ADVFN Logo

We could not find any results for:
Make sure your spelling is correct or try broadening your search.

Trending Now

Toplists

It looks like you aren't logged in.
Click the button below to log in and view your recent history.

Hot Features

Registration Strip Icon for alerts Register for real-time alerts, custom portfolio, and market movers

IMG Imagination Technologies Group

181.25
0.00 (0.00%)
28 Mar 2024 - Closed
Delayed by 15 minutes
Share Name Share Symbol Market Type Share ISIN Share Description
Imagination Technologies Group LSE:IMG London Ordinary Share GB0009303123 ORD 10P
  Price Change % Change Share Price Bid Price Offer Price High Price Low Price Open Price Shares Traded Last Trade
  0.00 0.00% 181.25 181.50 181.75 - 0.00 00:00:00
Industry Sector Turnover Profit EPS - Basic PE Ratio Market Cap
0 0 N/A 0

Imagination Technologies Share Discussion Threads

Showing 40401 to 40424 of 43000 messages
Chat Pages: Latest  1624  1623  1622  1621  1620  1619  1618  1617  1616  1615  1614  1613  Older
DateSubjectAuthorDiscuss
14/11/2016
16:17
Thinking on Denso didn't one of the overlords say recently they were looking at partnering up to speed up targeted ecosystem development ? Maybe it's that.

edit - from the 2016 year end statement "We will focus MIPS and Ensigma where we have, or can build, differentiated, defendable positions. We will also look to improve the prospects for both business units through partnership arrangements aimed at building the necessary ecosystems and reaching scale more quickly."

Maybe Denso tie up is related to this.

hammerd2
14/11/2016
13:32
We were (still are?)in some Harman automotive kit courtesy of OMAP5 as I recall:
milesy
14/11/2016
13:15
I couldn't spot any use of MIPS or PowerVR by Harmon for in car systems. Did I miss anything?
jamesrowe
14/11/2016
11:29
I thought they already were but can't find mention anywhere. Are they a potential chipmaker though ? More likely to incorporate an off the shelf chip wouldn't you think ?

More grist to the becoming the de facto choice for ADAS solutions mill though.

hammerd2
14/11/2016
11:01
I thought sharegroper meant try for free before taking a licence.

I dunno if they are already a customer.

twatcher
14/11/2016
10:51
Is there any potential buyer out there to whom the same argument that has been used re Apple takeover wouldn't apply ? Vis : why would you want to pay $1bn for a company when you can get what you want for a few million a year ? Denso would appear to want MIPS specifically for their own use in ADAS type chips rather than to license general CPU IP to others to mitigate the purchase costs. It's still Apple or no-one surely, (see above for counter argument)

Always pleased to see something that helps feed the takeover rumour mill though.

hammerd2
14/11/2016
09:38
Try before you buy, perhaps?
SG1

sharegroper
14/11/2016
07:13
DENSO : to Launch Joint Research on Multithreaded CPU Cores with Imagination Technologies
taffy100
13/11/2016
17:25
"Any National team turning up to a hugely important game dressed in pink, deserve a thrashing."



But their accessories were grand! :-)

bukko
12/11/2016
10:19
Lol, good to see you back good grief :-)
Dave

sweenoid
12/11/2016
09:18
... I thought we'd at least come away with the goal posts!!
goodgrief
12/11/2016
08:48
All a bit humiliating...
goodgrief
12/11/2016
07:38
Completely off topic

Any National team turning up to a hugely important game dressed in pink, deserve a thrashing.
WTF were they thinking of?
S

sweenoid
10/11/2016
11:18
Good post Ian
sweenoid
10/11/2016
10:56
Linely article has unsubstantiated claims, and poor research. They assume Apple are using fundamentally the same graphics IP as last years chip, and hence to get the advertised GPU increase have had to up the clock signficantly.

While it is possible that is the case, I can't recall if Apple has EVER used the same graphics IP in different Ax generations.

But regardless, there is no proof of the A10 having a throttling issue compared to the A9, in fact it performs better.

