Intel shuts down Granulate less than three years after $650M acquisition

[Maximus]

Intel shuts down Granulate less than three years after $650M acquisition

Intel is closing the Israeli startup it acquired in 2022, laying off its entire workforce of over 100 employees amid sweeping global cutbacks.

Intel is shutting down Granulate and laying off its entire workforce of over 100 employees. The cutbacks are part of broader layoffs taking place at Intel Israel this week.

Intel acquired Granulate for $650 million in March 2022. In recent months, Intel attempted to sell Granulate’s operations but ultimately failed to find a buyer.

Intel confirmed the closure, stating: "As part of Intel's transformation process, we continue to actively review each part of our product portfolio to ensure alignment with our strategic goals and core business. After extensive consideration, we have made the difficult decision to discontinue the Intel Tiber App-Level Optimization product line. We are committed to ensuring a smooth transition for our customers and will continue to provide support until the first quarter of 2025."

Intel shuts down Granulate less than three years after $650M acquisition | CTech

Intel is closing the Israeli startup it acquired in 2022, laying off its entire workforce of over 100 employees amid sweeping global cutbacks.

www.calcalistech.com

 

[Maximus]

The acquisition was done in 2022 so I assume Pat cannot blame it on previous CEO.

 

[hist78]

SANTA CLARA, Calif., March 31, 2022 – Intel Corporation today announced an agreement to acquire Granulate Cloud Solutions Ltd, an Israel-based developer of real-time continuous optimization software. The acquisition of Granulate will help cloud and data center customers maximize compute workload performance and reduce infrastructure and cloud costs. Deal terms are not being disclosed.

“Today’s cloud and data center customers demand scalable, high-performance software to make the most of their hardware deployments,” said Sandra Rivera, executive vice president and general manager of the Datacenter and AI Group at Intel. “Granulate’s cutting-edge autonomous optimization software can be applied to production workloads without requiring the customer to make changes to its code, driving optimized hardware and software value for every cloud and data center customer.”

Greg Lavender, chief technology officer, senior vice president and general manager of the Software and Advanced Technology Group at Intel, said: “We are building our portfolio of software optimization tools that offer flexible and scalable capabilities that allow us to meet the growing demand of the ubiquitous compute era. Granulate’s innovative approach to real-time optimization software complements Intel’s existing capabilities by helping customers realize performance gains, cloud cost reductions and continual workload learning.”

While cloud computing and microservices have created a new era of flexibility in distributed applications and deployment scalability, modern architectures have introduced more complex performance issues that are not easily managed by traditional operating systems and runtimes. Additionally, customers often deploy older Linux distributions and application libraries that are not up to date with the latest advancements in today’s high-performance CPUs. Intel is committed to helping its customers ensure they are right-sizing their compute clusters, instance types and cloud deployments.

Granulate’s autonomous optimization service solves these issues by reducing CPU utilization and application latencies. It does this by learning the customer’s application and deploying a customized set of continuous optimizations at runtime. This enables deployment on smaller compute clusters and instance types to improve application performance and drive down cloud and data center costs. Granulate’s service does not require developer intervention nor does it require the customer to make changes to its own code. Optimizations for the latest CPUs can be applied even on legacy Linux distributions and runtimes.

Granulate’s autonomous optimization service enables cloud and data center customers to significantly improve the performance of their deployments, reduce operational overhead and lower application costs.

“Together with Intel, we believe we can help customers achieve meaningful cost reductions and five times the throughput across workloads,” said Asaf Ezra, co-founder and CEO of Granulate. “As a part of Intel, Granulate will be able to deliver autonomous optimization capabilities to even more customers globally and rapidly expand its offering with the help of Intel’s 19,000 software engineers.”

Intel and Granulate’s relationship began in late 2019, when Granulate was part of the first graduating class of Intel® Ignite, the startup accelerator program that taps into Intel’s resources to help early-stage companies succeed. Over the past year, Intel and Granulate have worked together under a commercial agreement to collaborate on workload optimization on Xeon deployments. This collaboration resulted in gains in performance and decreases in costs for customers running on Intel processors. With the acquisition of Granulate, Intel will rapidly scale Granulate’s optimization software, including across Intel’s data center portfolio. Intel is investing in growth opportunities enabled by software, including disruptive end-to-end artificial intelligence and security platforms, services, and APIs.

Transaction Details and Timing

The transaction is expected to close in the second quarter of 2022, subject to typical closing conditions. At that time, Granulate’s approximately 120 employees will be integrated into Intel’s Datacenter and AI business unit.

