Thread regarding Marvell Technology Group Ltd. layoffs

MSFT acquiring ThunderX design team

anyone hear any rumors? grape vine says layoffs were to reduce the OpEx numbers for the sale... MSFT wants thunderX to compete with AWS...

by
| 5018 views | | 12 replies (last )
Post ID: @OP+16nYwSnO

12 replies (most recent on top)

What was the reason for the failure of the original Cavium Thunder project ?

by
|
Post ID: @qpou+16nYwSnO

The problem is in the deep-rooted buddy system in Silicon Valley. If you have friends in high places ready to cover your big mistakes, then nothing happens to you. You continue to get high positions and well paying jobs despite screwing up at multiple places. Then you return the favor by covering the a_s of the person who covered your a_s previously. This is the way hi-tech industry works.
These people who came as a 'Vulcan' project group from Broadcom had such a bad reputation in Broadcom. They worked on the chip design for over four years and yet kept missing the takeout for nearly a year. Each month they will say, we missed the takeout but it will happen next month. The next month they would say the same thing all over again. Finally when the chip came out it was such a bad failure that I have never witnessed such a big disaster in my life!. Broadcom decided to shut the division.

by
|
Post ID: @pezz+16nYwSnO

The sad thing about the situation is that "on paper" these individuals look as if they have many years of experience and end-up enjoying senior, well-paid positions. Invariably, they go on to replicate the failure inherent in their profound competence at some other company, so they tend to either bail at the 11th hour when the chickens come home to roost, leaving behind a spiders nest of broken RTL, or throw someone else into the firing line as a means of self-preservation. It works, and these people make a career out of it.

by
|
Post ID: @pvmz+16nYwSnO

Broadcom Vulcan team was a disgrace to work with. Especially the so called "architecture" team. They burned hundreds of millions of dollars in Broadcom for four years and when the chip finally came out for internal testing, nothing worked. You would be lucky if one or two cores in chip worked for you and that too at low speed of 1 GHz. It was a disgrace and Broadcom management shutdown the company. Then they found some id–ts in Cavium to buy them!. Make me wonder how bad Silicon Valley has gotten in last 10 years or so?. These people should never have found any job in valley but not only did they got hired by Cavium, but to top it off they got promoted. Now I heard the Chief architect of this losing team has found job in Nvidia!!!!

by
|
Post ID: @pdrk+16nYwSnO

Thunder was a failure long before the Vulcan guys ever showed up at Cavium. And if there decision to hire them was a mistake, it was only one in the litany of mistakes that had been made prior.

The place was/is a steaming cesspit of dysfunction, and you only have yourselves to blame for that.

by
|
Post ID: @imxj+16nYwSnO

There’s no managing to get anything. SoftBank decided to fund some companies who are doing arm servers. At that time it was supposed to be funded until 2022. With this pandemic and SoftBank’s loosing big time on all of its recent funding it wouldn’t be a surprise if they have pulled out. And this would be the reason for Mrvl looking for a buyer. If MSFT is going to buy then it’s actually a blessing.

by
|
Post ID: @8cmc+16nYwSnO

How did they manage external funding? what is the source ?

by
|
Post ID: @8hfg+16nYwSnO

Previous comments are bang on target and the very reason for server product line to be a failure. Have worked with the vulcan team initially when cavium bought them. Pretty soon realized that the whole team (like 60 of them) barring few was a big mistake that cavium ever did. And the hiring scams they did is unspeakable which brought in more sh–ty folks calling themselves engineers. Had it not been external funding they wouldn’t have lasted this long. I hope Microsoft sees through them and not repeat the same mistake that cavium did.

by
|
Post ID: @6rlo+16nYwSnO

The Cavium team at Marvell is now mainly the Vulcan team, which was the remnants of Netlogic. These people are notorious in the Bay for their bad reputation. Vulcan was such a disaster than Broadcom was forced to abandon it and sell it to Cavium for next to nothing (now X2). Most of these people can barely string together compilable Verilog, let alone design anything meaningful. Don't go near them.

by
|
Post ID: @4cxi+16nYwSnO

The Cavium team at Marvell is now mainly the Vulcan team, which was the remnants of Netlogic. These people are notorious in the Bay for their bad reputation. Vulcan was such a disaster than Broadcom was forced to abandon it and sell it to Cavium for next to nothing (now X2). Most of these people can barely string together compilable Verilog, let alone design anything meaningful. Don't go near them.

by
|
Post ID: @4oop+16nYwSnO

If i was MSFT and I am desperate then would buy the IP but not the team. The verif is a piece of sh– team.

by
|
Post ID: @2qgo+16nYwSnO

The series doesn't have the compute performance to be competitive and has fallen rather flat on it's face. Wouldn't be surprised if Marvell was interested in offloading the group to some third party. Also worth noting that much of the architecture talent from Cavium has long departed because things are being wound down. Microsoft already have most of the Centriq team in Raleigh though, so they already have this design capability, but from what I hear these are working in the quantum computing effort. Maybe a bargain price that's too good to turn down though.

by
|
Post ID: @owc+16nYwSnO

Post a reply

: