Intel launched new Xeon chips on Tuesday to offer better support to next-generation computing applications like artificial intelligence, driverless cars, and others, according to Naveen Rao, vice president of Intel’s Artificial Intelligence Products Group. There is no doubt that with its new chips, Intel has entered the war zone against NVIDIA and AMD, which recently challenged Intel’s near-monopoly position in the data center business.
Can Intel’s Xeon chips beat the competition?
What Intel could not understand all these years is the fact that it cannot stay ahead just by putting an increasing number of transistors on every chip. Such a strategy has become less effective with Moore’s Law slowing down. Competitors that designed their chips to perform better by way of parallel processing stand to benefit from improving their performance, notes Fortune.
However, Intel stated that its Xeon chips are 1.65 times faster on average compared to its previous generation when performing common tasks. The company also stated that software had been specifically designed to work better with its chips.
Lisa Spelman, vice president of Intel’s Data Center Group, told Fortune, “This represents the best of our 20-year history of data center innovation.”
According to Gartner analyst Martin Reynolds, Intel’s new microprocessors are more efficient than to their predecessors and will be more effective when it comes to handling artificial intelligence workloads and advanced storage. However, Reynolds also pointed to Intel’s over-dependence on a relatively small number of big data center operators.
“The challenge now is so much of their work is going to these big internet guys,” the analyst said.
Intel’s newest family of microprocessor targets Alphabet, Amazon.com and Microsoft, among others that operate big data centers and deploy thousands of computers to run their own services and offer the same services to clients who do not have their own computer systems, notes Reuters.
Although the Xeon chips were formally launched on Tuesday, some large customers such as Google have already received shipments. Intel has now shipped about 500,000 chips to its large customers, notes Fortune.
Analyst not favoring Intel
Intel was recently downgraded from Hold to Underperform by Jefferies, which stated that the reason for downgrading Intel is its inferior Xeon/Xeon PHI platform compared to NVIDIA’s products in emerging parallel workloads such as deep neural networking.
Jefferies also has various other areas of concern, such as Microsoft’s new support for ARM processors in Windows and the NVIDIA data center business, which is increasing at a rapid growth rate of 200%. NVIDIA is also getting popular among investors, as in May, SoftBank Group acquired a $4 billion stake in the company.
According to Bernstein Research analyst Stacy Rasgon, “This is the most competitive environment we have seen in a decade.” Intel wants to make it clear that it is still the supreme and that it is going to stay in this business positively, the analyst said. “Part of this is because there is so much hype around the competing products, they want a chance to tell their story,” said Rasgon.
On Tuesday, Intel shares closed up 0.80% at $33.92. Year to date, the stock is down over 6%.
