NVIDIA's Project Kal-El

[hans007]

i dont think nvidia stock will continue to rise. tegra 2 is first which got them some wins. but its not even clear its the fastest dual core out there.

you have the samsung chip that is already faster, t he qualcomm chips and marvell ones. its not like there is no competition. they just have a very slight lead.

if anything tegra 1 was mostly a failure, and no one cared about nvidia until they got tegra2 out. and if anything nvidia stock has always had wild swings. its been going from 6 to 40s to single digits for years. not t omention we are in the super fed liquidity QE2 "all boats rising" stock market of 2010/2011.

 

[cotak13]

Well IDC supposedly nV already has some signifigant contracts....

One of which it turns out went sour on them (LG). There might be a reason why LG drops the Tegra so suddenly and quietly (avoiding questions it appears).

Overall there's nothing much to see here. You guys are getting worked up over a marketing graph. Whether nvidia delivers is yet to be seen.

I for one am a bit sceptical of the rational behind nvidia's work. What will you do with a 4 core processor in a power limited package? Now before people jump on about how it enables quicker to idle and better performance means more can be done with the tablet and smart phone... No one's sure tablets are going to really become that big and in smart phones power is much more limited because you can only cram so much battery into something that fits in a pocket. Thing is if tablets keep their form factor (vs becoming netbook like) they aren't much good for anything except for consumption. Not many will want to do real work on a tablet (work such as spread sheet etc). A more powerful mobile chip can maybe enable more interesting development in terms of gaming. But there you run into the fragmentation issue. With mobile software being sold for peanuts you want to minimize your cost and maximize your returns. Which means if 1 or 2 phones have a very powerful 4 core mobile processor you'll still design your stuff for the vast majority of phones that has 1 or 2 cores. The same with many other apps.

So in some ways I think these powerful multi core chips are solutions looking for a problem.

 

[BD2003]

Tablets are going to be huge. They already are. The iPad at least, is incredibly good at what it does. The one thing it needs more than anything else right now other than less weight is more raw power.

 

[endlessmike133]

Tablets are going to be huge. They already are. The iPad at least, is incredibly good at what it does. The one thing it needs more than anything else right now other than less weight is more raw power.

It also needs a better screen; preferably OLED.

 

[BD2003]

It also needs a better screen; preferably OLED.

Oh, I could come up with a whole list myself. But near the top of it is quad core SoC. My iPad has almost completely replaced my desktop/laptop use at home, which is primarily web surfing. Pages don't render as fast as they could/should.

As soon as they get the prices down, tablets are going to be as ubiquitous as phones and TVs. They do so many things so much better than a laptop/desktop that their success is beyond question at this point. It was an open question pre-iPad, but 1 year later the case is closed, I'd bet my life on it.

The question of who will be powering them is still wide open, and even if its not NVIDIA, as long as they bring the heat we'll all benefit from the competition.

 

[cotak13]

Oh, I could come up with a whole list myself. But near the top of it is quad core SoC. My iPad has almost completely replaced my desktop/laptop use at home, which is primarily web surfing. Pages don't render as fast as they could/should.

As soon as they get the prices down, tablets are going to be as ubiquitous as phones and TVs. They do so many things so much better than a laptop/desktop that their success is beyond question at this point. It was an open question pre-iPad, but 1 year later the case is closed, I'd bet my life on it.

The question of who will be powering them is still wide open, and even if its not NVIDIA, as long as they bring the heat we'll all benefit from the competition.

I think you need to review your logic.

First you say your primary use for a computing device is surfing. Then you say that tablet do so many things better than a laptop and desktop that their success is beyond question. It seems to me for you it does one thing better rather than many things better.

I think their success is a huge question right now.

Put it simply. The faster and more capable you make a tablet. The more it's going to become like a laptop. You want faster processing you can't get away from more heat which means eventually there needs to be active cooling. You need more memory to make your apps and web site load and run faster which means more space required for memory. To keep the same battery life the only choice is to make the device bigger which you already did because of the need for active cooling. You start using it more like a laptop with more apps will mean requirement for more storage which eventually leads to more space requirement. At some point the whole thing starts looking like a laptop, weight like a laptop and for all intend and purpose is a laptop.

 

[BD2003]

I think you need to review your logic.

First you say your primary use for a computing device is surfing. Then you say that tablet do so many things better than a laptop and desktop that their success is beyond question. It seems to me for you it does one thing better rather than many things better.

I think their success is a huge question right now.