Intel to Acquire Granulate

Granulate’s autonomous optimization software helps drive greater performance and ROI in the cloud and data center.

www.intel.com

 

[hist78]

$650 million -- gone down the drain!

That amount could buy a lot of fruit for Intel employees for several years.

 

[Maximus]

$650 million -- gone down the drain!

That amount could buy a lot of fruit for Intel employees for several years.

While Pat is complaining the slow process of ChipAct funding, Intel perfectly show how they spend 7.6% of ChipAct funding they will get (8.5 billions).

 

[XYang2023]

I don’t think that has anything to do with the CHIPS Act. Under his leadership, a considerable amount of resources were wasted:
1. Failed Tower acquisition and associated penalties
2. Significant hiring increases since he took over
3. Closing down this acquisition
4. Unsuccessful RISC-V initiatives
5. Ramping up production due to COVID, leading to current write-offs

That said, Intel has spent $30 billion associated with the CHIPS Act, which is an order of magnitude greater than the wasted or misallocated spending.

However, I do believe he should have refrained from raising executive compensation at the start of this year.

 

[hskuo]

$650 million -- gone down the drain!

That amount could buy a lot of fruit for Intel employees for several years.

We might measure it as 2 Hi NA EUV scanners are drained.

 

[blueone]

I wonder how much longer it's going to take for Intel's BoD and the investment community to realize that PG is the wrong CEO for Intel, that he always has been, and the BoD acts to replace him?

 

[Daniel Nenni]

I wonder how much longer it's going to take for Intel's BoD and the investment community to realize that PG is the wrong CEO for Intel, that he always has been, and the BoD acts to replace him?

I highly doubt they would replace Pat from someone outside of Intel so it would be more of the same. If Intel was sold I think Hock Tan and Broadcom would be interesting. Hock just won the Morris Chang Award for Excellence! He would whip Intel into shape.

I think 18A is a great achievement and bringing design back in house and away from TSMC is the right thing to do. I also think keeping Intel design and manufacturing together is the right thing to do. Two wins for Pat!

The only mistake Pat has made with the foundry business is to set expectations too high and I think Pat has learned his lesson. The executive staff running the foundry group knows what they are doing. They just need the time and space to do it. My opinion.

 

[Artificer60]

I wonder how much longer it's going to take for Intel's BoD and the investment community to realize that PG is the wrong CEO for Intel, that he always has been, and the BoD acts to replace him?

Assuming that you believe that Intel's current strategy is the right one, who else would you find that is committed to dealing with another 2-3 years of financial hemorrhaging to get the foundry business back on its feet? I'd argue that most potential CEO's you could bring in would try to split the company and unload the foundry business. In my opinion that this would have a serious negative impact on the design business as they would lose the advantages of influencing the process development to meet the needs of the design. Given that, I think that Intel is stuck with Gelsinger, warts and all.

 

[hist78]

Assuming that you believe that Intel's current strategy is the right one, who else would you find that is committed to dealing with another 2-3 years of financial hemorrhaging to get the foundry business back on its feet? I'd argue that most potential CEO's you could bring in would try to split the company and unload the foundry business. In my opinion that this would have a serious negative impact on the design business as they would lose the advantages of influencing the process development to meet the needs of the design. Given that, I think that Intel is stuck with Gelsinger, warts and all.

Intel’s Board of Directors is itself a troubled one. Their egos and lacking vision and knowledge prevent them from admitting they hired the wrong person and approved the wrong strategy from the start. If forced to fire a CEO, they would likely prefer to cite unrelated reasons, such as “having a consensual relationship with an employee, violating the company’s non-fraternization policy.”

Intel’s Board of Directors can’t even retain or recruit well-recognized industry veterans and experts for the board itself. How can we expect them to hire the right CEO or dismiss one when necessary?

 

[DepthFirst]

Intel’s Board of Directors is itself a troubled one. Their egos and lacking vision and knowledge prevent them from admitting they hired the wrong person and approved the wrong strategy from the start. If forced to fire a CEO, they would likely prefer to cite unrelated reasons, such as “having a consensual relationship with an employee, violating the company’s non-fraternization policy.”

Intel’s Board of Directors can’t even retain or recruit well-recognized industry veterans and experts for the board itself. How can we expect them to hire the right CEO or dismiss one when necessary?

If there would be a leadership change at Intel it needs to start with clearing the board of members who aren’t industry experts. The comparison of Intel’s board to TSMC’s is utterly damning. The loss of Lip-Bu Tan was very disappointing, Intel needed more like him not less.

