A bit of Boombastic news from Asia, it is Digitimes who claimes that AMD will be delaying their ZEN2 announcements towards 2021, the reason, lack of competition.
Now, we doubt that AMD would have said that out loud, so the credits for 'lack of competition' as a statement go to the always reputable Digitimes, but yeah that's what they say could that mean no ZEN3 this year. We would expect to see announcements at CES 2021 in January now.
In Europe, AMD has 85% of the market, and the B550 and A520 motherboards have yet to be launched, making the current generation stronger than ever. AMD logically would continue selling current processors, which are manufactured using a process that is already super mature and optimized, and motherboard manufacturers continue to have a high level of sales and are releasing their current inventory. AMD has always claimed that these processors would arrive in late 2020, so we recommend taking this information with a bit of common sense and some salt.
The content below is translated and thus chinglish, but you'll get the point:
With the launch of the B550 chip on June 16th, motherboard makers will fully increase their capacity, followed by the A520, which will be released on July 7th, and the updated version of the 3 Ryzen 3000 series code-named Matisse Refresh that has not appeared on the previous platform blueprint will be processed. Devices, including Ryzen 9 3900 XT, Ryzen 7 3800 XT, and Ryzen 5 3600 XT, respectively, replacing the current 3900X, 3800X, and 3600X processors. In addition, on July 21, a low- and mid-level APU code-named Renoir followed.
Motherboard industry said that the Ryzen 3000 series is popular, and Supermicro has also extended its life cycle, and it is determined that the new generation plan will not be launched in September. It will use the TSMC 7nm EUV Zen 3 architecture Ryzen 4000 series processor, 2020 Mass production will only begin at the end of the year, announced at CES in January 2021, and it is currently unknown whether the process will be changed to 5 nm EUV.
This strategy also shows that the DT battle situation has been significantly reversed. Due to the mediocre performance of Intel Comet Lake, and the new products at the end of the year also stayed at the 14-nanometer process specifications, after the evaluation, it was decided to adjust the conversion schedule of the old and new platforms.
Motherboard industry further pointed out that Intel's DT platform strategy in recent years has also made the supply chain indistinct, facing the advancement of Super Micro Ryzen, and Comet Lake's performance is mediocre, but it has not seen Intel adjust its pace, let the 10nm DT processor advance Plan to market.
It is understood that after Comet Lake, Intel will launch the Rocket Lake processor platform as soon as the end of 2020, but it still stays in the 14nm process. The first 10nm Alder Lake processor on the DT platform will not be released until the end of 2021. It is as long as one and a half years, Intel still failed to fully enter the 10nm generation.
At present, it is observed that motherboard manufacturers such as ASUS, Gigabyte, MSI and ASRock, as well as HP, Dell, Acer and Lenovo have continued to increase the proportion of ultra-micro platform shipments and cases opened quarter by quarter. Near the end of the year, the market share of Super Micro DT is expected to approach 30%, and the DIY channel layout is expected to write a new high in more than 10 years.
Motherboard manufacturers said that due to the booming demand for ultramicro platforms, the profitability of high proportion of motherboard manufacturers has been quite good, which has also caused the market situation to form an ultra-micro momentum and continue to strengthen the atmosphere. It has also promoted the sales of ultra-micro NB and server platforms. If Intel Failure to amend the blueprint or start a price war in the second half of the year will lead to an increase in the loss of market share in 2021. Not only PCs, but also servers may be snatched by 10%. Performance growth momentum may be constrained.
In addition, it is worth noting that NVIDIA and Supermicro are also quite confidential about the launch schedule of the next-generation graphics chips. NVIDIA has released the Ampere (Amp) architecture A100 GPU developed for artificial intelligence and high-performance computing, and GeForce consumer GPUs have been released in August. Supermicro will not reveal Big Navi's annual new product until NVIDIA's new product is launched. It is currently expected to be extended to the fourth quarter.
With the successive launch of new products from Supermicro and NVIDIA, and the expansion of orders for TSMC's 7nm, 7nm EUV, and 5nm processes, the two major manufacturers have become key customers for TSMC to fill Huawei's capacity gap.