Put it simply. The faster and more capable you make a tablet. The more it's going to become like a laptop. You want faster processing you can't get away from more heat which means eventually there needs to be active cooling. You need more memory to make your apps and web site load and run faster which means more space required for memory. To keep the same battery life the only choice is to make the device bigger which you already did because of the need for active cooling. You start using it more like a laptop with more apps will mean requirement for more storage which eventually leads to more space requirement. At some point the whole thing starts looking like a laptop, weight like a laptop and for all intend and purpose is a laptop.

You're conveniently forgetting that performance per watt increases over time as well. Otherwise we wouldn't have tablets, smart phones, or even the modern pc as we know it. The iPad already has the performance of a several year old laptop right now. It'll have today's laptop performance several years from now. Probably at half the price, half the weight and twice the battery life of the current iPad. And there will be a dozen competitors. Storage isnt an issue either. I remember paying $100 for a 16mb compact flash card less than a decade ago, now I can pick up a 16gb stick for half that. God only knows what I'll be able to get for $50 when stark comes out.

Don't be so short sighted.

 

[cotak13]

You're conveniently forgetting that performance per watt increases over time as well. Otherwise we wouldn't have tablets, smart phones, or even the modern pc as we know it. The iPad already has the performance of a several year old laptop right now. It'll have today's laptop performance several years from now. Probably at half the price, half the weight and twice the battery life of the current iPad. And there will be a dozen competitors. Storage isnt an issue either. I remember paying $100 for a 16mb compact flash card less than a decade ago, now I can pick up a 16gb stick for half that. God only knows what I'll be able to get for $50 when stark comes out.

Don't be so short sighted.

And you are forgetting performance per watt is a tide that lifts all boats. And you are forgetting that appetite for storage goes up never down. Improvement in tech just results in people eating it up one way or another. When I started building computers 32 megs of ram was a lot. Today even smart phones need more to be competitive. We haven't gotten to the point where improvement are pointless. If your argument was valid we'd all still be using pentiums and be very happy.

I think you are the one being short sighted because you look at the possible improvement and thinks wow that's going to be great. While forgetting improvement is never enough as computing history has demonstrated over and over.

In the end a faster tablet means even faster laptops and desktops. A tablet with N amount of storage will mean laptops and desktops with XN amount. Does anyone honestly think we'll stop at 1080p for video content? When we have imaging sensors that records twice that resolution already. And what about when 3D is actually viable for the masses? Storage, processing and memory will be used up.

 

[BD2003]

And you are forgetting performance per watt is a tide that lifts all boats. And you are forgetting that appetite for storage goes up never down. Improvement in tech just results in people eating it up one way or another. When I started building computers 32 megs of ram was a lot. Today even smart phones need more to be competitive. We haven't gotten to the point where improvement are pointless. If your argument was valid we'd all still be using pentiums and be very happy.

I think you are the one being short sighted because you look at the possible improvement and thinks wow that's going to be great. While forgetting improvement is never enough as computing history has demonstrated over and over.

In the end a faster tablet means even faster laptops and desktops. A tablet with N amount of storage will mean laptops and desktops with XN amount. Does anyone honestly think we'll stop at 1080p for video content? When we have imaging sensors that records twice that resolution already. And what about when 3D is actually viable for the masses? Storage, processing and memory will be used up.

I'm not at all saying the high end laptop or desktop is going to be replaced by tablets. Im not giving mine up either. You are absolutely correct. But performance of SoCs has now reached the point where they are a viable option for a general purpose computing device with it's own set of advantages or disadvantages.

To be perfectly honest, I hadn't really intended on getting into this kind of debate. I've so fully integrated a tablet into my own personal computing ecosystem that I honestly cant imagine how their success in the near future is even in question, and I'm honestly surprised there's anyone that does. But if you look closely at what I said, you'd see that we agree far more than we disagree.

 

[cbn]

And you are forgetting performance per watt is a tide that lifts all boats. And you are forgetting that appetite for storage goes up never down. Improvement in tech just results in people eating it up one way or another. When I started building computers 32 megs of ram was a lot. Today even smart phones need more to be competitive. We haven't gotten to the point where improvement are pointless. If your argument was valid we'd all still be using pentiums and be very happy.

I think you are the one being short sighted because you look at the possible improvement and thinks wow that's going to be great. While forgetting improvement is never enough as computing history has demonstrated over and over.

In the end a faster tablet means even faster laptops and desktops. A tablet with N amount of storage will mean laptops and desktops with XN amount. Does anyone honestly think we'll stop at 1080p for video content? When we have imaging sensors that records twice that resolution already. And what about when 3D is actually viable for the masses? Storage, processing and memory will be used up.

