Ugh!
I really am not up to typing this out as I am ill. I took receipt of two very nice Vape juices and within a couple of hours I was going down with my first does of the flu for a few years! Typical.
I have been wading through the Internet and YouTube as there are now announcements regarding AMD's Polaris, even brief one of Zen, and nVidia's Pascal coming thick and fast!
Only ... it has been a little odd, I am not the only one to have noticed and I am now wondering whether or not people are actually going to wake up to being fooled as well as being misled?
Basically nVidia announced there cards and everyone went nuts. Except I thought the cards sounded complete shite, many people missed an important point but some did spot it.
In fact I sometimes wonder to people's sanity and their own self proclaimed levels of intelligence as I have seen many arguments with the typical nVidia fanboys stating "The 1080 is 2.5 times the power of the 980!" Or they stated the 980Ti, cannot remember, does not matter ... because no it does not!
I wanted to chuckle each time someone then replied with "Nooo! It is 2.5x the performance in Virtual Reality, did you even WATCH the video?!" LMAO!
But then once corrected they are harping on about the actual 30% increase in power, so wwe have moved 150% to 30% or 2.5 times to 0.33 times the power.
Umm ... did anyone note the insane clocks?
A node shrink and insane clocks? I am sorry but I expected a fair bit more than that to be honest and I thought it would be in excess of 50% at the very least.
Oh yeah, and the node shrink? Absolutely everyone completely forgot about one other thing regarding node shrinks that it seemed nVidia forgot all about too?!
It wont be to scale but as an example 14nm is half the size of 28nm. They pay for the wafers used to grow chips so for the same money they get, theoretically but like I said not quite to scale, double thee amount.
This equates to being half the price of the previous equivalent chips. NOT far more expensive!
Yet there will be these morons that think themselves sooo smart that will fight tooth and nail to defend nVidia while being taken advantage of while their back is turned that they might as well bend over and drop their pants!
These same people were also bitching about AMD's Polaris chip and how it needs to be less than $300?! A chip no one has tested. Jesus Christ, do these people even HEAR themselves?! Do they even read back the utter tripe they type?!
So when I then saw a headline in a report that stated that AMD have just announced $200 for their Polaris 10 card I was ... taken aback and laughing like a drain! Hmm, why do we say that? Drains do not laugh, lol.
So all these morons would finally shut the hell up as the Polaris is not only cheaper than the £300 they harped on about but was a whole $100 cheaper than that? Umm ... welllllll .... noooo...
You see now what they are saying is that the card must be absolute shit to be priced that cheap ... despite the fact that AMD have clearly iterated and re-iterated many times that they were going to release the cards cheap, get VR to the masses and that everyone would be shocked.
Nope! The small minded fanboys who have no life popping off to anyone to get a rise out of them with all their spelling mistakes ad typos still troll around the Internet looking for someone or something they want to disagree with.
I will give you a little info ... I have lost count at the number of times I have wanted to reply and rip the living daylights out of these people and I ... resist. I think, no, no don't. They think they are smart but they are not and will not listen because they actually believe that their deductions are so good that they are both immediate and infallible. You cannot help nor educate people like that just let them go off and blow that same old trumpet.
All they succeed in doing is leaving one hell of a trail on the Internet that can be traced back for bloody eons proving over and over again that they are small minded fools.
But ... Polaris being $200? Wow, not even I expected that!
I am very interested to see how it benchmarks, with what drivers and if they are stable or not, as this might give us an insight into the faster cards from AMD coming between October and early 2017.
I like the idea of them moving the 'X' moniker to after the 'R' and doing away with the '9' and '7' as it was stupid and confused less tech savvy people. Oddly everyone is still expecting a Radeon RX-480X. Maybe a X Series Radeon RXX-480XXX? LMFAO.
I am expecting an RX-490 and two Fury cards and all based on Vega. You were told about Polaris 10 and 11 and have now been shown them and told when. Next is Vega.
I cannot remember now, was there a 470 mentioned?
I had noted that the compute units had dropped in number from the R9-390 which was interesting. I realised that there is a large area for increases in these CU's to 9 increase in number. A very large area and you just know that the top Vega will have more of them than thee Fiji chips? I feel pretty confident about this.
Hmm ... unless that ends up being the Navi chip?
At the end of the day they will both have to space out the increases in power, compute units and chip sizes until it is time to move on from both 14 and 16 nanometres and likely onto 10 nanometres?
Do this too quickly by competing too much and you end up with things being squeezed in without the proper engineering and you know what happens then, right? Things get too hot ... oooh wait a minute?!
