Analytics With the significant decline in revenue and profitability over the past six months, Intel has made some tough decisions ahead as it seeks multibillion-dollar cuts as the embattled semiconductor company tries to implement its grand plan for a comeback.
As announced last week, the US chip maker plans to Reducing spending by $3 billion annually Starting next year and then increasing that to $10 billion by 2025.
While much of these spending cuts will come from laying off a “meaningful number” of people, the company also said it would also consider “portfolio cuts.” This means that Intel will be looking to cut out some of the products it makes and sells.
In some cases, Intel may look to offload products to a willing buyer, as it has reduction NAND flash and solid state drives work. But in other cases, the company may decide to discontinue and write off products, as it has done with its own Business Memory Optan Back in July.
“We remain committed to improving value creation efforts through portfolio optimization, reallocation of resources for higher returns, higher growth companies, mergers and acquisitions and, where possible, divestment,” Intel CEO Pat Gelsinger said of the company. 3rd Quarter Earnings Call Since a week
While we don’t yet know where Intel will be cutting products, we’ve decided to make some informed guesses about where the company could cut back on its portfolio.
Havana chips AI
In December 2019, Intel made the seemingly hasty decision to pay $2 billion to Acquisition of AI chip company Habana Labsjust weeks after the company showcased its impressive deep learning processors from its previous acquisition, Nervana Systems.
Knowing it needed a solid answer for Nvidia’s fast-growing AI chip business, it seems Intel realized it had better prospects with Habana, so it decided to end of development From Nervana deep learning chips to focus on new acquisition.
Three years later, Intel has made some progress with Habana. Its Gaudi training chip is now available in cloud instances from Amazon Web Services, there are Habana-supported systems available from Supermicro, and the division is working on second generations of its own training and inference chips, the latter of which are mapped to new servers from Supermicro, Inspur and WiWynn.
The problem is that Habana likely won’t make progress fast enough to deliver a meaningful blow to Nvidia’s multi-billion dollar AI chip business anytime soon. So far, Intel has Few details revealed On the end users who have used Habana chips, one of them is Recently Mobileye section spun out.
Habana chips are still not supported before oneAPIIntel’s answer to CUDA platform from Nvidia. Then there’s the fact that Intel didn’t make enough money from Habana to disclose sales in earnings.
Intel has positioned Habana chips as more efficient alternatives to GPUs for training and inferring AI at scale, so if processors can do their job, it may be worth it for Intel to stick with them in the long run. But since the company is badly in need of extra cash and Habana isn’t making a lot of money, it might be best for Intel to separate the division to focus on More flexible GPU architecture.
And yes, we know Intel GPUs don’t make That much money now too. But if Intel really wants to become a more viable competitor to Nvidia’s AI prowess, it needs a dedicated acceleration chip of some sort. And how Intel can recover GPU development costs Multiple production linesGPUs may make more financial sense.
Blockscale ASIC
Here is a question worth asking: Is the cryptocurrency mining industry stable enough for Intel to continue to choose? Goliath x86 just started shipping its first bitcoin mining accelerator chip, the Blockscale ASIC, back in junepromising to provide a power-saving alternative to power-hungry GPUs.
The problem is that things weren’t great for crypto-mining companies when Blockscale hit the market, and these conditions haven’t improved.
The financial times reported on Wednesday That crypto-mining companies are on shaky footing due to rising energy costs and declining cryptocurrency values that began earlier this year. The so-called crypto winter has already caused one of the data center providers to crypto companies, Computer North, to program file for bankruptcyOther companies are facing major financial problems.
This includes one of Blockscale’s early clients, Argo Blockchain, who warned last week that it was at risk of halting operations due to a lack of funding. Another early Blockscale customer, Griid, is face allegations It is the lender that the company has defaulted on its main line of credit.
Blockscale’s revenue can’t get much at the moment given that its product business unit, Rapid Computing and Graphics Systems, brought in just $185 million for Intel in the third quarter, with most of that likely coming from the company’s GPUs.
Perhaps conditions for the crypto-mining industry will suddenly improve, and Intel’s bet on Blockscale will seem worthwhile in the long run. But in the midst of a faltering economy that could only get worse, the company may run out of patience with Blockscale’s viability sooner rather than later.
RealSense stereo cameras
Remember when Intel said last year that it was “finish” RealSense Computer Vision Business? Turns out that wasn’t quite the case as the company decided to stick with it Stereo camera equipment Eliminate LiDAR, face authentication, and trace production lines.
Intel obviously has enough customers to justify the continued development and sales of its stereo cameras, but we’re sad to say that these premium robotics and IoT devices will be a solid choice for the chopping block, whether that means discontinuing them or selling them to another company.
Simply put, RealSense cameras are out of focus of the core Intel chip. Yes, Intel makes use of its ability to design the chips that go inside these cameras, but the sensors themselves? Developing such devices may not be a priority when a company needs to focus on making its chips — primarily CPUs — more powerful and efficient than competing products.
Pat Gelsinger, CEO of Intel He said it himself When this reporter asked about his decision to terminate RealSense’s business.
“Hey there are some good assets we can harvest, but they don’t fit into one of those six business units I’ve built,” he said last year. The six divisions he referred to are Intel’s Personal Computers, Data Center, Networking and Edge, Graphics, Foundry, and the recently formed Mobileye business.
Mini PC NUC
Don’t get us wrong. Intel’s NUC mini PCs are great with small feats of engineering, especially those that can fit up to Core i7 CPU and mid-range graphics cards in a box That fits easily in your backpack.
Intel deserves a lot of credit for pushing the boundaries of mini PC design and finding new ways to balance power, performance, and space in small form factors. But now that the company needs to make tough spending decisions, it may be time to reassess NUC’s future under Intel.
Back in May, the chipmaker a statement It has sold over 10 million NUCs since it introduced the Next Unit Computing brand a decade ago. Ten million is anything but impressive, but we have to remember that hundreds of millions of computers are shipped every year. Dell alone shipped nearly 60 million last year out of 340 million total PCs sent in 2021.
It’s not that the new urban communities don’t offer value. While New Urban Communities (NUCs) are enjoyed by enthusiasts, they are also used by many companies around the world. Just NUC sales probably represent a small percentage of the company’s total revenue, and manufacturing small, well-designed computers falls outside the company’s core chip strategy. Fortunately, Intel is far from being the only company Makes small computers Currently.
Intel had no comment on this issue at the time of publication. ®
Do you have any ideas about where Intel might make cuts in its product lineup? Leave us a comment. Or, better yet, if you know what Intel plans already, you can always Email us in confidence.