Carmack: Nvidia's Project Denver Will Be Good
John Carmack seems to like Project Denver.
As previously reported last week during CES 2011, Nvidia excited all of us with news that the company would be using an ARM design to create a CPU/GPU hybrid dubbed as Project Denver. Currently Nvidia already has hands-on experience using ARM-based designs for its Tegra chips, however this new processor will "usher in a new era for computing" for PCs, servers, supercomputers and beyond.
"ARM is already the standard architecture for mobile devices," said Nvidia's chief scientist Bill Dally. "Project Denver extends the range of ARM systems upward to PCs, data center servers, and supercomputers. ARM’s modern architecture, open business model, and vibrant eco-system have led to its pervasiveness in cell phones, tablets, and other embedded devices. Denver is the catalyst that will enable these same factors to propel ARM to become pervasive in higher-end systems."
id Software mastermind John Carmack seems equally excited about Nvidia's Project Denver, enough so to show his support via Twitter. "I have quite a bit of confidence that Nvidia will be able to make a good ARM core," he said. "Probably fun for their engineers."
Given that Nvidia is taking such a huge leap forward in creating its CPU/GPU processor, it's probably good to know that the highly respected designer behind the DooM, Quake and Wolfenstein series has their support. It will be interesting to see what Carmack and id Software does in the future in regards to support for the new Project Denver hardware.
In other ARM-related news, sources at Taiwanese laptops makers claim that AMD is currently working on GPUs to support the ARM architecture. The drawback is that AMD's pace is slower compared to the success of Nvidia's current Tegra 2 hotness. "AMD will be able to change the ecosystem and score points in the global notebook market if it moves faster," sources said.
- john-carmack ,
- tegra ,
- project-denver ,
- ARM ,
- id-software
- Nvidia May Be Reviving its Old ''Ti'' Marker
- FEAR 3 Now Pushed Back to May
- Nintendo 3DS Gets Region Codes
- Next-Gen SandForce Controller Seen on OCZ SSD
- Watch Us Shoot at an SSD With a Shotgun & Rifle
- NZXT Unveils Bunker USB Locking Device
- Thermaltake Reveals Level 10 GT Chassis
- IBM Wins the 2010 Patent Shootout
- Cataclysm Sells 4.7M Units in One Month
- AMD CEO Left Over Mobile Strategy Disagreement
- Crysis 2's 3D is ''the Same as Avatar''
- Intel CEO Talks About Competing with Arm
- Microsoft Battles Apple for App Store Trademark
- Gamers Raid Medical Server to Host Black Ops
- Intel Records Record Quarter and Record Year
- Intel Staffers Get Huge Bonuses after Record Year
- IBM's Supercomputer Beats Humans at Jeopardy
- Steve Jobs is Sick Again, Takes Another Leave
ARM on desktops? not gonna happen anytime soon, nvidia..
granted, the architecture has great potential, and may even have the capability to surpass x86 in terms of processing power, but - a change of architecture is no small thing. Even if only for the sake of compatability, x86 will reign for at least 8-10 more years
ARM on desktops? not gonna happen anytime soon, nvidia..granted, the architecture has great potential, and may even have the capability to surpass x86 in terms of processing power, but - a change of architecture is no small thing. Even if only for the sake of compatability, x86 will reign for at least 8-10 more years
True, especially given the Intel - Nvidia settlement confirms that Intel holds a patent to any x86 emulator. Without that, mass adoption will be postponed for years.
Which is of course sad, as ARM is an inherently more efficient and better scaling architecture.
That's not to say I'm a pessimist about desktop ARM's chances. It will start by securing its most important bridgehead: Linux distro's already support ARM based architectures and have a rather nice catalogue of software compatible with it.
Then it will foray into the netbook and ultraportable market. I can see Apple's MacBook Air being an early adopter for example: Apple's engineers have loads of ARM experience thanks to iOS and will no doubt want to jump on the ARM ship (battery life is far better on ARM than on x86 and that's Apple's main selling point).
Being a Linux derivative, Chrome OS will no doubt join the ARM party too, likely long before anyone else.
Given Windows will support ARM too in the nearby future (and most likely Office to go with it), quite a few Windows netbooks will change lanes as well. After all, browser makers also have near ARM ready code lying around (all major browser makers have a mobile ARM based version)
And then there's the big one: servers and HPC. ARM's efficiency and scalability are great in desktops, but even more so in the server world. If it can make even a small foothold there, commercial interest in ARM will soar.
At that point two things can happen:
- ARM breaks through and becomes a major architecture. In order to cash in on its growth, Intel uses its highly advanced manufacturing facilities and tons of R&D money to return to the ARM market. In order to push that source of revenue, it might even release an x86 emulator, perhaps subject to royalties.
- ARM fails to see mass adoption, but becomes a viable and financially healthy niche.
Either way, ARM will be rather successful.