Another things that everyone noticed very quickly was that when nVidia made their announcement o their 1070 and 1080 cards the Internet and YouTube was awash with them.
You could not move for all the videos and all the convenient memories of tech journalist fanboys, or on the payroll, for crap about how brilliant the 1070 and 1080 was. I was furious at the over use of the "2.5x" which was a load of shit and quite deliberately misleading.
AMD have an announcement and you would be doing damn well if you heard any of those same journalists so much as break wind!
Luckily and quite fortunately I was far from being the only one that noticed this.
Sorry but I think that being paid off is not only completely immoral it is the very basis for how corruption starts and is so rife. It is both expected and hated in equal measure and many times seemingly by equal amounts?
Now all I have to do is chew on my fingernails until October.
What's all that blood? Lol.
Showing posts with label GPU. Show all posts
Showing posts with label GPU. Show all posts
Sunday, 5 June 2016
Thursday, 17 March 2016
THE APPROACHING NEW ERA
Well it has been awhile since I typed anything out about the up and coming CPUs and GPUs.
But then I really do not like to use every excuse to publish something, like many like to do, and then harp on for the first couple of paragraphs explaining how I had previously stated this with a series of dates that go back a year or two, YAWN!
I always check though and have a read around once or twice a week, though sometimes I leave it a couple of weeks so that news and rumours can build up.
So at the moment it is looking like AMD's Polaris is going to be with us in a couple of months time and thankfully their Zen processor in October, see? I ... SNIP!
This is great news and better still seems to be the buzz about them. The lower power draws for Polaris 10, as they are naming it, sound absolutely great as does that being predicted for Zen.
The latest seems to be projecting an 8 core Zen releasing in October of this year, 2016, but I am still expecting something with a larger core count than 8 and purely because of the 14nm Finfet lithography they are using. This is not merely just a shrink but, as I understand it to be, a stacking too.
So if you can make an 8 core processor on 28nm you can make a 16 on 14nm, theoretically though the size may have to alter slightly.
Now start to stack them?
I would imagine that in the first versions the stacking may only be usable if done once, but later on with some refinement may stack three layers high, with luck. So HBM memory could go on the third layer, which I would wager is what the Zen+ turns out to be? We will see.
The only thing I worry about is that the consoles will hold back many game developers from utilising this extra power ... unless ... the Nintendo NX, fingers crossed, uses these new Polaris GPUs or that the XBOne can be upgraded, via a plug in graphics card. Microsoft have alluded to as much and lets be honest, Sony had that Vaio Carbon laptop several years ago that had that separate GPU in a box ... umm, thingey! Lol.
As they are meant to be so similar it stands to reason that the Playstation 4 can probably be upgraded graphics chip wise too? Umm, it is Sony!
Of course I am simply going to have to build one!
My only worry is that I build an 8 core Zen PC build only to have them release a 12 or 16 core Zen chip two or three months later?! Grr! Maybe when Zen is finally released we would have found out whether or not a higher core count Zen will release in early 2017?
I also note that in the wake of Polaris things were at first quiet on the nVidia front but then a flurry of information started to surface.
It is times like this that puts credence in both the new info, nVidia, to that previously released, AMD, as one is seen as a knee-jerk reaction to the other releasing details. At least that's one of the techniques I have used to separate the wheat from the chaff, and I do so hate chaff.
It will be such an interesting time to see what Zen, Polaris and Pascal can do and of course added to this will be Intel's own knee-jerk reaction to AMD's releases. So if you think that the second half of 2016 is going to be both mad and exciting, well just wait until the first half of 2017!
For me the Intel chips have done bugger all in the way of progression and added to that they have remained artificially high, some VERY high, with the higher core count chips of 6 and 8 cores. This is so bloody typical of Intel of old and it seemed like forever back in the day to go from a 386 to a 486 CPU, the two chips that preceded the Pentium, which was in essence the 586.
Though I use an AMD A10-7850K right now I will be building with a solely Zen CPU core next time around. The highest core count they release or at least intend to release between October 2016 and probably March 2017, if my nerves can hold out that long.
Saying that I am still interested in what APUs they will be releasing using the Zen cores along with what they will do with the GPU side of the APU. I mean to think of something as could be as powerful as a R9-380 or a Maxwell in an APU? Maybe if it turned out to be better still these could have a dramatic effect on the Virtual Reality headsets also releasing over similar time frames to all these other chips.
I like the idea now, finally, of having a really small form factor PC that can do everything I ask of it and store everything I ask of it and do tasks in a fraction of the time they do now? Oooh.
