console.
Will they? AMD seems interesting in leveraging gaming, Intel? It is unclear but it seems that they won't push more than 4 big cores in the personal realm which means more and more silicon free to be invested in the GPU. If both go in that direction, Nvidia will be in a world of pain if they can't establish them selves in the ARM realm, to do so they need competitive CPU foremost.
They may maintain a niche in the professional market for their high end GPU and the matching software (software being the most important part as for IBM), but will they generate enough money to compete (in the long run) with imo the upcoming wave of many cores products (imo IBM is there with its power a2 and the products based on the arch, Intel won't give up on the segment Xeon Phi are is just the first gen of products, I expect many cores based on ARM v8 to make it to the market in not that far future). Though successfully turning into a lesser IBM won't be that easy as I'm not sure Nvidia will have the software that really lock in their costumers.
EDIT
Intel could get serious about games
With consoles setting a ceiling on the PC realm for the upcoming years, it should not be that tough even for Intel to ship in the upcoming years APUs that have perfs in the ball park of next generation console.
Will they? AMD seems interesting in leveraging gaming, Intel? It is unclear but it seems that they won't push more than 4 big cores in the personal realm which means more and more silicon free to be invested in the GPU. If both go in that direction, Nvidia will be in a world of pain if they can't establish them selves in the ARM realm, to do so they need competitive CPU foremost.
They may maintain a niche in the professional market for their high end GPU and the matching software (software being the most important part as for IBM), but will they generate enough money to compete (in the long run) with imo the upcoming wave of many cores products (imo IBM is there with its power a2 and the products based on the arch, Intel won't give up on the segment Xeon Phi are is just the first gen of products, I expect many cores based on ARM v8 to make it to the market in not that far future). Though successfully turning into a lesser IBM won't be that easy as I'm not sure Nvidia will have the software that really lock in their costumers.
EDIT
Intel could get serious about games
Last edited by a moderator:
Nvidia one of few highlights in an overall down market day as the company hosted its annual investor conference and announced a plan to return 1 billion $ to investors through dividend increases and stock buybacks.
The stock ended the day up 2.51% at 13.09 with an intraday peak of 13.23.
The stock ended the day up 2.51% at 13.09 with an intraday peak of 13.23.
Weird. Net income took a hammering down from last quarter (as expected) but it's still better than I thought it would be given AMD's graphics departments better results as well.
Something just doesn't add up however because PC shipments were way down. I suppose the logical argument is that the low end APU shipments were down but midrange to higher end (with discrete gpu) remained steady.
Something just doesn't add up however because PC shipments were way down. I suppose the logical argument is that the low end APU shipments were down but midrange to higher end (with discrete gpu) remained steady.
Silent_Buddha
Legend
Or just as many predicted. Tablets are only eating into the low end budget PC segment. The PC gaming segment continues to have modest growth and hence demand for discrete graphics remains unchanged or with positive growth.
We're basically seeing similar market trends as when everyone was predicting notebooks and laptops would soon be replacing all or most desktop PCs. Only this time it's doom and gloom for the PC industry as a whole due to tablets. There will be a balance reached at some point or the novelty of tablets will wear off just as it did with notebooks and laptops over a decade ago.
Regards,
SB
We're basically seeing similar market trends as when everyone was predicting notebooks and laptops would soon be replacing all or most desktop PCs. Only this time it's doom and gloom for the PC industry as a whole due to tablets. There will be a balance reached at some point or the novelty of tablets will wear off just as it did with notebooks and laptops over a decade ago.
Regards,
SB
Doom and gloom stories for the PC platform were persistent as far, as I can remember myself being interested in the industry (early 90's that is). I can't remember how many times the "elite" RISC/UNIX combo was just about to crush WinTel and drive the x86 architecture into oblivion -- a long running narrative throughout the 90's.
What's interesting is, in the history of computing, it's almost always something "coming from below." Mini-computers replaced mainframes (although not totally), and workstations replaced mini-computers, and PC replaced workstations. Very rarely something was "coming from above," actually I can't think of an example right now.
It's predicted, years ago, that ARM based computers could "come from below" to replace current PC. Just like mainframe still exists right now, PC and PC based workstations, PC based servers, are probably going to be exist for quite some time, but I won't rule out the possibility that ARM based mobile devices could "come from below" to replace a large portion of PC's current market share.
Silent_Buddha
Legend
That's true for the RISC/Unix/Linux/etc. rants from the minority of users.
But not true for the notebook/laptop "revolution" that happened in the late 90's early 2000's as notebooks and laptops became affordable (1000 USD and less). At that time you saw the same predictions as we're now having with tablets due to notebooks and laptops taking away significant market share from desktops.
And while desktops saw a brief drop in sales (around the release of WinXP we saw an almost exact same dip in desktop sales as we're seeing in PC sales currently), market share eventually stabilized and desktop sales went on the rise again.
I'm expecting something similar here as tablets become more ubiquitous and the novelty wears off (wanting to upgrade every year). Just as it did with notebooks where people felt they had to upgrade their notebook every 1-2 years. And where after X amount of years everyone that was going to replace their desktop had already replaced their desktop. And then some of which started buying desktops again.
The only thing I can see changing this dynamic is if tablets became true computing devices with a full blown OS. Either Microsoft with Win8 or Apple with MacOS. Something capable of running everything (productivity and entertainment) that can be run on a desktop. But for that to happen we also need the hardware for tablets to increase significantly in power, although Intel is quite close already. As someone on this forum has mentioned (sorry I can't remember your name) in a thread somewhere, Clovertrail based tablets can already handle all of his productivity software smoothly with the exception of large Excel spreadsheets (lots of data, not screen size.
). If Intel can get i7 or even i5 based performance into something in the 8-10 hour battery range at bit of a weight reduction, then I think we might be there.Then again, does that change anything since it's still basically a PC even if it's a tablet?
Regards,
SB
The problem they (x86) have is that their low end desktops still make up the vast majority of their sales. If this is what is being eaten by tablets then they still have to find something to use those low end chips on. Many of them are of course the worst chips of the wafer - the Celerons and Pentiums at 65W etc.
If you basically just say that market is disappearing, then what is going to happen? Is the whole market going to shift up in price so that Celerons are priced at i3 prices? Or are they just going to throw these lower end cpu's away? There has to be a knock-on effect somewhere.
AMD is in deep trouble under these circumstances because they just don't have the performance chips that are going to be left, and they are obligated to pay $1 billion per year to GF for wafers. Kaveri better be something special or they better have a really good plan B.
If you basically just say that market is disappearing, then what is going to happen? Is the whole market going to shift up in price so that Celerons are priced at i3 prices? Or are they just going to throw these lower end cpu's away? There has to be a knock-on effect somewhere.
AMD is in deep trouble under these circumstances because they just don't have the performance chips that are going to be left, and they are obligated to pay $1 billion per year to GF for wafers. Kaveri better be something special or they better have a really good plan B.
- Status
Similar threads
