By: Peter Greenhalgh (no.delete@this.thanks.com), February 4, 2015 11:56 am
Room: Moderated Discussions
Hi David,
Hope you are doing well.
I'm going to avoid being drawn in to the discussion here, though it's great to read the speculation! :)
However, I will shed some more light on some of the questions & comments in this thread with information we announced yesterday in San Francisco.
Firstly, there are micro-architectural enhancements throughout the Cortex-A72 design which improve both IPC and power. In fact, the Cortex-A72 power improvements are achieved on the same process with the same library as Cortex-A57. We aren't relying on a process shrink to achieve the power improvement or boost performance purely through frequency. Depending on the workload we're seeing anywhere between 10-50% more clock-for-clock performance than Cortex-A57 under identical system conditions while also reducing power. I'm talking about a range of decent sized, representative workloads, not micro-benchmarks. More information on the micro-architecture and said workloads will be disclosed in due course! :)
With respect to the performance uplift numbers over Cortex-A15 and Cortex-A57 that we quoted at the launch. These do factor in process technology. Cortex-A15 is predominantly a 28nm product, Cortex-A57 is mainly on 20nm and Cortex-A72 we expect to see on 14nm/16nm. While the audience here on RWT is very technical and understands the nuances of process geometry, many of the press we speak to want an appreciation of how much faster the processor in their favourite phone, tablet, may be in the future. To that end we quote performance uplift as a combination of the IPC improvement and the extra frequency afforded to our silicon partners due to advances in process technology.
Overall we're extremely pleased to have increased IPC while meaningfully decreasing power across various use cases. It's going to be great to see Cortex-A72 in partner platforms. As we've got more than 10 partners extremely active with the processor I hope the wait won't be too long! [in a big.LITTLE configuration with Cortex-A53 of course ;)]
I'm going back to lurk-mode now.
Best Regards,
Peter Greenhalgh (Director of Technology, CPU Group, ARM)
Hope you are doing well.
I'm going to avoid being drawn in to the discussion here, though it's great to read the speculation! :)
However, I will shed some more light on some of the questions & comments in this thread with information we announced yesterday in San Francisco.
Firstly, there are micro-architectural enhancements throughout the Cortex-A72 design which improve both IPC and power. In fact, the Cortex-A72 power improvements are achieved on the same process with the same library as Cortex-A57. We aren't relying on a process shrink to achieve the power improvement or boost performance purely through frequency. Depending on the workload we're seeing anywhere between 10-50% more clock-for-clock performance than Cortex-A57 under identical system conditions while also reducing power. I'm talking about a range of decent sized, representative workloads, not micro-benchmarks. More information on the micro-architecture and said workloads will be disclosed in due course! :)
With respect to the performance uplift numbers over Cortex-A15 and Cortex-A57 that we quoted at the launch. These do factor in process technology. Cortex-A15 is predominantly a 28nm product, Cortex-A57 is mainly on 20nm and Cortex-A72 we expect to see on 14nm/16nm. While the audience here on RWT is very technical and understands the nuances of process geometry, many of the press we speak to want an appreciation of how much faster the processor in their favourite phone, tablet, may be in the future. To that end we quote performance uplift as a combination of the IPC improvement and the extra frequency afforded to our silicon partners due to advances in process technology.
Overall we're extremely pleased to have increased IPC while meaningfully decreasing power across various use cases. It's going to be great to see Cortex-A72 in partner platforms. As we've got more than 10 partners extremely active with the processor I hope the wait won't be too long! [in a big.LITTLE configuration with Cortex-A53 of course ;)]
I'm going back to lurk-mode now.