They cite poor performance in Futuremarks's Physics test. Apple A series chips have ALWAYS performed really badly in the physics test. Why Linley doesn't know this, I don't know. But here is an article from the benchmark company from over 3 years ago, attempting to explain why iphone5s shows virtually no improvement over the iphone5 in the physics test. And why Linley are even mentioning it in relation to graphics I have no idea. You'll see in the futuremark article that the physics test is a CPU test, NOT a GPU test.



"In the Physics test, which measures CPU performance, there is little difference between the iPhone 5s and the iPhone 5"

And in relation to GFXbench tests, the results of the batterylife tests on the recent anandtech review show that although frame rate drops from 60 to 50 after around 5 mins, it sustains around 50fps for over 2+ hrs more, right until the battery gives up. The accompanying table shows that the iphone7 acheives a terminal fps which is close to x5 more than all the other phones bar 1 (and it's still x1.6 that one). Crucially, the finally fps is 50% better that in the iphone6s. How does that indicate that the A10 has a heat issue that requires more throttling than the previous chip ?



So to sum up
1) They assume the GPU IP is fundamentally the same as the previous chip, Apple don't have a history of doing that.

2) They use futuremark physic test as evidence that there is a GPU issue, when physics is a CPU test and Apple chips have always showed poor performance in it.

3) They cite some GFXbench test as evidence, but don't bother to cite the one test in that suite which actually tests sustainable FPS and shows that it's final fps is 50% better than the A9 chip in the iphone 6s.

Now maybe Linley have some secret information they aren't revealing, but based on the evidence they cite in the piece I'd file it under, "poorly researched"

twatcher
09/11/2016
19:27
Especially odd since IMG has recently been extolling the virtues of PowerVR for consistent frame rate VR & AR!

Has Apple really messed up by pushing the clock speed too far or has Linley completely misinterpreted the results based on a pre-conception about Apple not going for series 8 but just instead increasing clock speed?? Both options appear unlikely but one of them must be true. Qualcomm could almost have written the article themselves.

jamesrowe
09/11/2016
17:17
Ive had half an hour trying to find any evidence of the A10 throttling and found nothing. Quite the opposite in fact. Which makes the Linley claim all the more curious. Although I am loath to say they've just made it up after the EE times debacle :-)
hammerd2
09/11/2016
17:02
Agree H2. There seemed to be as much speculation as fact in that article - facts for performance but speculation about what was actully being tested and how the benchmarks reflect performance making it difficult to draw any valid conclusions at all. Unusual for Linley as previous articles of theirs that I have read seemed pretty good.
jamesrowe
09/11/2016
14:31
I'd say he has an extra hole too but it seems we're no longer allowed to comment on things without actual personal knowledge. Thankfully I haven't had that sort of experience.

Odd that Linley is claiming throttling on an Apple GPU. It's never happened before and I have seen nothing indicating it is happening now either except in this piece. Their rationale seems to be that the chip doesn't kill every benchmark. The same as they always haven't on certain benchmarks.

Add in that I haven't seen a definitive statement of the IP in the A10 - is it just an overclocked 7600 from the A9 or has it moved on yet again - then maybe their conclusions are a little premature. Wiki list the A9X as a do-deca core so maybe it's a further customised version of this ? Links to analysis greatly appreciated.

hammerd2
09/11/2016
13:29
Nearly there, who would have thought that Trump has a spare!?
goodgrief
09/11/2016
12:39
Linley sees PowerVR bested by Adreno 530. Apple's customisation of PowerVR enables it achieve peak performance which beats all other smartphones out there but cannot maintain this performance for a long time.
rbalakrishnan
09/11/2016
09:46
Dave, 'I hate TRUMP' - but you don't know him. I alwys thought your posts were intellingent but a statement like that is plain stupid. I think we'd better stick to IMG.
colsmith
09/11/2016
09:40
Any disaster for shares in general always seems to be good for IMG for some strange reason so I am just waiting for the share price to rise. :-)
jamesrowe
09/11/2016
09:31
Btw Trump was 150-1 when he initially ran, was probably as worthy of a punt as TMZ/FST are/were (I say TMZ as I still can't bring a quote up for them on my iphone finance app or google finance etc ;))
richardc77
Chat Pages: Latest  1624  1623  1622  1621  1620  1619  1618  1617  1616  1615  1614  1613  Older

Your Recent History

Delayed Upgrade Clock