Tonight at 8pm PT / 11pm ET NVIDIA will be hosting its annual press conference from CES 2015. While I fully expect the majority of the information to be on Tegra-related products, there is a chance they could slip in some GeForce products as well. I leave nothing past NVIDIA with a 2 hour window.
You can watch the live stream of the event right here and I am going to be live blogging from the event as well, pending Internet capability inside the conference room.
PC Perspective's CES 2015 coverage is sponsored by Logitech.
Follow all of our coverage of the show at https://pcper.com/ces!
Tegra? more we will be
Tegra? more we will be providing tegra chips for cars and there Infotainment systems. Yay, faster mp3 and satellite radio. 1000fps gps navigation screens.
Wonder if they announce how no-one took them up on there IP offer or how they are suing companies because of it to try and make money. They also would have to mention how the Samsung lawsuit against them was also allowed to proceed.
Samsung’s ITC Case Against Nvidia Moves Forward
http://www.pcmag.com/article2/0,2817,2474163,00.asp
Just to point out, the
Just to point out, the purpose of using powerful, energy efficient chips in a car is not for GPS or radio, it is for the future of driving. Cars are becoming much smarter with sensors, being able to identify objects, speed limits, displaying information on a windshield. Imagine a guided arrow similar to what you would see in Forza or other racing games with green, yellow, and red to show you when/where you need to turn and if you are going to fast.
Now, imagine if you had sensors that could determine driving conditions. What if its a bit wet, snowing, or the roads are icy. What if the car could then guide you to to drive faster or slower to get around a bend in those conditions based on the bank of the road?
These are just a couple of really really easy examples that I thought of just now. You better believe that calculating that kind of stuff would require a decent bit of power in a chip. I’m not saying that is happening today, or tomorrow, but it is coming and nVidia is working on hardware to be able to do that kind of workload.
Facts are awesome
The Tesla
Facts are awesome
Audi is still in development
Audi is still in development and its a hybrid system.
What’s up with the focus on
What’s up with the focus on Autos, what about the 8 64 bit CPU cores, hopefully they tweaked the Denver cores a little more, but the K1’s Denver core could execute 7+ IPC per core, so the X1’s cores should at least be good for 7+ IPC times 8 cores, that’s a lot of CPU power, maybe for a Chromebook, Tablet, device. Let’s see how this benchmarks against Broadwell M, and Intel’s not so great graphics. If only there could be some tablet SKUs running full Linux. The Tegra X1, running a chromebook, would probably give a Chromebook running a core i3, or i5, some stiff competition, and that GPU will make mincemeat out of Intel’s tablet SKUs weak graphics. Nvidia should begin to transition the Tegra, with more GPU power(32 bit FP units) for the laptop market, and even more than 8 CPU cores.
Why even bother with a
Why even bother with a chromebook? Wouldn’t a full Windows laptop be more useful with this hardware inside?
Not with Arm Holdings
Not with Arm Holdings reference design CPU cores, less than half of the Denver core’s IPC, It looks like Apple will continue to lead in the custom ARMv8 ISA based SOC market in 2015, at least until AMD gets their custom ARMv8 ISA based APUs to market around 2016. 8 Denver cores would have been a very powerful SOC, but the ARM reference cores are not as wide order superscalar as the Apple, and Denver custom cores. Imagination Technologies, who make the PowerVR GPU cores that Apple, and others use in their mobile SOCs, purchased MIPS a few years back, and has since produced a 64 bit MIPS ISA based CPU, so expect there to be more completion among the RISC based SOCs, and those 64 bit MIPS processors will be paired with PowerVR graphics, and there are options for the IT’s processors to have SMT( MIPS I6400 CPU includes up to four threads per core ).
For Graphics workloads, unless the GPU can handle FP 32, at the very least, any compositing tasks will be slower, so not in a laptop, at least not the AUTO grade GPU that only has FP 16. Hopefully Apple will get a pro tablet, that can run OSX, and get the PowerVR wizard, with the hardware ray tracing, those Enhanced Cyclone cores in the A8X should be fine for the task, and 8 of those cores could power a laptop, and the PowerVR slightly customized GPUs in the A8X certainly have better graphics than Intel’s. Nvidia is not going to be alone, in producing a SOC for third party OEMs to place in their products, once AMDs k12 custom ARMv8 ISA based APUs reach the marketplace, and Nvidia will need its Denver cores to remain competitive.
Why is the press not getting
Why is the press not getting more Info of the K1’s CPU cores. I’m reading people say that they are using ARM reference cores(Please let this be wrong), One Denver 64 bit core can out perform 2 ARM Reference 64 bit cores, who wants the ARM cores, when the Denver cores are so much more powerful.
Oh Well the Nvidia website
Oh Well the Nvidia website lists Arm A57-A53 cores, 4 Denver cores could outperform these 8 reference cores! What’s up with the Denver cores, Nvidia, well here is waiting for AMD’s custom 64 bit ARM cores, for its custom ARMv8 APU product, come next year.