Intel finds a couple more 11th-gen Core chips, one hits 5.0GHz in laptops Intel has found another pair of 11th-gen Core processors and announced them at Taiwan’s Computex conference, then revealed its 12th-gen “Alder Lake” architecture is “just on the horizon.” The 11th-gen Core i7-1195G7 boasts Intel Iris X graphics, four CPU cores, 12MB cache, base speed of 2.9GHz, and the ability to send one core …
This post has been deleted by its author
Ah CPU fanboyism. How.. sad.
10 GHz base-clock? come back in ten years and maybe someone will be doing it. Maybe. There’s good reasons why CPUs still run at a 2~3 GHz base-clock, despite the big gains in transistor switching speeds. The technologies that allow faster switching also allow increased density (i.e., more cores) instead, which gives better real-world performance than just bumping up the operating clock and having it spin while the chip waits for the DRAM.
I’ve no loyalty to any part vendor, but “nanometres” also means very little in this context. The gate density of Intel’s “10 nm” parts is pretty much the same as TSMC’s “7 nm” parts, and Intel’s “7 nm” parts have a higher density than TSMC’s “5 nm” parts. This is a stupid and confusing situation, but there’s no industry agreement on how to define “feature size”, so it persists. What is true is that Intel squandered its lead on process, has failed to get its 7 nm process up and running fast enough, and now it’s paying the price for that failure. This is good. Companies should not be allowed to fail to deliver without it hurting their bottom line.
You’re right that AMD has eaten Intel’s lunch in the enthusiast market, but as a business that makes lots and lots of money selling laptop chips, Intel’s biggest threat isn’t AMD, but rather ARM-based hybrid designs like Apple’s M1 that provide low TDP without sacrificing single-core performance. It’s hard to believe that Intel’s switch to a hybrid-core design for the 12th generation CPUs was unrelated to Apple dumping Intel in favour of rolling their own hybrid-core ARM chip. AMD has followed suit, and Zen5 will now be a hybrid-core design too.
"The gate density of Intel’s “10 nm” parts is pretty much the same as TSMC’s “7 nm” parts..."
No, it is not. Please stop repeating this nonsense while trying to sound reasonable and neutral. The only people who believe this are Intel partisans, and they believe it because it's what Intel, with nothing else to say, have been saying for a couple of years now. When your competitors are beating you in every segment, with every part, on every metric across the board, what can you do? You tell lies and rely on partisans to repeat them often enough that others start to believe them, because admitting that your products are not competitive with someone else's, usually including those of a previous generation and/or an entire market segment down, just isn't how corporate marketing people operate.
The simple fact is that TSMC 7+ is the best large-scale process in existence today, and it offers density greater than anything Intel have working (5nm and 3nm at scale are coming too, probably before Intel get their 7nm process working). That is why Intel's competitors (plural) are building processors with 2-4x as many cores and 2-10x as much cache as Intel in similar physical form factors, at similar or lower TDP, and at lower price points. If Intel 10nm were just as dense, their products would be competitive by these metrics. They are not, and one doesn't need brand loyalty to see that.
We've seen this before, 20 years ago, when Intel were trying to push Itanic and suppress their own Yamhill rebellion while AMD initiated the obvious revolution they declined to pursue. 10 years later, AMD had joined Itanic in the deadpool, having squandered their huge lead through a series of horrific blunders. Anyone foolish enough to develop brand loyalty (regardless of which) had their ass handed to them at least once in that timeframe. Will history be repeated? I have no idea, but I know that making apologies for Intel's current weak offerings isn't going to be what swings the tide. Intel's 10nm process is not as good as the TSMC processes their competitors are using, and their 7nm process still isn't turning out marketable dies. As a result, Intel's products are not competitive on any axis of price, performance, and power. Claiming otherwise tells a reader much about you, nothing about microprocessors. The smart move is to buy from the winner, not marry them; tomorrow you may well want to buy from someone else, and since you didn't disclose an economic interest in Intel, I know you don't owe them anything.
You are mistaken in thinking that quoted node size and gate density are the same thing. Intel's “10 nm” process yields chips with approximately 90 million transistors per square millimetre. TSMC’s “7 nm” process yields chips with approximately 90 million transistors per square millimetre, and its current 5 nm process raised this to 150M transistors per millimetre. Guess what Intel’s 7 nm process can pack onto a square millimetre... it's the same 150 million transistors. That is the measure that matters.
Microscopy scans of chips on both the Intel 10 and TSMC 7 processes do not show a 30% smaller feature size on the TSMC parts, it’s more like 5-10%, with Intel making more use of complex structures to level up that small difference. Frankly these days, “nanometre” it’s as much of a marketing term as a real measure of process size. You may as well judge CPU architectures by MHz...
If we could see Qualcomm or Apple building their chips at Intel foundries, we’d have a better view of the real differences, but until that happens (an outside chance, as they have claimed that they will now take on third-party jobs), or until Intel hires TSMC to build CPUs (not likely), you can’t make a valid comparison. Process is just one factor in power consumption, just as core-count is largely meaningless once you change between core designs.
