AMD unleashes more info about Torrenza
http://www.dailytech.com/article.aspx?newsid=6512
I think Torrenza could be the platform that gets Apple's interest.
Quote:
This week at CeBIT 2007, AMD revealed more details about its "accelerated computing" platform, codenamed Torrenza. AMD's goal behind Torrenza is to create a platform where application-specific processors can interact cost effectively and offer better performance than a general purpose CPU, while remaining compatible with off the shelf platforms.
AMD guidance revealed this week that future processors will also have integrated "accelerators" embedded into them. A Torrenza system will have at least two sockets, and both will accept accelerators and accelerated CPUs.
One accelerated-processor project on AMD plate, slated for 2008 under the codename Fusion, and combines a dedicated GPU or GPU accelerator onto the same package or even the same silicon die as the main CPU. AMD has already set the ground-work for Fusion processing with its Stream Computing initiative -- utilizing ATI-based graphics adaptors for heavy number crunching.
Other Torrenza ready projects are also coming to light. Clearspeed announced its CSX600 math-coprocessor plug-in last year, with the stated intention of creating a socket plugin version for Torrenza. Los Alamos National Labs is currently building the world's fastest supercomputer, Roadrunner, with Opteron and Cell processors on the Torrenza platform.
Torrenza is not just locked within the compounds of the CPU sockets. According to AMD, Torrenza systems will accept accelerators in a PCI-Express interface too, allow for multiple application specific accelerators to access system memory and processor functions directly. Mercury systems announced a PCIe plug-in accelerator late last year.
While Torrenza is well on its way to seeing daylight, Intel is also working on its own open architecture platform. Notorious for keeping its CPU platform a closely guarded technology, Intel indicated that it was working on a competitive technology to AMD's HyperTransport, dubbed CSI, allowing direct CPU and memory access.
Intel guidance suggests the company will announce its Torrenza competitor sometime in mid-2008.
This week at CeBIT 2007, AMD revealed more details about its "accelerated computing" platform, codenamed Torrenza. AMD's goal behind Torrenza is to create a platform where application-specific processors can interact cost effectively and offer better performance than a general purpose CPU, while remaining compatible with off the shelf platforms.
AMD guidance revealed this week that future processors will also have integrated "accelerators" embedded into them. A Torrenza system will have at least two sockets, and both will accept accelerators and accelerated CPUs.
One accelerated-processor project on AMD plate, slated for 2008 under the codename Fusion, and combines a dedicated GPU or GPU accelerator onto the same package or even the same silicon die as the main CPU. AMD has already set the ground-work for Fusion processing with its Stream Computing initiative -- utilizing ATI-based graphics adaptors for heavy number crunching.
Other Torrenza ready projects are also coming to light. Clearspeed announced its CSX600 math-coprocessor plug-in last year, with the stated intention of creating a socket plugin version for Torrenza. Los Alamos National Labs is currently building the world's fastest supercomputer, Roadrunner, with Opteron and Cell processors on the Torrenza platform.
Torrenza is not just locked within the compounds of the CPU sockets. According to AMD, Torrenza systems will accept accelerators in a PCI-Express interface too, allow for multiple application specific accelerators to access system memory and processor functions directly. Mercury systems announced a PCIe plug-in accelerator late last year.
While Torrenza is well on its way to seeing daylight, Intel is also working on its own open architecture platform. Notorious for keeping its CPU platform a closely guarded technology, Intel indicated that it was working on a competitive technology to AMD's HyperTransport, dubbed CSI, allowing direct CPU and memory access.
Intel guidance suggests the company will announce its Torrenza competitor sometime in mid-2008.
I think Torrenza could be the platform that gets Apple's interest.
Comments
Sebastian
http://www.dailytech.com/article.aspx?newsid=6512
I think Torrenza could be the platform that gets Apple's interest.
How would this work? Would Apple make a FCP machine? A PS accelerated machine?
How would this work? Would Apple make a FCP machine? A PS accelerated machine?
No I don't believe that's how it works. I haven't actually researched it much but I read a little bit about it a few months ago. The idea is there would be one CPU or APU as AMD calls it, that holds the CPU, GPU, and other specialized hardware. So there would be one unit handling everything and AMD can make APUs with say... a core for Artificial Intelligence, a core for sound, a core for physics, etc.
That article I read is probably out of date so I'll need to look into it further.
Sebastian
AMD is pusing Fusion( their GPU/CPU hybrid) along with the Torrrenza platform with Hypertransport connected links and their Stream technology (I'll have to look into this more) as a way to attack areas that are ill suited for general purpose CPU.
It's amazing how cyclical the computing market is. We go through eras of integration of technologies into a single piece of silicon only to return back to multi-processors. Parallel technologies give way to serial technologies and the inverse.
I'm figuring that Apple is well positioned to take advantage as they have Unix based OS and they actively promote Macs in BioTech and other High Performance computing areas.
if Mac Pro and xServe has option for AMD platform (forgot about consumer Macs for now)
somethings are NOT rite in this world
Sebastian has it right.
Not entirely. Whilst the information available would suggest that the "accelerators" would be hard-wired for one particular function, that wouldn't necessarily be the case. AMD are also working very closely with FPGA companies on this. Using FPGAs as the accelerators would allow for dynamically re-configurable hardware.
So Apple could implement come kind of "CoreComputation" framework that does the dynamic re-configuration. Much like CoreImage works out the best way to perform an image-processing function on the GPU, CoreComputation would work out the best way to perform a function on an FPGA accelerator, re-configure the FPGA accordingly, and then use it to perform the function. It should be noted that this is not at all trivial and would require a lot of development. I hope (and have done so ever since the switch to Intel) that Apple are working on it right now.
Not entirely. Whilst the information available would suggest that the "accelerators" would be hard-wired for one particular function, that wouldn't necessarily be the case. AMD are also working very closely with FPGA companies on this. Using FPGAs as the accelerators would allow for dynamically re-configurable hardware.
So Apple could implement come kind of "CoreComputation" framework that does the dynamic re-configuration. Much like CoreImage works out the best way to perform an image-processing function on the GPU, CoreComputation would work out the best way to perform a function on an FPGA accelerator, re-configure the FPGA accordingly, and then use it to perform the function. It should be noted that this is not at all trivial and would require a lot of development. I hope (and have done so ever since the switch to Intel) that Apple are working on it right now.
Very nice. CoreComputing is exactly what those in the Science, BioTech and Engineering verticals would love IMO. Apple just posted some ADC videos on iTunes centered around OS X and Science and I avoided them because I knew it would be hopelessly over my head
I'm still gobsmacked that Apple got ZFS going in OS X so quickly. The first thing I though was "Apple's never going to support something like ZFS so quickly" and the next I know fairly evolved ZFS support is in OS X. Apple has allayed my fears that they aren't looking forward to new technologies. In fact OS X is so portable by design I think they probably have to reign themselves back in from trying to support too many things.