 

[blueone]

If there would be a leadership change at Intel it needs to start with clearing the board of members who aren’t industry experts. The comparison of Intel’s board to TSMC’s is utterly damning. The loss of Lip-Bu Tan was very disappointing, Intel needed more like him not less.

Unfortunately, the lack of technical and industry expertise is endemic to almost all large US corporate BoDs. Take, for example, Nvidia's BoD. Arguably one of the most successful US semiconductor companies ever, and yet, take a look at the BoD:

https://investor.nvidia.com/governance/board-of-directors/default.aspx

A former Taco Bell executive?

A former chief marketing officer for the National Football League?

A former CIO for Eli Lilly?

Two college professors and a bunch of VCs.

Seriously?

 

[hist78]

I highly doubt they would replace Pat from someone outside of Intel so it would be more of the same. If Intel was sold I think Hock Tan and Broadcom would be interesting. Hock just won the Morris Chang Award for Excellence! He would whip Intel into shape.

I think 18A is a great achievement and bringing design back in house and away from TSMC is the right thing to do. I also think keeping Intel design and manufacturing together is the right thing to do. Two wins for Pat!

The only mistake Pat has made with the foundry business is to set expectations too high and I think Pat has learned his lesson. The executive staff running the foundry group knows what they are doing. They just need the time and space to do it. My opinion.

"The only mistake Pat has made with the foundry business is to set expectations too high and I think Pat has learned his lesson. The executive staff running the foundry group knows what they are doing. They just need the time and space to do it. My opinion."

I believe Pat Gelsinger made more than one mistake with Intel's foundry business strategy:

His ego and lack of restraint on his mouth have damaged many relationships and potential business opportunities. For instance, under his leadership, Intel released a series of videos attacking Apple Macs. Then he turned around and said he hopes Apple will return to Intel as a foundry customer. Even if everything in those videos were true, shouldn’t this kind of ad campaign have been left to companies like HP, Dell, Lenovo, or Microsoft, rather than Intel?

Another example is Pat Gelsinger’s repeated comments about TSMC’s location, due to a vicious and aggressive neighbor across Taiwan Strait. He has damaged the Intel-TSMC relationship that was built over more than 40 years started by founders like Morris Chang, Bob Noyce, and Andy Grove. When TSMC needed more internal and external CoWoS capacity to meet exploding AI demand, the CoWoS outsourcing went to Amkor, not Intel.

Under Pat Gelsinger's leadership, Intel proposed building too many fabs, facilities, and R&D centers in too many locations. At one point, Intel had new foundry/manufacturing and R&D projects across Malaysia, Vietnam, Oregon, Arizona, New Mexico, Ohio, Israel, Ireland, Germany, France, Italy, Spain, and Poland. I may have even missed a few. Now, several of these projects are on hold due to funding and demand issues. Sadly, Intel has also begun laying off more than 15,000 employees.

How can Pat Gelsinger be allowed to continue with such a reckless approach to Intel’s foundry strategy?

 

[siliconbruh999]

I highly doubt they would replace Pat from someone outside of Intel so it would be more of the same. If Intel was sold I think Hock Tan and Broadcom would be interesting. Hock just won the Morris Chang Award for Excellence! He would whip Intel into shape.

I think 18A is a great achievement and bringing design back in house and away from TSMC is the right thing to do. I also think keeping Intel design and manufacturing together is the right thing to do. Two wins for Pat!

The only mistake Pat has made with the foundry business is to set expectations too high and I think Pat has learned his lesson. The executive staff running the foundry group knows what they are doing. They just need the time and space to do it. My opinion.

I think he is missing DC GPU delaying Falcon Shores was not a good idea

They finally have simplified the core design they said it in the earnings call and going forward both P and E core will be unified into one like AMD

 

[samwilde]

They finally have simplified the core design they said it in the earnings call and going forward both P and E core will be unified into one like AMD

That is a good idea. It would cut R&D on CPUs by like half without losing much in terms of market coverage.
I also think Intel should trim down their middle management.

In the long term Intel will also need proper PDKs for DUV FinFET, 3nm, and 18A processes so they will be able to capture external clients for the foundry.

I disagree with the idea that Intel's investment into RISC-V was a bad investment. It helped provide an alternative to ARM in the low end. Which means it will fragment ARM's income stream and make it easier for Intel to compete against them. At the same time RISC-V is a growing market and the investment would pay itself off if Intel holds on to it for long enough.