Also about that time 4K might be becoming commonplace and in so doing mobile telecoms companies and other IP providers will have to up their speeds and stop screwing the public for simply looking at a few webpages and downloading a few small files.
Tech simply has to keep up with tech. But they are such lamentable human beings and completely greedy that I think it will take an uproar from customers and government then forcing them before they do this.
A 4GB download limit was utterly laughable when I first saw them set the figure several years ago, now beyond ridiculous and soon to be 'way out there' ridiculous.
But then I really do not like to use every excuse to publish something, like many like to do, and then harp on for the first couple of paragraphs explaining how I had previously stated this with a series of dates that go back a year or two, YAWN!
I always check though and have a read around once or twice a week, though sometimes I leave it a couple of weeks so that news and rumours can build up.
So at the moment it is looking like AMD's Polaris is going to be with us in a couple of months time and thankfully their Zen processor in October, see? I ... SNIP!
This is great news and better still seems to be the buzz about them. The lower power draws for Polaris 10, as they are naming it, sound absolutely great as does that being predicted for Zen.
The latest seems to be projecting an 8 core Zen releasing in October of this year, 2016, but I am still expecting something with a larger core count than 8 and purely because of the 14nm Finfet lithography they are using. This is not merely just a shrink but, as I understand it to be, a stacking too.
So if you can make an 8 core processor on 28nm you can make a 16 on 14nm, theoretically though the size may have to alter slightly.
Now start to stack them?
I would imagine that in the first versions the stacking may only be usable if done once, but later on with some refinement may stack three layers high, with luck. So HBM memory could go on the third layer, which I would wager is what the Zen+ turns out to be? We will see.
The only thing I worry about is that the consoles will hold back many game developers from utilising this extra power ... unless ... the Nintendo NX, fingers crossed, uses these new Polaris GPUs or that the XBOne can be upgraded, via a plug in graphics card. Microsoft have alluded to as much and lets be honest, Sony had that Vaio Carbon laptop several years ago that had that separate GPU in a box ... umm, thingey! Lol.
As they are meant to be so similar it stands to reason that the Playstation 4 can probably be upgraded graphics chip wise too? Umm, it is Sony!
Of course I am simply going to have to build one!
My only worry is that I build an 8 core Zen PC build only to have them release a 12 or 16 core Zen chip two or three months later?! Grr! Maybe when Zen is finally released we would have found out whether or not a higher core count Zen will release in early 2017?
I also note that in the wake of Polaris things were at first quiet on the nVidia front but then a flurry of information started to surface.
It is times like this that puts credence in both the new info, nVidia, to that previously released, AMD, as one is seen as a knee-jerk reaction to the other releasing details. At least that's one of the techniques I have used to separate the wheat from the chaff, and I do so hate chaff.
It will be such an interesting time to see what Zen, Polaris and Pascal can do and of course added to this will be Intel's own knee-jerk reaction to AMD's releases. So if you think that the second half of 2016 is going to be both mad and exciting, well just wait until the first half of 2017!
For me the Intel chips have done bugger all in the way of progression and added to that they have remained artificially high, some VERY high, with the higher core count chips of 6 and 8 cores. This is so bloody typical of Intel of old and it seemed like forever back in the day to go from a 386 to a 486 CPU, the two chips that preceded the Pentium, which was in essence the 586.
Though I use an AMD A10-7850K right now I will be building with a solely Zen CPU core next time around. The highest core count they release or at least intend to release between October 2016 and probably March 2017, if my nerves can hold out that long.
Saying that I am still interested in what APUs they will be releasing using the Zen cores along with what they will do with the GPU side of the APU. I mean to think of something as could be as powerful as a R9-380 or a Maxwell in an APU? Maybe if it turned out to be better still these could have a dramatic effect on the Virtual Reality headsets also releasing over similar time frames to all these other chips.
I like the idea now, finally, of having a really small form factor PC that can do everything I ask of it and store everything I ask of it and do tasks in a fraction of the time they do now? Oooh.
Also about that time 4K might be becoming commonplace and in so doing mobile telecoms companies and other IP providers will have to up their speeds and stop screwing the public for simply looking at a few webpages and downloading a few small files.
Tech simply has to keep up with tech. But they are such lamentable human beings and completely greedy that I think it will take an uproar from customers and government then forcing them before they do this.
A 4GB download limit was utterly laughable when I first saw them set the figure several years ago, now beyond ridiculous and soon to be 'way out there' ridiculous.
Subscribe to:
Posts (Atom)