-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
Another day, another broken Intel process node: Intel announces 7nm will be delayed
- Thread starter Grim5
- Start date
Soldato
- Joined
- 19 Feb 2011
- Posts
- 5,849
Dont think they are "Dumping" their fabs, just think they are struggling with their processes in their fabs, and maybe looking at switching them to other working fabs and altering the process accordingly no? Kinda smart move because if they port across to say TSMC and the chips are really good, they can say "We was confident our design was always going to be great no matter where it was made" however if its a dud they can blame "Translation to non native process" as being a key contributor lol.
Everything will be thought about with multiple outcomes planned for i am sure, Intel as we know are masters of Spin, although though recently a lot of their spin has been terrible, transparent and blatantly biased but thats what you get when you hire morons like Shrout to PR for you.
Could see though them spinning off their main CPU's to competitors for a good while, as if they fill the capacity in say TSCM, with their financial might they could offer the Fabs a more lucrative deal than AMD can which would also hamstring AMD as well who as we know are fabless. That way they can use their own fabs to meet demand on their current stack and respin newer stuff to older nodes for other demands...
I don’t know how intel can outsource their fab process to TSMC. There must be some issues with IP that both companies want to protect. TSMC is happy to partner with AMD as AMD has no fabrication processes and therefore there is no risk of any information about their own production falling into the hands of intel/Samsung etc.
for intel to outsource it will hurt their margin and potentially spell the end of the fabrication pioneering and become reliant on external resources and it is not something I think intel will be looking to do.
for intel to outsource it will hurt their margin and potentially spell the end of the fabrication pioneering and become reliant on external resources and it is not something I think intel will be looking to do.
Soldato
- Joined
- 22 Nov 2018
- Posts
- 2,840
AMD started work on Zen in 2012 when Jim Keller joined. Considering Ryzen didn't release until 2017, that's a 5 year development time.
So based on that, if Intel were on the ball and started developing something new when Zen released in 2017, the 5 year development time won't be up until 2022.
Obviously it won't take Intel 5 years because they normally save time by recycling old architecture from decades ago, hence all the security holes.
Associate
- Joined
- 13 Jun 2016
- Posts
- 1,519
- Location
- UK
I have invested in AMD. The market hasn't taken the news well - AMD up nearly 12% today, Intel down just over 14%. That should apply some more pressure to the C-suite at Intel.
Soldato
- Joined
- 31 May 2009
- Posts
- 21,457
lol
Unfortunately intel used their 7nm process to shrink the mic to heard you, and it failed, so they missed the orchestral sorrow.
It will be hard for Intel to outsource to TSMC in the short-medium term, AMD reportedly has TSMC's extra capacity on lock ever since TSMC had to stop producing Huawei's chips. It's rumoured that Nvidia cannot even get enough 7nm capacity at TSMC to produce all of their Ampere GPUs on the same node, hence why they will be turning to Samsung for some of their new GPUs. Intel are either going to have to pony up for extra capacity or they will have to take whatever scraps they can get.
I don’t even think it is a matter of capacity. Intel cannot outsource for the fact that they will loose the competitive edge. If they outsource to TSMC how can they still justify their current advanced fabrication. They will retain some level of production but it will be low grade. That means it will leave only TSMC in the game and they can hold everyone to ransom.
Yeah. Massive demand for TMSC's 7nm and 5nm production lines. Apple, Qualcomm and AMD have that locked on as massive customers. These are long-term TSMC customers with massive orders for those product lines who plan to be with TSMC for the foreseeable future. Intel will have to compete on all of these in order to be able to win that.
What's more likely is that Intel will work with Samsung, as they've done this before as well.
Intel have to be seriously looking at porting at least some of their new designs to TSMC, that in itself isn't a given mainly due to capacity constraints at TSMC for the near future at least.
Intel are well and truly stuck. They don't have their internal 10nm process where they need it yet and it's, arguably, obsolete before it's even fully viable - except for some of the SKU's near the bottom of their stack. They look lost but they still have their marketing dept' which only be able to save them so long. Yes they have a massive revenue stream and still a fair amount in the bank, but I wouldn't see them really start to hemorrhage money Comp'ing people not to switch to the competition - and that means both Arm and AMD now!
In this state they could easily take 3-5 years to fully sort their issues if not more, and they've had over 2 already, but looking at what is coming out of Intel it looks like they have only just started to be honest, and I don't mean fully honest either, about the delays with their manufacturing.
I'm sure Ryan is on the case making yet more misleading benchmarks, it's his job description after all, with which to hoodwink most of the tech-press. They need to stop with the misleading bullcrap and focus on the engineering! I'm sure they are in this predicament mainly because the marketing boys took over a while back. It took a while for the ship to turn as Intel had such a lead, AMD were dead in the water, and all Intel could do was drip feed 3-5% increments per generation and segment the market up the wazzu..
You reap what you sow.
Intel are well and truly stuck. They don't have their internal 10nm process where they need it yet and it's, arguably, obsolete before it's even fully viable - except for some of the SKU's near the bottom of their stack. They look lost but they still have their marketing dept' which only be able to save them so long. Yes they have a massive revenue stream and still a fair amount in the bank, but I wouldn't see them really start to hemorrhage money Comp'ing people not to switch to the competition - and that means both Arm and AMD now!
In this state they could easily take 3-5 years to fully sort their issues if not more, and they've had over 2 already, but looking at what is coming out of Intel it looks like they have only just started to be honest, and I don't mean fully honest either, about the delays with their manufacturing.
I'm sure Ryan is on the case making yet more misleading benchmarks, it's his job description after all, with which to hoodwink most of the tech-press. They need to stop with the misleading bullcrap and focus on the engineering! I'm sure they are in this predicament mainly because the marketing boys took over a while back. It took a while for the ship to turn as Intel had such a lead, AMD were dead in the water, and all Intel could do was drip feed 3-5% increments per generation and segment the market up the wazzu..
You reap what you sow.
Associate
- Joined
- 23 Jul 2006
- Posts
- 944
It's almost ironic that the 14nm++++++++++ jokes are going to become true. The fact they aren't going to have 7nm fro DC until half way through 2023 at the earliest, realistically puts any desktop parts into 2024, and 10nm might as well not exist either.
How are Intel going to market their way out of stagnation for the next 3-4 years? I am sure they'll spend a lot of money pulling the wool over peoples eyes, or trying to 'convince' OEM's that they should stick with them, this smells like the dawn of the AMD Athlon 64 all over again, where INtel will simply have no answer other than MOAR!!!! clock speed.
How are Intel going to market their way out of stagnation for the next 3-4 years? I am sure they'll spend a lot of money pulling the wool over peoples eyes, or trying to 'convince' OEM's that they should stick with them, this smells like the dawn of the AMD Athlon 64 all over again, where INtel will simply have no answer other than MOAR!!!! clock speed.
What else is left for 14nm performance? I was shocked at how much performance they got out of these 10 series dies. However, they barely overclock because they're pushed to the limit and they've shaved the die for temps. Assuming 10nm doesn't suddenly pop up, what's the next move?
Caporegime
- Joined
- 9 Nov 2009
- Posts
- 25,366
- Location
- Planet Earth
While they'll continue to get good and possibly increased performance per core from 14nm, the die sizes (creating yield, manufacturability and cost issues) will leave AMD with an easy goal to score. If Intel are forced to keep desktop parts on 14nm for longer than expected then AMD can increase the core counts of their SKU's easily (with the *600 lineup being made up of 8 rather than 6 cores. Intel can't compete against that at all. because the prices of those chips, and availability would be impossible to solve.