Intel discloses newest microarchitecture and 14 nm manufacturing

Intel logoIntel today disclosed details of its newest microarchitecture that is optimized with its 14nm manufacturing process.

Intel disclosed details of the microarchitecture of the Intel Core M processor, the first product to be manufactured using 14nm. The company says that the new microarchitecture along with the advanced manufacturing process will contribute to smaller, cooler and more silent devices.

Intel architects and chip designers have achieved greater than two times reduction in the thermal design point when compared to a previous generation of processor while providing similar performance and improved battery life.

Intel's first 14nm technology in volume production uses second-generation Tri-gate (FinFET) transistors.It will be used to manufacture a wide range of high-performance to low-power products including servers, personal computing devices and Internet of Things.

The first systems based on the Intel Core M processor will be on shelves for the holiday selling season followed by broader OEM availability in the first half of 2015.

Intel discloses newest microarchitecture and 14 nm manufacturing

Intel wants a greater foothold in the mobile market. And with Broadwell-Y they believe they finally have what they need to accomplish that goal.

Core M

Let's start with some information about Broadwell's GPU. Generally, Broadwells GPU is a continuation of the Intel Gen7 architecture first seen in Ivy Bridge. While there are some changes, this is still the same GPU architecture that weve seen from Intel for the last two generations.

Broadwells GPU has been upgraded to support the latest graphics APIs -- Direct3D feature level 11_2 and Direct3D 12 -- meaning Intel will not longer lag behind Nvidia's and AMD's offerings.

Intel Core M

Intel has confirmed that Broadwells GPU will offer support for OpenCL 2.0, including OpenCLs shared virtual memory.

Broadwell-Y, like Haswell-Y before it, implements a single slice configuration of Intels GPU architecture. However the composition of a slice will be changing for Broadwell, and this will have an impact on the balance between various execution units.

Intel isnt talking about overall GPU performance, at least for now. The company is reiterating the benefits of their 14nm process, noting that because 14nm significantly reduces GPU power consumption it will allow for more thermal headroom, which should further improve both burst and sustained GPU performance in TDP-limited scenarios relative to Haswell.

Broadwells GPU also brings the 'Duty Cycle Control'. The idea is to extract more idle power gains. Transistors require a minimum voltage to operate, which means that after a certain point Intel can no longer scale down their voltage and reduce the poewr consumption further. Intels solution to this problem is is to start turning off the GPU instead -- the process of duty cycling. By putting the GPU on a duty cycle Intel can run the GPU for just a fraction of the time down to 12.5% of the time which gets around the threshold voltage issue entirely.Control of the duty cycle is then handled through a combination of the GPU hardware and Intels graphics drivers.

Duty cycle control is especially important for the TDP and battery life constrained Y SKU, but ultimately all mobile SKUs would stand to benefit from this feature.

Intel is also upgrading their GPUs media capabilities for Broadwell. The increase in sub-slices and the resulting increase in samplers will have a direct impact on the GPUs video processing capabilities the Video Quality Engine and QuickSync further increasing the throughput of each of them, up to 2x in the case of the video engine. Intel is also promising quality improvements in QuickSync.

Broadwells video decode capabilities will also be increasing compared to Haswell. On top of Intels existing codec support, Broadwell will be implementing a hybrid H.265 decoder, although the shaders will be also used for decoding, making it less power efficient than doing everything in hardware.

Last but not least, Broadwells display controller will be updated and will support HDMI 1.4 and DP 1.2/eDP 1.3a. The Y SKU will also is get native support for 4K, although you should not expect Broadwell-Y to have the performance necessary to do intensive rendering at this resolution.

Regarding Broadwells CPU architecture, Intel is shooting for a better than 5% IPC improvement over Haswell.

In order to deliver that IPC increase Intel will be relying on architectural tweaks in Broadwell: Bigger schedulers and buffers for better feed of the CPU cores. Broadwells out-of-order scheduling window for example is being increased to allow for more instructions to be reordered, thereby improving IPC. Meanwhile the L2 translation lookaside buffer (TLB) is being increased from 1K to 1.5K entries.

The TLBs are also receiving some broader feature enhancements that should again improve performance. A second miss handler is being added for TLB pages, allowing Broadwell to utilize both handlers at once to walk memory pages in parallel.

Meanwhile, Broadwells branch predictor will see its address prediction improved for both branches and returns, allowing for more accurate speculation of impending branching operations.

In addition, both multiplication and division are receiving a performance boost thanks to performance improvements in their respective hardware. Floating point multiplication is seeing a sizable reduction in instruction latency from 5 cycles to 3 cycles, and division performance is being improved by the use of a Radix-1024 (10bit) divider. Even vector operations will see some improvements here.

Finally, Intel is once again targeting cryptography for further improvements.

Specifically for the Core M, Intel has made a number of alterations to everything from the CPU packaging to the process node itself.

Intel's biggest goal with Core M is to offer a processor for 10" tablets under with 10mm in thickness that are passively cooled - -a device that would require a sub-5W SoC. In addition, the Core M variant is designed to further reduce power consumption by optimizing the resulting transistors for lower power, lower voltage, lower clockspeed operation. By doing this Intel was able to further reduce power consumption in all of the major areas over what would be a traditional 14nm Intel process.

Source: CDRinfo

Tags: 14 nm, CPUs, Intel

Comments
Add comment

Your name:
Sign in with:
or
Your comment:


Enter code:

E-mail (not required)
E-mail will not be disclosed to the third party


Last news

 
The smartphone could come with a 4,200mAh battery
 
New iTunes backup password verification system exists in parallel to the much stronger, older iOS 9 system
 
 
ARM's new Bifrost architecture, which focuses on high-end 4K and VR experiences
 
Apples 2017 iPhone could employ the use of Samsungs Flexible OLED Atomic Layer Deposition Technology
 
Redmond not willing to pay so much for Twitter
 
Foxconn and Pegatron will both be behind the production
 
Samsung Galaxy TabPro S - a tablet with the Windows-keyboard
The first Windows-tablet with the 12-inch display Super AMOLED
June 7, 2016 /
Keyboards for iOS
Ten iOS keyboards review
July 18, 2015 /
Samsung E1200 Mobile Phone Review
A cheap phone with a good screen
March 8, 2015 / 4
Creative Sound Blaster Z sound card review
Good sound for those who are not satisfied with the onboard solution
September 25, 2014 / 2
Samsung Galaxy Gear: Smartwatch at High Price
The first smartwatch from Samsung - almost a smartphone with a small body
December 19, 2013 /
HP Slate 7 is a 7-inch Android 4 Tablet PC with good sound
A cost-effective, 7-inch tablet PC from a renowned manufacturer
October 25, 2013 / 4
 
 

News Archive

 
 
SuMoTuWeThFrSa
    123
45678910
11121314151617
18192021222324
252627282930 




Poll

Do you like Windows 10?
or leave your own version in comments (32)