- Oct 16, 2019
- 2,746
- 6,653
- 146
This is not a good picture Charlie's painting of whatever's left of Intel's roadmap.
For the record, this is also significantly worse than any expectations I ever had as well. Far, far worse.
You have convinced me. From now on I believe the opposite of what he says, because you have been following him and you know.What use is obviously biased material though? I guess it depends on your priorities.
It's why I started following AnandTech in the first place, Anand Shimpi was incredibly professional and even handed. He called it exactly the way it was in an objective fashion. Or today like with Tech Jesus.
If company A does 7 dumb things, and 3 good ones, and company B does 8 good things, and 2 stupid ones, then I want someone that tells it honestly, without a vendetta or a case of the shills.
Write this down here a hundred more times in case abody missed it the first couple of times and then we will start to believe you.
What happened again when you called him out on twitter?
Other than feelings got hurt, I mean.
To everybody who has some emotional investment, anything that states the obvious, in specific language, like Charlie uses will appear as he is hating Intel.
But he is not hating them. He just points blatant incompetency that is inside the company, lately. You may not like it, but Charlie was talking about 10 nm woes since 2015. Like they have turned untrue, right?
He was making stuff up because he hated Intel, right?
Give Charlie credit, when its due. He was right way more often on Intel topic, than not right, at all.
Oh, and last thing. Charlie on topic of Intel has way more credibility than some random forum user that claims to have spoken with Asian Suppliers.
lol not trying to win hearts and minds here. I’ll enjoy you all asking for your good shepherd Charlie to come up with more fud with icelake sp ships next year. You know to keep you guys gaslit 🤡
Let me ask you. Why do you say it’s likely true? Do you have anything to back up that claim?The part about Icelake-SP being delayed into 2021 is likely true. They could of course pull a Cascade Lake-AP and announce that it's "shipping" but not actually sell it.
Let me ask you. Why do you say it’s likely true? Do you have anything to back up that claim?
he denied it. Maybe you don’t care about journalistic integrity from your moms basement, in between chugs of Mountain Dew, bites of KFC and sweaty dota rounds but there need to be some standards for integrity. Clearly you don’t care but others might. It’s certainly not a thought crime to point it out lol.
Charlie is a troll plain and simple. I don’t care that he blocked me, he has nothing useful to share vs my resources. What he does do it make for a lot of intel hate circle jerking in this forum though - including from your sorry self
lol not trying to win hearts and minds here. I’ll enjoy you all asking for your good shepherd Charlie to come up with more fud with icelake sp ships next year. You know to keep you guys gaslit 🤡
Lol you really are special. Do you have many friends or just live to post snarky replies. I haven’t seen you say one intelligent value added comment in this thread. Speaking about kindergarten, did you even pass?I'm not sure who are you talking about, but you accidentally quoted my postAlso if that's really your intellectual level of arguing on a tech forum - making wild assumptions of someone's living, family, snack- and beverage consumption, God knows why - then you've said more than enough, you deserve my quick and effective ignore here
You speak just like every little boy I knew back then in kindergarten
"I'm disputing every single claim with cocky personal attacks - but please believe me, I don't care at all!!!!11"
![]()
![]()
Let's be rational here. If IceLake-SP is supposed to ship at healthy volumes (read: unlike Cascade Lake-AP where nobody really wanted the things anyway), the Big Cloud Boys should be getting their early samples now or within the next three months. Google, Amazon, MSFT, etc. Then they'll get chips in volume in maybe March/April or so . . . well before the rest of the market.
I haven't heard one thing from Intel about IceLake-SP sampling yet. Milan was sampling in October.
Why is it obvious? Not trying to be pedantic but want to see if you have anything beyond general assertions and the word of CharlieBecause it's obvious the yields are bad, and I also believe Charlie on that specific point.
I think Intel can salvage 10 nm for products that are very small dies, but to even sell like client parts at a real volume it will need to be chiplets.
icelake sp has been sampling for months.