That is true, but it seems that smartphone hardware advances are the closest to being the most practical.

For example, I am sure IT professionals need plenty of processing cores and RAM for the tasks they do, but what about the average person? Will they be able to do all their web browsing, Office tasks, photo editing on lesser hardware like ARM? If so, that hardware has the chance of becoming the standard.

Maybe a good example of what I am talking about is the bobcat x86 processor? AMD saw a need for lesser (not greater hardware) for certain users.

 

[cbn]

Put it simply. The faster and more capable you make a tablet. The more it's going to become like a laptop. You want faster processing you can't get away from more heat which means eventually there needs to be active cooling. You need more memory to make your apps and web site load and run faster which means more space required for memory. To keep the same battery life the only choice is to make the device bigger which you already did because of the need for active cooling. You start using it more like a laptop with more apps will mean requirement for more storage which eventually leads to more space requirement. At some point the whole thing starts looking like a laptop, weight like a laptop and for all intend and purpose is a laptop.

I do agree that ARM is on the way to being installed in laptops or laptop like devices.

Maybe one saving grace of ARM is that it has traditionally been built on low leakage silicon whereas Intel has classically used High leakage silicon specially binned for ULV for its CULV chips. (Someone correct me if I am getting these details wrong.)

 

[drizek]

You can't compare todays memory prices to 2001. The industry is on a completely different scale now.

Mobile processors had been neglected up until around 2007, but since then they have been catching up to offer the same technologies with the same manufacturing as desktop processors. But you can't expect them to continue improving at this pace, particularly when you consider the very strong TDP limitations in mobile.

Eventually they are going to start following a 2 year cycle like desktop CPUs. Nvidias desktop graphics approach of building a bigger and hotter GPU every six months and gluing a bigger vacuum cleaner to it isn't going to work on a phone.

I think what we will see is some distinct performance categories based on battery size and a devices ability to dissipate heat. 10", 7", 4", 3". Right now Apple is using a 3" SOC in a 10" tablet and there is a lot of room for adding more performance and power consumption there. But within each segment, I don't think we will see meaningful improvements more quickly than about once every 18-24 months.

 

[badb0y]

Next 4 years are going to be interesting...I am probably going to end up buying a Tegra based tablet.

 

[cbn]

First have a look at this video

http://www.youtube.com/watch?v=VATNyBTPltI&feature=related

They test Kal-El with coremark benchmark.

From www.coremark.org

http://www.coremark.org/faq/index.php?pg=faq

what do you guys think ??

It looks like Coremark does scale perfectly with multiple cores. (assuming each Tegra 3 cpu core is the same as a Tegra 2 cpu core)

So according to the results (found in the video) each Cortex A9 cpu core gets roughly half the coremarks of a 2 Ghz Core2duo T7200 cpu core.

Hmmm.....A 1 Ghz Cortex A9 cpu core actually has greater than 50% of the coremarks of a 2Ghz Core 2 duo core? Wouldn't the Core 2 duo have much better IPC? (Something tells me this benchmark is not optimized for Intel )

 

[Mopetar]

Eh, it's a T7200, a 65nm Merom part from 2006 with a clock rate of 2 GHz. From what I've seen, slides that discussed the Tegra 3 (Which Kal-El most likely is) said the cores would run at 1.5 GHz.

It's also only one benchmark. Unless you have an entire set of them it's possible that Nvidia picked the best looking one they could find.

 

[IntelUser2000]

Hmmm.....A 1 Ghz Cortex A9 cpu core actually has greater than 50% of the coremarks of a 2Ghz Core 2 duo core? Wouldn't the Core 2 duo have much better IPC? (Something tells me this benchmark is not optimized for Intel )

The Nvidia chips are using newer and more aggressively optimized versions than the Core 2 one. There's a 1.6GHz Core 2 that scores the same as the 2.0GHz version shown above.

It's also worth pointing out, that from Anand's CPU bench, the 2 core Core 2 clocked at 1.6GHz is more than 3x faster versus the 230, while from CoreMark, the differences are only 2x.

 

[Dribble]

I don't think the likes of TI, Samsung, Qualcomm are just going to roll over and sit idly by while Nvidia "wins big".

Qualcomm are a relatively small company - they are not used to going up against the really big boys and are unlikely to win long term. They bought some AMD IP and have successfully used that but does that mean they can keep up when the market gets really cut throat? If this market really takes off chances are someone big will buy them out.

