Imagination Technologies Share Discussion Threads

[adventurous] 05/4/2017 12:31

Orkney The cars will have to pass the test, those that fail will get a thrashing by John Cleese :-)

[mallorca 9] 05/4/2017 12:25

Excellent article.
Surely all here will now short this via a spread bet and recover some losses.
It cannot be worth more than £20m.

[orkney] 05/4/2017 11:11

Re driverless cars. Driving standards in the uk imo have worsened over the years, what with new technology i.e. Sat navs, etc... And hate to say it but the influx of unqualified foreign drivers. Driverless cars will only worsen affairs.
What will happen to the Driving test?

[miahkaysor] 05/4/2017 11:08

Back up to 115p?

[tsmith2] 05/4/2017 11:08

Intra day reversal from here with a flat positive finish would be v v bullish

[mr.elbee] 05/4/2017 11:07

richard..just another symptom of the current mania for all things high tech and scientific, with no regard whatever to the precautionary principle.

Diesel cars anyone?

wifi health issues next

[mallorca 9] 05/4/2017 11:06

and the long fall begins ...

[borromini1] 05/4/2017 11:06

Sheep Herder - quite likely, I was pointing to the contradiction of Apple staying mute and IMG having meaningful discussions. How long before IMG make their next announcement if Apple stonewall all attempts at engagement?

Taking a new GPU design avoiding others patents into critical product with no empirical precedent seems to increase the risk of receiving a surprise failure maybe along the lines of previous failure mechanisms or maybe a new one. I was airing the notion that Apple might have been working on GPU design for longer than we think.

If Apple are moving an office relating to GPU design to St Albans that is literally next door to IMG in Kings Langley. My last question was about the balance between employees that change company and patents that don't. If IMG are correct in their assertion that their IP is comprehensively defended by patents and all rights are reserved as part of the licence contract. At what stage does taking an employee engrained with IP from another company and applying their knowledge, experience and expertise become copying the IP?

[richardc77] 05/4/2017 10:16

We will have to agree to disagree on the DC market. I get there are scenarios that it will work for, just think there are too many issues to overcome with a mixed pool of some driverless and mostly non driverless on the road for it to be widespread in the next 40-50yrs.

[nearlythere] 05/4/2017 10:03

Richardc77

Driversless cars may take time but for companies driverless lorries, taxis buses etc will have major cost saving. For HGV no more problems with driver hours they can used a lorry 24 /7, so imo driverless will be very big.

[sheep_herder] 05/4/2017 09:35

COLSMITH - just look at what technologies are coming out and ask yourself whether they need graphics processing power or more general compute power. Yes, many of these can be done on GPGPU but it's not as efficient as on a dedicated processor. What happens when the non-graphics compute requirement outstrips the graphics compute requirement? A tipping point perhaps.

[richardc77] 05/4/2017 09:34

I think you make a good point re "driverless" cars. A minority want it, and I don't mean 51/49 split! I suspect the reality is the idea won't work until the majority want "driverless" so it may end up being a "white elephant".

[colsmith] 05/4/2017 09:27

A few lateral thoughts or snake oil - I don't know.

I suppose one question that comes to mind in all this is what is the requirement for more and more GPU performance. Have APPLE found the holly grail of the next big thing customers want and it needs a huge uplift of processing to deliver; or is it GPU compute for highly intensive processing stuff like weather calculations etc. Can't see IOT stuff needs massive processing power but maybe it does.

Of course it may simply be that APPLE want to save the few million they pay to IMG; but that doesn't seem logical. Because they could have bought IMG flogged the non GPU stuff and saved the royalty and got a full GPU team. Not sure there is very clear thinking here but of course if they know over the next few years of APPLE new stuff what customers want, or will be convinced they want, it may be a good strategy. Sure intriguing patents issues aside. Maybe Softbank and ARM is what APPLE should be thinking about.

I wonder though with APPLE products being very expensive is there need to slash price of their products as others are starting to eat their market; as much of this stuff is becoming commodity; is that part of it?

I wonder also if the great and good in the boardrooms are struggling to find the next big thing - in cars shedloads are going to self drive but do folk in the main want to be driven - has anyone asked the driving community? Clearly there is a market for self drive but is it say 90% of drivers want it or 5% - fiscally makes a difference. There is a safety issue (*) as well - after all software apart from small processes isn't provable and confidence can come from testing in use. Landing a plane is simple stuff against driving in the M25 or north circular! I guess self drive will need GPU processing - maybe that that's it, the next big GPU market needing more power than today is self drive?

