Three weeks to Intel’s (INTC) – Get report stock plump on news of a resurgence of production technology, the company shares many details on how it intends to compete with AMD (AMD) – Get report and Nvidia (NVDA) – Get report over the next few years.
At this year’s edition of its annual Architecture Day event, Intel revealed fresh details about its soon-to-be-launched Tiger Lake notebook processor line, as well as its next-gen Alder Lake PC CPU platform and its plans to combat AMD and Nvidia in a host of GPU markets.
And remarkably – after last month’s revelation that the launch scheme for CPUs on its next gen, 7nm, the production process node is being pushed out by about six months due to manufacturing issues – Intel promises significant transistor performance for chips made by today’s Thanks to 10nm node, thanks to a set of improvements of transistor and capacitor technology it refers to as SuperFin.
Intel claims that SuperFin, which will be used by Tiger Lake processors and Intel’s first discrete GPUs, will provide 17% to 18% better transistor performance than Intel’s first 10nm transistors.
The company also said it is working on “Enhanced SuperFin” transistors; they deliver extra performance gains and will be used by both future GPUs and Intel’s Sapphire Rapids server CPU platform, which arrives in the second half of 2021. Another version of SuperFin will be used by Intel’s Ice Lake server CPU. line, which is set to ship at the end of the year and will compete against the expected AMD’s third-gen Epyc server CPU line (Milan).
Tech analyst Patrick Moorhead noted that Intel’s recent advances seem to give it a lead in the industry in terms of transistor performance. Taiwan Semiconductor (TSM) – Get report, which recently began mass production for its 5nm node (seen as competitive with Intel’s planned 7nm node), currently has a clear advantage in terms of transistor density.
Tiger Lake and Alder Lake
Thanks in part to ‘dramatic’ clock speed improvements, made possible by SuperFin, Intel promises to unveil its 10nm Tiger Lake line, which will be unveiled at a Sept. 2-event and reliance on a new CPU core microarchitecture, known as Willow Cove, will deliver “more than a generation increase in CPU performance.” The company also reiterates that Tiger Lake’s integrated GPUs – based on a new architecture known as Xe-LP – a 2x performance improvement will deliver relative to Intel’s older Gen11 integrated GPUs, and says Tiger Lake will deliver “massive” performance gains for processing AI / deep learning algorithms.
If Intel makes good on its performance / clock speed requirements, Tiger Lake should put Intel on a stronger footing in the notebook processor market, with AMD seeing strong growth following the launch of the Ryzen Mobile 4000 line in January. Last month, AMD said that revenue for notebook processor more than doubled in Q2 annually.
As for the 10nm Alder Lake line, which is due to launch in the second half of 2021 and is intended for desktops and notebooks, Intel said the CPUs (like many ARM architecture CPUs) will have a mix of high performance cores based on a next-gen microarchitecture known as Golden Cove and low power cores based on a microarchitecture known as Gracemont.
Such architecture could improve the life of notebook batteries. However, there is some skepticism about the merits of its use in the high-end desktop CPU market, where AMD has become quite competitive and performance matters most more than power consumption. Until Alder Lake arrives, Intel will continue to rely on its older 14nm process node for desktop CPUs – the company is thought to be working on a new 14nm desktop CPU line, known as Rocket Lake.
Separately, Intel took the wraps off of Client 2.0, a long-term initiative to develop PC processors whose various processing and I / O functions are placed on several different chips and integrated using Intel’s packaging and interconnect technologies. The company claims that Client 2.0 can reduce percentage development times by a year as such relative to making a traditional, monolithic processor, while Intel can also produce processes to mix and match and reuse chip IP.
Intel’s GPU plans
For the discrete GPU market, Intel promises to launch several offerings between now and the end of 2021. Specifically:
- Xe-HP and Xe-HPC, a few high-performance server GPUs next year. With the exception of one Xe-HP variant, the GPUs will rely on multiple chip “tiles” connected via high-speed interconnects. Xe-HPC, named Ponte Vecchio and intended as Intel’s flagship server GPU, will rely on a mix of Intel and third-party manufacturing processes.
- Xe-HPG, a gaming GPU architecture that will support real-time ray tracing and is promised to meet the needs of “enthusiastic gamers.” It will be produced by a third party (possibly TSMC) and arrive sometime next year.
- DG1, a discrete PC GPU for developers that first appeared in January and is promised to ship at the end of the year.
- SG1, a server GPU that combines four DG1 GPUs and is expected to ship this year. Intel says SG1 is intended to support “low-latency, high-density Android cloud gaming and video streaming.”
The GPU’s announcements come three months after Nvidia, which remains dominant in the server GPU market, unveiled a versatile flagship server GPU, known as the A100, which relies on its new Ampere architecture. And it comes just ahead of a Sept. event where Nvidia is expected to unveil its first Ampere gaming GPUs.
AMD meanwhile is offering gaming GPUs based on a new architecture known as RDNA 2, as well as a server GPU based on a new architecture known as CDNA. Both are due at the end of the year.
Intel’s stock fell 1.5% in trading on Thursday, amid broader weakness in hardware exposures to Cisco Systems (CSCO) – Get report soft October quarter guidance.
.