Most of the others develop these ARM soc's but not as their primary business. e.g. Samsung doesn't make it's money from ARM soc's - they sell the end hardware. This is proven by the fact that despite having their own soc they quite happily use tegra where it works better. The galaxy tablet 2 is tegra2, and even their new galaxy s 2 phone is going to be sold with either a hummingbird or tegra2 chip. i.e. same phone, different but similarly performing soc. Samsung care a lot more about the end phone then they do about the chip in it.

This is unlike nvidia who are all about the computing hardware, that is their one and only focus - and they are used to having to fight tooth and nail against Intel/AMD who have successfully crushed most opposition.

 

[BD231]

Honestly I first laughed when I found out about Nvidia making ARM based processors, but I'm starting to think they may have something here.

Then I guess you were laughing when they started making chipsets and integrated video device's

They've always seemed like the latch on and suck from anothers TEET kinna company when it comes to new business ventures. Intel went Mike Tyson and choked them till they were blue in the face.

Wondering if perhaps the mobile market will do the same.

 

[Idontcare]

Qualcomm are a relatively small company - they are not used to going up against the really big boys and are unlikely to win long term.

http://www.icinsights.com/news/bull...-Forecast-To-Top-10-Billion-In-Sales-In-2010/

Qualcomm is small?

 

[Sinanju]

Ironman confirmed to be faster than superman!!

 

[Davendork]

I'm excited also by the potential of RISC back on the desktop with Windows 8 ARM support and all the competition / brains going at it in the ARM market thanks to smartphones.

I'm, however, really confused by the naming convention of the types of ARM cores. Some cores are A9 and some are A15. It'd be awesome if there was a breakdown of the differences in the core architectures at a high level. Like, A15 is a SoC with graphics, networking, max speed, max cores, etc.

 

[Dribble]

http://www.icinsights.com/news/bull...-Forecast-To-Top-10-Billion-In-Sales-In-2010/

Qualcomm is small?

lol I admit you might have a point there...still be interesting to come back to this thread in 3 or 4 years and see how it has all panned out.

 

[nismotigerwvu]

I'm excited also by the potential of RISC back on the desktop with Windows 8 ARM support and all the competition / brains going at it in the ARM market thanks to smartphones.

I'm, however, really confused by the naming convention of the types of ARM cores. Some cores are A9 and some are A15. It'd be awesome if there was a breakdown of the differences in the core architectures at a high level. Like, A15 is a SoC with graphics, networking, max speed, max cores, etc.

Well to be honest, its the A15 that really threw a wrench in the works. There had been nice generational steps from A8 to A9 and then *BAM* A15. Things get a lot muddier when companies like Qualcomm don't just stick to "reference designs" (not meaning this in a bad way, it just makes concise lines difficult to paint). Take the Snapdragon for instance, which is a highly modified take on the A8 design that has more floating point grunt than anything else we've seen so far (but is hobbled by the mostly horrible, but rapidly improving Adreno gpus). To make a long story short, you have to look at benchmarks man...

 

[LoneNinja]

Then I guess you were laughing when they started making chipsets and integrated video device's

They've always seemed like the latch on and suck from anothers TEET kinna company when it comes to new business ventures. Intel went Mike Tyson and choked them till they were blue in the face.

Wondering if perhaps the mobile market will do the same.

I didn't actually pay attention to what was inside a computer until around 2005, so Nvidia already had IGP and chipsets. lol

I also knew very little of ARM, or how many devices they're actually inside when first hearing about Nvidias venture into the ARM. There was also the rumor of Win 8 running on ARM and Nvidia wanted to create APU's to compete with AMD/Intel and X86. I would still be laughing if their main target was desktops and Win 8.

I also think Tablets are a fad, and once the fad ends, they'll be nothing more than a niche market. My Android phone is great, so is the mobility and it's great having Internet wherever I go, but outside of calls/texts I don't touch the thing when I'm home. My desktop, even for basic internet browsing, is superior for me. Than of course productivity is vastly superior due to far greater performance. A tablet to me removes the portability that a phone has, while also removing the ease of use and productivity of a desktop or even laptop.

Don't get me wrong, I'd love to see more powerful phones, but it's my honest opinion tablets will loose popularity quickly. I could very well be wrong, guess I'll have to watch the market and find out as time goes by.

 

[cbn]

I'm, however, really confused by the naming convention of the types of ARM cores. Some cores are A9 and some are A15. It'd be awesome if there was a breakdown of the differences in the core architectures at a high level. Like, A15 is a SoC with graphics, networking, max speed, max cores, etc.

I really like these charts/graphs. (particularly the second one)