Best Regards,
Peter Greenhalgh (Director of Technology, CPU Group, ARM)
Topic | Posted By | Date |
---|---|---|
ARM announces A72 | Maynard Handley | 2015/02/03 12:36 PM |
ARM announces A72 | anon | 2015/02/03 01:53 PM |
ARM announces A72 | Hugo Décharnes | 2015/02/03 02:20 PM |
ARM announces A72 | juanrga | 2015/02/03 05:15 PM |
ARM announces A72 | Wilco | 2015/02/04 01:58 AM |
ARM announces A72 | Eric Bron | 2015/02/04 02:48 AM |
ARM announces A72 | none | 2015/02/04 03:24 AM |
ARM announces A72 | Eric Bron | 2015/02/04 03:42 AM |
ARM announces A72 | Exophase | 2015/02/04 08:01 AM |
ARM announces A72 | Anon | 2015/02/04 08:35 AM |
ARM announces A72 | Exophase | 2015/02/04 08:58 AM |
ARM announces A72 | Groo | 2015/02/04 10:24 AM |
ARM Marketing, BS up to my ears | David Kanter | 2015/02/04 11:51 AM |
ARM Marketing, BS up to my ears | Maynard Handley | 2015/02/04 02:59 PM |
ARM Marketing, BS up to my ears | David Kanter | 2015/02/04 03:21 PM |
ARM Marketing, BS up to my ears | Groo | 2015/02/04 03:30 PM |
ARM announces A72 | juanrga | 2015/02/04 05:23 AM |
ARM announces A72 | Wilco | 2015/02/04 04:01 PM |
ARM announces A72 | juanrga | 2015/02/04 05:06 PM |
ARM announces A72 | Anon | 2015/02/04 02:28 AM |
ARM announces A72 | juanrga | 2015/02/04 05:31 AM |
ARM announces A72 | Aaron Spink | 2015/02/04 07:49 AM |
ARM announces A72 | Ronald Maas | 2015/02/03 08:23 PM |
ARM announces A72 | Seni | 2015/02/04 01:19 AM |
ARM announces A72 | Maynard Handley | 2015/02/04 11:42 AM |
ARM announces A72 | Seni | 2015/02/04 01:33 PM |
ARM announces A72 | dmcq | 2015/02/04 01:57 PM |
ARM announces A72 | Ronald Maas | 2015/02/04 07:42 PM |
ARM announces A72 | anon | 2015/02/04 06:19 AM |
ARM announces A72 | Exophase | 2015/02/04 08:31 AM |
ARM announces A72 | David Kanter | 2015/02/04 11:25 AM |
ARM announces A72 | Exophase | 2015/02/04 02:33 PM |
ARM announces A72 | anon | 2015/02/04 11:27 PM |
ARM announces A72 (fixed format) | anon | 2015/02/04 11:29 PM |
ARM announces A72 | Exophase | 2015/02/05 12:11 AM |
ARM announces A72 | anon | 2015/02/05 01:02 AM |
ARM announces A72 | anon | 2015/02/04 06:57 PM |
ARM announces A72 | Wilco | 2015/02/03 02:39 PM |
ARM announces A72 | Maynard Handley | 2015/02/03 03:13 PM |
ARM announces A72 | anon | 2015/02/03 03:29 PM |
ARM announces A72 | Wilco | 2015/02/03 03:44 PM |
ARM announces A72 | David Kanter | 2015/02/04 10:56 AM |
ARM announces A72 | Peter Greenhalgh | 2015/02/04 11:56 AM |
ARM announces A72 | Aaron Spink | 2015/02/04 12:59 PM |
ARM announces A72 | Alberto | 2015/02/07 11:22 AM |
ARM announces A72 | Exophase | 2015/02/07 11:47 AM |
ARM announces A72 | Alberto | 2015/02/07 01:44 PM |
ARM announces A72 | Exophase | 2015/02/07 03:35 PM |
ARM announces A72 | Alberto | 2015/02/08 02:09 AM |
ARM announces A72 | Exophase | 2015/02/08 12:05 PM |
ARM announces A72 | David Kanter | 2015/02/08 01:39 AM |
ARM announces A72 | dmcq | 2015/02/08 05:14 AM |
ARM announces A72 | Michael S | 2015/02/08 05:38 AM |
ARM announces A72 | Gabriele Svelto | 2015/02/10 06:11 AM |
ARM announces A72 | Jouni Osmala | 2015/02/10 12:24 PM |
slit vs unified | Michael S | 2015/02/10 02:57 PM |
slit vs unified | dmcq | 2015/02/11 06:44 AM |
ARM announces A72 | Doug S | 2015/02/08 10:00 AM |
ARM announces A72 | Exophase | 2015/02/08 11:57 AM |
ARM announces A72 | dmcq | 2015/02/04 02:10 PM |
ARM announces A72 | David Kanter | 2015/02/04 03:28 PM |
ARM announces A72 | Wilco | 2015/02/04 02:59 PM |
ARM announces A72 | Aaron Spink | 2015/02/04 10:31 PM |
Intel 32nm vs 14 nm | Michael S | 2015/02/05 02:03 AM |
Intel 32nm vs 14 nm | Wilco | 2015/02/05 03:27 AM |
Intel 32nm vs 14 nm | David Kanter | 2015/02/05 10:05 AM |
Intel 32nm vs 14 nm | carop | 2015/02/05 12:12 PM |
Normalize to drawn or effective width? | David Kanter | 2015/02/05 12:45 PM |
Normalize to drawn or effective width? | carop | 2015/02/05 03:40 PM |
Normalize to drawn or effective width? | David Kanter | 2015/02/06 01:44 PM |