* as software is art not engineering (in a provable sense) is there a legal issue when a self drive crashes - is it the driver (a software entity) responsible and so the developer or will the passenger be liable?

[pal44] 05/4/2017 09:26

Bear trap I think... gla

[mallorca 9] 05/4/2017 09:16

short term target .... 37p

[sheep_herder] 05/4/2017 09:10

borromini1 - Apple will never release any details to IMG about their new design. They have no need to pander to the company's requests. And they certainly have no legal obligation to do so.

You are correct in your assertion that they will have product soon but they are still building the team to implement it, so the 12-15 month time frame is probably correct for getting to tape out. You don't need silicon to prove performance - this will have been modeled long ago. Major developments like this take 75-100 man years so the initial team looking at it will have been in place for at least 2. Whether they release the product on time depends on whether they can build the team fast enough. They already have plans to move the office from London to St Albans with the hope of recruiting more staff from the Cambridge area.

I'm not sure what your last question about transfer of knowledge is asking? You hire employees with experience in the area of expertise you are lacking specifically for their knowledge in this area.

[mister md] 05/4/2017 08:55

Bottom line is IMGs results over recent years were not pretty. Declining revenues and profits. 40m net debt. And now their biggest customer is about to walk. A great trading opportunity here from the lows the other day, but I'm afraid thats likely to be it unless any other players are desperate to acquire their technology for a meaningful amount

[borromini1] 05/4/2017 08:54

If Apple are putting a new GPU design into all their mobile products in 15 to 24 months they must have designed the GPU, arranged manufacture of silicon, tested and proved it matches or exceeds the performance of PowerVR 7XT Plus and compared it favourably with Furian's projected capabilities. Apple would not risk untested silicon in the iPhone product line that has generated the vast bulk of the value of the worlds biggest company by market capitalisation. All this with no visible GPU patent support. No wonder IMG asked Apple for proof that they were not utilising IMG IP or IMG confidential information. So Apple won't divulge any details on their new GPU design to IMG but IMG have already started an attempt at meaningful discussions on new licensing arrangements and will make an announcement in due course. How long can Apple stay mute in this process? Presumably a following IMG announcement will be sooner rather than later.

So last year when Apple declined to offer a takeover bid for the whole of IMG they had already advanced their own GPU implementation to the point that they were pretty confident they did not need PowerVR and associated baggage. Presumably they were not offered only the stripped down PowerVR unit and related assets as this breaks the IMG business model of a collection of related IP adding up to greater than the sum of the parts. Instead they cherry picked some IMG employees under the guise of assisting PowerVR implementation. What options does the IMG or say ARM IP licencing model offer for protecting against the effects of the transfer of knowledge from supplier to client? Will this be a case that breaks the model?

[sheep_herder] 05/4/2017 08:52

They can be in discussion for whatever they want but Apple aren't going to agree to any raise in royalties. What are you on about?

[adventurous] 05/4/2017 08:49

Sheep_Herder - Oh, I know plenty about contracts! ;-)

However IMG "Apple's notification has led Imagination to discuss with Apple potential alternative commercial arrangements for the current license and royalty agreement."

I doubt they are talking about lowering the royalties for the current.... when Apple have asserted they will not longer be using IMG IP!

[mallorca 9] 05/4/2017 08:26

Short now in place.
Drip drip down followed by a huge fall as the Co has to restructure.
I wonder when the creditors will be calling in the £40m net debt.

[sheep_herder] 05/4/2017 08:21

adventurous, I love your enthusiasm that IMG could "up the royalty fee". lols. Ever heard of a contract? The royalty rates have long been cast in stone for any licenses Apple have signed.

[0penallhours] 05/4/2017 08:06

So Apple can get it wrong when it comes to all things GPU:



FWIW I cant help thinking someone will make a bid for IMG in 2017 and then take the battle to Apple to offset the price paid for the company. If it was someone like AMD, Intel or Samsung they could afford the drawn out battle to recoup the purchase price and would then still own the IP and engineers to produce and use their own world-leading GPU designs.