Fulcrum (Calabasas, Calif.), a privately held, fabless company, produces high bandwidth network switching chips based on an asynchronous switching fabric for data center network providers. The acquisition of Fulcrum could give Intel a head start on an end-to-end solution for data centers.
"Intel is transforming from a leading server technology company to a comprehensive data center provider that offers computing, storage and networking building blocks," said Kirk Skaugen, Intel vice president and general manager of the company's Data Center Group, in a statement. Skaugen said Fulcrum's switch silicon complements Intel’s processors and Ethernet controllers and would enable Intel to offer customers new levels of performance and energy efficiency while improving the economics of cloud service delivery.
Fulcrum, founded in 2000, was previously included on early integrations of the The EE Times 60 Emerging Startups list. The company has announced several rounds of venture funding over the years totaling at least $55 million for the development of its clockless 10 Gigabit Ethernet and 40 Gigabit Ethernet chips. Fulcrum's technology removes the central clock, which can result in chips consuming less power while performing at gigahertz performance in standard commodity manufacturing processes.
Intel (Santa Clara, Calif.) said the acquisition would fulfill an important component in the company's strategy to deliver comprehensive data center building blocks, from server processors and technologies to storage and networking.
The acquisition agreement remains subject to the approval of Fulcrum Microsystems shareholders, as well as regulatory approval and other customary closing conditions, Intel said. It is expected to close in the third quarter, the company said.
@iniewski: your comment is on the right track -I think the answer to Intel's intentions of buying Fulcrum is both! Intel would be certainly interested in asynchronous processor core tech for lower latencies but the market opportunity with Fulcrum's FocalPoint switches (for which Intel was one the early NIC OEM's) is also appealing.
Fulcrum in Asynchronous went a similar route as Celoxica with C to Gates - first technology - then concentrate on product.
Asynch can achieve a new speed step, so Achronix and Fulcrum fit this - compared to the incremental speed of smaller geometries. Even if only used in buses and peripherals.
Looking forward to implementation - e.g. asynch on-chip buses.
So it's now only a question of time before Intel also picks up Aquantia. A perfect fit with Fulcrum for the data center plan.
Now, that said, we shouldn't forget Intel's previous spree into a similar albeit back then smaller market when they picked up Level One and that Danish company paying billions for both like 12-14 years ago...
Some years ago, Intel gave away the IXP network processors section because they wanted to focus on PCs. Now they would have been major leaders in network embedded devices and they wouldn't need to buy other companies.
It is not signal shift but back to the future Dylan...ten years ago Intel was acquiring optical networking companies but later dumped that IC Internet portfolio (there is still design group in Vancouver left that was doing networking design but today they work on memory sub-systems)...Intel also was developing and selling network processors...so there is a long history of Intel trying to get into Internet/data centers, usually without much success...so this acquisition must be about something else as other speculate above...Kris
Seconding what JMUNN said.
I remember back when Fulcrum was started--they advertised that they had a method of designing asynchronous logic state machines, including their own design toolsets. This was touted as how they were getting such low latency in their chips. Now they don't talk about it at all. (Trade secret?)
I wouldn't be surprised if this is what Intel is after and they don't even care about the switch chips.
@JMUNN-It's a good question. I thought about that. A lot of people felt there was more to that Achronix deal than Intel let on, and the asynchronous nature of Achronix' technology seened appealing to Intel. There may be a thread there.
Are they going after Asynchronous IP and/or patents? They are working with Achronix FPGA which is also asynchronous to be their fab. Is this an attempt to someday move to an asynchronous processor core development?
Actually, I think Intel is over billing this in its press release (Gee, that's never happened before!) The company has long had a secondary business selling Ethernet chips and adapters. The Fulcrum addition butresses that business nicely, but I suspect they will still play second fiddle in Ethenet to companies such as Broadcom.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.