Link? Intel hasn't announced it. Meanwhile:
![]()
AMD Dishes on Zen 3 and Zen 4 Architecture, Milan and Genoa Roadmap
A now-removed presentation from an HPC conference shines the light on AMD's roadmap and change to its next-gen processors.www.tomshardware.com
To me its simple. And I said this before.Intel is a victim of their 14 nm super-refinement.
By refining 14 nm so much they got much better clocks out of it along with better efficiency but that had the cost of eating into 10 nm's advantages: not so much on the power efficiency and a lot more on clocks side of things. They did this thinking they would be able to get 10 nm working without problems, and thus it wouldn't be a big deal to have 14 nm eat up some of 10 nm advantages. Also, by having such a big performance advantage over AMD's CPUs, even if something were to go wrong, it shouldn't be bad.
Guess what: getting 10 nm to work has been a nightmare and, all of a sudden, those super-refined 14 nm parts are making the lower clocks they have been getting from 10 nm pretty much useless, as far as upgrades to current CPUs go, and not even the better efficiency is saving them because the clocks they have been getting with 10 nm have been THAT BAD, and even 10 nm's much better IPC (supposedly) isn't enough to counter this.
Guess what: AMD made a 52% IPC increase to their CPUs when they launched Ryzen, and have been increasing that number with subsequent Ryzen launches and ... there goes Intel's performance advantage ...
So now Intel has an architecture that has better IPC and efficiency but significantly lower clocks and they can't sell it as an upgrade because ... well ... it's a downgrade in everything except efficiency ... and to top it off, they are having REALLY BAD yields with it. And since AMD is now breathing hard on their necks, they don't have margin for error, which is what 10 nm has been up to now.
This is how i understood it.
This was May 2019 and they said "in the coming months". Is that years ? where are they ? Rome is here and has annihilated anything they have posted about,They said it at their May investor day but if you don’t believe me:
I'm not sure who are you talking about, but you accidentally quoted my postAlso if that's really your intellectual level of arguing on a tech forum - making wild assumptions of someone's living, family, snack- and beverage consumption, God knows why - then you've said more than enough, you deserve my quick and effective ignore here
You speak just like every little boy I knew back then in kindergarten
"I'm disputing every single claim with cocky personal attacks - but please believe me, I don't care at all, I'm just declaring my superiority!!!!11"
![]()
To moderators: sorry for off topic. No need for a warning, it won't go any further, the ignore is in place.
Lol you really are special. Do you have many friends or just live to post snarky replies. I haven’t seen you say one intelligent value added comment in this thread. Speaking about kindergarten, did you even pass?
To me its simple. And I said this before.
Currently, they have nothing to compete with Rome, or even close.
They have nothing to compete with HEDT, currently 3960x and 3970x
They have nothing to compete with desktop. The 3950x specifically, let alone the 3900x and the 3800x and the 3700x, etc.. Down the the 3600.
These are facts (aside from gaming the 9900 series wins by a little)
So, regardless of what you think of Charlie, Intel is sitting in the basement. When will this change ? Not soon from all we see. So even if you hate him, and even if he is anti-Intel, where is he wrong (essentially, not to nitpick)
Ok so wait till 2h next year. The one key thing Rome has not annhilated: intelThis was May 2019 and they said "in the coming months". Is that years ? where are they ? Rome is here and has annihilated anything they have posted about,
I don't count vaporware.
They beat the best Intel servers by over100% almost. Do I need to link the benchmarks ? Hnag on a sec... I will.Ok so wait till 2h next year. The one key thing Rome has not annhilated: intel
for a moderator you really wear red on your sleeve.
They said it at their May investor day but if you don’t believe me:
So now Intel has an architecture that has better IPC and efficiency but significantly lower clocks and they can't sell it as an upgrade because ... well ... it's a downgrade in everything except efficiency ... and to top it off, they are having REALLY BAD yields with it. And since AMD is now breathing hard on their necks, they don't have margin for error, which is what 10 nm has been up to now.
Yeah, not great at all actually.Ice Lake does definitely not have better efficiency (than Zen 2).