[The Register] Intel Buys Soft Machines

[Arachnotronic]

http://www.theregister.co.uk/2016/09/09/intel_soft_machines/

Soft Machines honestly came off to me as basically snake oil, but maybe they've actually developed some worthwhile IP?

 

[frozentundra123456]

http://www.theregister.co.uk/2016/09/09/intel_soft_machines/

Soft Machines honestly came off to me as basically snake oil, but maybe they've actually developed some worthwhile IP?

Strikes me as a bit desperate, but I suppose 250 mil is petty cash to intel.

Edit: for a company that should have some very impressive business minds, seems like Intel's acquisitions have turned out very poorly.

 

[witeken]

Way to go Intel! Extremely surprising, given their bold claims. That's what we wanne see. The only the paranoid survive mentality. Doing everything to bolster their CPUs.

Anandtech has article about it, BTW.

As it stands the design promises a lot, but because we really need to see the proper silicon implementation, it might be hard to visualize until a company in the technology ecosystem decides to make that step. It would be an interesting differentiation point for sure, but it requires investment to reach utility in mass production. That makes a number of analysts wary and conservative with good reasons, especially with the assumptions made on that data graph.

http://www.anandtech.com/show/10025/examining-soft-machines-architecture-visc-ipc/

 

[Nothingness]

A new low in Intel acquisitions? Soft Machines has never been credible. But I guess Intel is not that stupid, so they might have some interesting things, such as some patents.

 

[Arachnotronic]

A new low in Intel acquisitions? Soft Machines has never been credible. But I guess Intel is not that stupid, so they might have some interesting things, such as some patents.

Ah, so my intuition that Soft Machines was basically selling snake oil is not unique.

 

[Nothingness]

Ah, so my intuition that Soft Machines was basically selling snake oil is not unique.

Their VISC is a joke.. They demonstrated a CPU running at 350 MHz and claimed large IPC. Funny.

Add to that you need to change your software and you have a winner.

 

[Arachnotronic]

Their VISC is a joke.. They demonstrated a CPU running at 350 MHz and claimed large IPC. Funny.

Add to that you need to change your software and you have a winner.

LOL!

 

[witeken]

Ah, so my intuition that Soft Machines was basically selling snake oil is not unique.

Intel is not stupid. And SM have products taped out, so it isn't like it's still a design on paper. I'll have to re read the article AT published, but it's likely they have some valuable things that Intel could use. For instance, the thing is ISA neutral. If some of the things this company does match with what Intel is planning for their future Tocks (like your 2022 job listing), then it would be a logical buy. If nothing gets disclosed (or if this rumor isn't correct), then we're just left guessing.

Part of Soft Machines' product package is a low level virtual software layer that will translate a guest instruction set and convert it into the VISC ISA. This is to allow VISC to be used with existing software, and to more easily integrate into current environments rather than trying to establish an ecosystem for a new architecture in 2016. Soft Machines tells us that two instruction sets are supported, one of which will be ARMv8. It was implied that x86 would be the other, although they were reluctant to outright confirm it (ed: x86 translation is likely not to be looked upon fondly by Intel). Meanwhile we were told that writing additional translation layers, while not trivial, can be done and that they plan to support other guest ISAs in future.

 

[dark zero]

Their VISC is a joke.. They demonstrated a CPU running at 350 MHz and claimed large IPC. Funny.

Add to that you need to change your software and you have a winner.

They want to get rid of VISC since AMD, IBM, Samsung and Qualcomm wanted to promote it... now that path is destroyed.

 

[CHADBOGA]

Intel is not stupid.

They keep blowing billions on ridiculous acquisitions.

McAfee was always a ridiculous purchase and so it proved to be.

 

[Arachnotronic]

They keep blowing billions on ridiculous acquisitions.

McAfee was always a ridiculous purchase and so it proved to be.

I'm interested in seeing how the Altera buy plays out. Will it work out well, or will it just be more like AMD buying ATI?

 

[itsmydamnation]

look at a company that is an expert in technology acquisition's, Cisco. Even Cisco has at best a 50% ( more like 33%) hit rate of successful acquisition's. its not just about how good the technology is, its just hard.

The problem with buying mcafee is they have a product for everything and none of them are that great.

 

[NTMBK]

7th acquisition that Intel have made so far this year. They're really throwing around the pocket book!

 

[sm625]

I consider this to be one more piece of the puzzle. I've been predicting for a while now that Intel is heading towards reprogrammable execution units in its cores. It's purchase of Altera and now Soft Machines quite clearly illustrates that this is where they are heading. The potential for ASIC style improvements in performance per watt is huge. Consider what quicksync has done for encoding performance per watt and apply that to something like a javascript engine. IPC could increase tenfold for applications that can make use of programmable logic located extremely close to the cpu core.

 

[FIVR]

Altera was not a bad acquisition IMO, although it remains to be seen what exactly intel does with the company. Soft machines is not in the same league. Altera had a working, competitive product pipeline and a customer base. What tangible benefit does soft machines bring to the table?

 

[witeken]

If Intel takes over a company for $250M, are they obliged to explain their stuff?

 

[DrMrLordX]

If Intel takes over a company for $250M, are they obliged to explain their stuff?

Only if they try peddling it to anyone.

I remember when VISC was first discussed around here . . . well maybe not the first time it was discussed, but anyway it was awhile back. Conspirators at the time thought this tech would be used in AMD CPUs if I recall correctly. Doesn't look like that's happening now.

 

[cytg111]

Getting out of mobile and into IoT. How is this a good fit?

Of course if it pans out, we will finally have a GOOD reason to go beyond 4 cores.

 

[Exophase]

I've read that Soft Machines received $200m in investment funding. A $250m acquisition would probably be a loss, meaning that the investors must have not had a lot of faith in the company's profitability.

They could still have viable ideas and technologies, but lacked enough in other areas needed to be competitive overall. Areas that Intel is strong in. Or maybe Intel had plans to do something that overlapped with their existing patents and they worried about future patent trolling.

 

[CHADBOGA]

What was the last acquisition by Intel that was an unambiguous success?

Has there ever been one?

 

[Nothingness]

What was the last acquisition by Intel that was an unambiguous success?

Has there ever been one?

Transmeta

 

[NTMBK]

What was the last acquisition by Intel that was an unambiguous success?

Has there ever been one?

Their StrongARM business was pretty successful, before they got rid of it for religious reasons.

 

[Arachnotronic]

What was the last acquisition by Intel that was an unambiguous success?

Has there ever been one?

Some of the smaller "tuck in" acquisitions for technology's sake have been pretty good. The QLogic Infiniband assets are serving as the basis for Omni Path (which isn't living up to the hype but is emerging as a credible lower cost alternative to InfiniBand). I also think the Fulcrum Microsystems acquisition which got them into Ethernet switches went okay.

Anyway, Intel truly sucks at big acquisitions, but the smaller ones tend to be okay since they are bought for technology, not for revenue.

 

[Nothingness]

One could also consider Infineon acquisition might eventually be considered as a success.

 

[Arachnotronic]

One could also consider Infineon acquisition might eventually be considered as a success.

It's getting better, Intel's finally inside of the iPhone, so that's a plus. Long way to go before it really pays off though, it has actually just been a gigantic money pit.