It was also a missed opportunity. Intel could have used RISC-V to add another customer to its foundry business. By cancelling the Intel-SiFive collaboration, SiFive were forced to go with ESWIN, a Chinese chip designer, and are now fabbing their chips at TSMC instead. In several years time people will see this as a failure similar to Intel losing out the Apple smartphone SoC bid.

 

[hskuo]

"The only mistake Pat has made with the foundry business is to set expectations too high and I think Pat has learned his lesson. The executive staff running the foundry group knows what they are doing. They just need the time and space to do it. My opinion."

I believe Pat Gelsinger made more than one mistake with Intel's foundry business strategy:

His ego and lack of restraint on his mouth have damaged many relationships and potential business opportunities. For instance, under his leadership, Intel released a series of videos attacking Apple Macs. Then he turned around and said he hopes Apple will return to Intel as a foundry customer. Even if everything in those videos were true, shouldn’t this kind of ad campaign have been left to companies like HP, Dell, Lenovo, or Microsoft, rather than Intel?

Another example is Pat Gelsinger’s repeated comments about TSMC’s location, due to a vicious and aggressive neighbor across Taiwan Strait. He has damaged the Intel-TSMC relationship that was built over more than 40 years started by founders like Morris Chang, Bob Noyce, and Andy Grove. When TSMC needed more internal and external CoWoS capacity to meet exploding AI demand, the CoWoS outsourcing went to Amkor, not Intel.

Under Pat Gelsinger's leadership, Intel proposed building too many fabs, facilities, and R&D centers in too many locations. At one point, Intel had new foundry/manufacturing and R&D projects across Malaysia, Vietnam, Oregon, Arizona, New Mexico, Ohio, Israel, Ireland, Germany, France, Italy, Spain, and Poland. I may have even missed a few. Now, several of these projects are on hold due to funding and demand issues. Sadly, Intel has also begun laying off more than 15,000 employees.

How can Pat Gelsinger be allowed to continue with such a reckless approach to Intel’s foundry strategy?

IMO, when Pat accepted the job, the goal had been set. The problems are the targets to be achieved and timeline. What I saw is, Pat set too many targets and it became defocus. It was not that easy to develop 5 HVM nodes in 4 years and constructed several fabs in parallel. Even tsmc can not do it. The troubles are expected and good luck for intel.

 

[hist78]

That is a good idea. It would cut R&D on CPUs by like half without losing much in terms of market coverage.
I also think Intel should trim down their middle management.

In the long term Intel will also need proper PDKs for DUV FinFET, 3nm, and 18A processes so they will be able to capture external clients for the foundry.

I disagree with the idea that Intel's investment into RISC-V was a bad investment. It helped provide an alternative to ARM in the low end. Which means it will fragment ARM's income stream and make it easier for Intel to compete against them. At the same time RISC-V is a growing market and the investment would pay itself off if Intel holds on to it for long enough.

It was also a missed opportunity. Intel could have used RISC-V to add another customer to its foundry business. By cancelling the Intel-SiFive collaboration, SiFive were forced to go with ESWIN, a Chinese chip designer, and are now fabbing their chips at TSMC instead. In several years time people will see this as a failure similar to Intel losing out the Apple smartphone SoC bid.

As an IDM, Intel can focus solely on x86 and disregard other architectures. However, as a foundry, Intel must be open to working with x86, ARM, RISC-V, and any other technologies that could bring in significant revenue, even if they might jeopardize Intel's own x86 market. This inherent conflict of interest between Intel’s IDM and foundry roles will persist until the day Intel’s Design/Product division and Manufacturing division split into two independent companies.

 

[XYang2023]

IMO, when Pat accepted the job, the goal had been set. The problems are the targets to be achieved and timeline. What I saw is, Pat set too many targets and it became defocus. It was not that easy to develop 5 HVM nodes in 4 years and constructed several fabs in parallel. Even tsmc can not do it. The troubles are expected and good luck for intel.

He should set 3 targets:
1. Copy Nvidia's strategy
2. Cost cutting
3. IDM 2.0 (questionable). He should select the most cost-effective way to achieve the strategy, for example, just focusing on Ireland and Arizona.

 

[delong.height]

Just like what I said in the past, Intel should upgraded its smart capital strategy. No more fab shell buildout; outsource that to Samsung. Outsource to Asia, There're plenty of countries beside China, and Taiwan that are willing to do fab business. Like Japan.

Intel shall also never ever pursue acquisition. Every deal they go for have failed miserably.