Today Intel had a major Data Center event in San Francisco. It was a multi-hour announcement showcasing all the different products Intel is/will be launching.
Some interesting background that Intel talked about was that only 2% of worlds data has been analyzed and a 5G will be a major driver move compute to the edge. They also touted that over 50% of AI workloads are inference and runs best on Intel (X86). Noticeably absent at the beginning of the presentation was Intel’s work on training.
We found the most interesting parts of the announcement being that AWS talked about custom versions of Intel’s CPU and a Up to 14X inference improvement from just July 2017 in its XEON processors. Overall there is a 30% gen/gen improvement in XEON, the biggest jump in 5 years. While staying at 10NM, Intel is able to continue to squeeze performance gains out of the server.
It is important as we hit the limits of process geometry that everything be accelerated, especially with Intel having a similar view of AI workloads and that’s what we saw with Intel. New Optane memory, new persistent memory, faster Adapter cards which will lead to more Smart NIC announcements, and a 10nm FPGA. It was clear at the event that large cloud provers like AWS, Azure, and Tencent are looking at all avenues to increase performance and reduce power consumption via software and hardware advancements.
Some interesting highlights included AWS touting over 100 unique instances that leverage Intel processers with more SAP instances running on AWS than anywhere else and Formula 1 using of 65 years of historic race data to train its models in order to make real-time race predictions.
Western Digital Corporation, also known as WDC, held its investor meeting yesterday. We highlight four important disclosures: (a) it expects HDD to survive, despite flash's advantages, (b) WDC's plans to use RISC and launch a new memory interface, (c) an update on current business trends, and (d) the company's plans for its storage systems business.
First, since the company owns both Hard Disk Drive (HDD) and Flash divisions, it has an interest in keeping investors informed of the relative competitiveness between the two. Two and a half years ago, WDC closed on its acquisition of flash pioneer Sandisk. Around that time, the company was asserting that there is, and the company anticipates it will remain the case, that HDDs will retain at least a 10-times advantage versus flash on a cost per bit ($/GB) stored basis. At the company's meeting yesterday the company reiterated its view that this 10x advantage will continue at least till the year 2022.
The 10x $/GB differential is important because, if true, it means HDDs will not go away for many applications, especially for storage of big data at hyperscalers. Consider this - storage systems vendors such as Pure Storage are announcing software systems that allow premises-based flash storage to be seamlessly moved to the cloud for longer-term storage, which means that cloud hyperscalers like Amazon Web Services (AWS) are likely to be HDD customers for a very long time. Nevertheless, flash will keep on taking share from HDD. The company has been reducing its manufacturing capacity for HDD, with significant layoffs and factory shutdowns over the past several years. In fact, it plans significant ongoing cost cutting in its HDD manufacturing, it the range of 15-25% Y/Y decline next year.
The company also announced an ambitious plan to transition from licensed CPU cores to RISC based cores, for which it expects to pay no licensing and/or royalty fees. WDC says it ships over 1 billion CPU cores per year, so this is a significant shift. Also, the company plans to introduce an open-sourced memory interface called OmniXtend memory fabric, which will pit it against Intel (using DDR4, etc.) who has historically launched its own interfaces. WDC is now a very big player in storage, having built through organic growth and acquisitions. It has more market might and these new initiatives have a better opportunity to launch than in the past.
Additionally, the company said that its near-term business trends are under pressure due to cyclical market fluctuations. The memory semiconductor industry has historically endured a boom/bust cycle and now the company is explaining it has entered the bust part of that cycle. Since its earnings call on October 25, 2018, it says conditions have deteriorated somewhat more, partly the result of hyperscalers continuing to reduce inventories and partly due to mobile phone companies still reducing forecasts of demand. The company expects 2019 will see flash demand below the historical range and that hyperscalers should see a return to growth in 2H19.
Lastly, the company's Data Center Solutions group, which employs a vertically integrated strategy to compete with traditional storage systems companies such as DELL-EMC, NetApp, HPE, Pure Storage and others, just experienced a record quarter on a revenue basis and is approaching break-even in its operations. The company has the goal of becoming a top 5 player in data center solutions, which we take to mean, it is planning to take share from the current players. The group has experienced 17x revenue growth from Q1FY19 compared to Q1FY16, according the presentation (of course, the group has made acquisitions that bolster this number), has shipped 3 Exabytes in the months in calendar 2018 (which isn't over yet) and has shipped 8,500 systems and platforms since inception. The company plans to experience "double digit" revenue growth rates for this unit in the future.
There were two main announcements, a new relationship with Google Cloud Platform and a new flash device - the AFF A800. Also, in our interviews with NetApp, we learned about the future of Fibre Channel at the hyperscalers.
Google. Google Cloud Platform now integrates NetApp Cloud Volumes as a drop-down menu capability as part of the Google console. This allows enterprise customers, for instance, to use Cloud Volumes to manage their data on Google's cloud service while simultaneously managing their data on premise. This relationship with Google now rounds out the NetApp relationships with the main hyperscalers - it already has in place relationships with both Amazon (AWS) and Microsoft (Azure). NetApp Cloud Volumes on Google Cloud Platform is currently available as a "preview" capability (sign up at www.netapp.com/gcppreview) and is expected to go to commercial status by the end of 2018. Customers will pay Google for the use of NetApp Cloud Volumes.
AFF A800. New flash hardware available from NetApp, which besides having impressive density and low-latency capabilities supports NVMe-over-Fibre Channel. Of course, the product also supports 100 Gbps Ethernet. From a historical standpoint, it is interesting that NetApp, a company whose heritage was driven by storage over Ethernet, is touting Fibre Channel. But, that's what its customers are asking for in order to accelerate their on-premise workloads such as database (Oracle), ERP (SAP) and other mission-critical enterprise workloads. In our interviews with NetApp, we were told that Fibre Channel is growing faster than Ethernet - this makes sense given the company's foray in recent years to flash and low-latency workloads.
Fibre Channel at the hyperscalers? We asked about what is going on with the hyperscalers' architecture to adapt to AI/Deep Learning workloads. NetApp executives explained that AI workloads are different from traditional workloads; they are random, low latency workloads connecting to GPUs. This type of workload, we were told by NetApp, works very well when attached to Fibre Channel. From NetApp's perspective, if customers want to run AI workloads fastest, they would likely do so on-premise, using Fibre Channel. Yet, many customers run their workloads on hyperscalers, all of which use Internet Protocol and the underlying Ethernet infrastructure. We have always been skeptical that hyperscalers would adopt Fibre Channel. We believe the hyperscalers may work with vendors such as NetApp to develop additional software capabilities to address the changing workloads relating to AI/ML/GPU workloads in the future - on top of IP/Ethernet infrastructures.