More details of Intel’s plans for improving graphics onboard its processors have been revealed, the Xe LP graphics for Tiger Lake processors will apparently be comparable to AMD’s GPU technology in terms of performance, although ‘It remains to be seen whether Apple will use it. before switching completely to Apple Silicon.
While previously a look at Intel’s future mobile processors would include a discussion of what might be included in future MacBook refreshes, that may not be the case for the Xe LP and Tiger Lake. Intel has suggested that the first Tiger Lake processors could ship before the end of 2020, but it seems unlikely that they will even be used by Apple.
During WWDC 2020, Apple announced that it was moving away from Mac and macOS from Intel processors to its own Apple Silicon, chips designed and produced by Apple itself. Apple intends to launch its first Apple Silicon products before the end of 2020 and hopes to transform its entire line of Macs for use within two years.
This means that for now, Apple will continue to use Intel chips in its products, so these new details may still be relevant. However, the fact that Apple Silicon is on the horizon and the gradual transition to it is less of an incentive for Apple to make major updates to its products using Intel processors. Consumers are also unlikely to want to purchase Intel Macs in the future if they are replaced by Apple Silicon substitutes within two years of purchase.
Apple can still refresh Macs with new processors, but given its own timeline and Intel’s penchant for delaying processor releases, it seems more likely that Apple will choose to use a current-gen processor rather than work. to incorporate another generation.
Could still appear on Macs
In January, before we knew about the arrival of Apple Silicon, Intel provided informationon about its Tiger Lake processors and the integrated Xe architecture it would use, an architecture that could provide a massive boost to graphics on mobile devices. During its recent Intel Architecture Day, Intel offered even more plans for the future direction of its GPU architecture.
Xe LP is a low power version of the Xe architecture for laptops and mobile devices rather than the Xe HP and Xe HPC variants for data centers and servers and the Xe HPG as potential discrete graphics cards for game lovers. As an entry-level solution, it will be used on processors intended for low power systems.
The chipmaker plans to run Xe LP with thermal design point designs ranging from 10 watts to 30 watts, shared with the processor, Tom’s Hardware reports. To achieve this with low power requirements, Intel has strived not only to improve its floating point per second (FLOPS) operations, but also to improve performance per FLOP and reduce the power required.
Intel claims that Xe LP will perform better in terms of performance than its previous Gen 11 graphics, with the ability to run at higher frequencies and better performance at the same voltage. Alternatively, it can also deliver performance comparable to Gen11 when using lower amounts of power.
The processor improvements start with Intel integrating 50% more Execution Units (UEs) into the GPU than the Gen11 versions. When combined with higher clock speeds, it is believed that Xe LP will be twice as fast as the previous generation.
An overview of Intel’s Xe LP architecture.
While Gen11 used two arithmetic logic units (ALUs) which were 4 wide and fractional work, so one was used for floating point and integer tasks while the other used enhanced math, Xe a at the places each running UE on a vector unit of 8 wide. floating point and whole number calculations. A separate enhanced math unit is executed simultaneously with the integer pipeline, with the two UEs also sharing a single controller.
Switching to provide floating point and integer work on eight ALU pipelines will effectively double performance right off the bat. There is also additional support for Int16 and 4x Int8 computations, along with an L1 cache to reduce latency and demands on the L3 cache, the latter having also doubled to 16MB.
There is also additional support for end-to-end data compression, which is said to be better than the delta color compression used by AMD and Nvidia. This support can further reduce bandwidth and power requirements, making it more suitable for high performance tasks.
Other updates include support for up to four displays, support for DisplayPort 1.4b and HDMI 2.0, an enhanced media engine with support for four 4K60 displays, full AV1 decoding acceleration, and the ability to simultaneously decode seven 4K60 HEVC streams.
We have no possible way of comparing Intel’s claims about future performance with Apple’s much less specific claims about the same for Apple Silicon. Publicly, Apple has said it wants to create Macs with significantly faster and constantly improving performance, so we can expect the target to be at least equal to or better than what Intel promises.
There’s also, however, the fact that Intel is significantly behind on its own projections of where its processors are today. Apple may have switched to its own processors because it doesn’t expect the situation to improve.
This would also be due to the fact that Intel’s quality assurance is poor. Apple Silicon will have to keep its promise of better performance, but it is at least likely to meet Apple’s timelines and quality control.