But really, nobody who knows what they’re talking about is claiming that Intel’s troubles are because TSMC is at “five” while Intel is still only on “seven”. The reason Intel is screwed is because TSMC got to volume production of its 150MTr/mm process (called “five”) about 18 months before Intel will (calling it “seven”).
i5-10600K will happily do 4.6GHz all-core.
But clock frequency is only part of the story - the IPC of that chip is not up to current AMD standards.
Personally I like high IPC CPUs as they run cooler.
As for the current marketing effort - yes they are behind and so they have to draw attention to the best features of the products, same as AMD had to do when they were behind. It's our job as customers to ignore the marketing smoke & mirrors and seek out the best price/performance for our applications.
BTW - there is no winners tape in this race. It's not even a race... the reality is that most CPUs sold are midrange where the #2 company can still make plenty of money. The leading company gets the privilege of selling high-end CPUs at eye-watering prices, but actually very few people buy them.
100 GHz? Much of what you've written is true, but that's ridiculous.
The highest-clocked general-purpose big-core processors on the market (i.e., leaving aside one-off research projects) are IBM's z15; they operate up to 5.2 GHz; previous generations reached 5.5. They also offer nearly a GB of cache each, though they have only 12 cores. No one is offering a 10 GHz part today nor anything close to that. With exotic cooling, it's possible to reach 8-9 GHz on a single core, but this is not commercially practical. 100 GHz may well be impossible; certainly it is not achievable using today's technology or anything on anyone's roadmap for the next 5 years. Intel were never 10 or even 5 years ahead of the rest of the industry; they may have been 1-2 years ahead 10 years ago, though that would mostly ignore IBM, who despite having incompetent management and marketing have long been and still are leaders in absolute performance.
The relationship among feature size, clock rate, and power dissipation is a harsh and brutal physical truth. Someone will operate an off the shelf processor at 10 GHz someday; I have no doubt of it. Whether anyone will ever offer that as a commercial product is questionable. For a bit less than twice the per-core performance of today's mainstream parts, they'd need to remove at least a kilowatt from an area no more than a couple centimetres on a side, not counting the power needed to operate that cooling mechanism itself. The economics don't work. And 100 GHz... well, I don't see it. People have claimed "physically impossible" before and been wrong, so I won't go that far, but this isn't on anyone's horizon. You'd be needing to remove upwards of 120 kW from the die. Even with 5 or 6 more process shrinks -- there are only 3 in the research pipeline -- you'd still be talking about something like 20-30 kW. Even if it's physically possible it makes no economic sense. The idea that Intel would be offering such a processor if they hadn't squandered their 1-2 year lead over AMD (and perhaps parity with IBM) is silly.
Hey its great to get such speed, B U T... it is also great to buy a sports car that can go 200 MPH.
You see the problem?? not that much money about, not enough places that will allow high speeds, not enough roads that are smooth, and long enough,
need new expensive, PC (home build) with 'everything' to get that speed... would you pay £5000 for that, and 4Gb/s net on top??
If this is Gelsinger's effect as new CEO, it doesn't bode well.
Even if he's not been there long enough to revamp *product*, he needs also to change *attitudes* (inside and outside Intel).
And that doesn't seem to be happening. Is there any sign of any Gelsinger effects at all? It's not as if he's unfamiliar with the company and its challenges.
https://www.theregister.com/2021/02/02/pat_gelsinger_intel_board/
Intel. The ex-86 company.
...I haven't found Intel CPUs interesting in the least.
Intel knows they're not competitive, so they're playing feature game at the moment. (AVX512, anyone?)
But, I'm very interested in those Xe graphics cards (or Xe cores on Xeons if that ever happens) for 3D virtualisation under VMware.
Current-ish Xeon graphics give 3 FPS for relatively simple 3D Printing models which is just unusable. I'm not expecting high-end Tesla level performance, just something reasonable that wouldn't need its own licensing. (*cough*, NVidia)
I know I'm only regurgitating what reviewers have already said about these - but CAVEAT EMPTOR!
A CPU that can go to 5GHz for 5ns then throttle to 2.8GHz because of terrible thermal design is pointless. Buying a laptop with an i9 label does not mean it's better than an i7 because of this.
Thank god some system integrators have recognised the capability of the hardware and fitted an appropriately scaled thermal arrangement.
I don't particularly regard battery life as a concern for my usage; however, Intel vs. AMD in battery consumption land is hugely in favour of the latter too.
What does this mean for prospective buyers who don't know what they are doing? They get an overpriced, poor performing lump they'll get annoyed with, and possibly think of looking at the alternatives instead. (Cough, M1...Or hell, even Chromebook.)
"A CPU that can go to 5GHz for 5ns then throttle to 2.8GHz because of terrible thermal design is pointless."
Because of the physics there's bound to be an upper limit on speed for any given device and package. Heat generation is proportional to switching speed and the density of components changing state as it's source is the energy of the state transition. For any given package it's essentially volumetric, but dissipation (and therefore cooling) is proportional to surface area, so there's order of difference between heating and cooling. Ultimately, for any given package there's a speed and density that can not be cooled sufficiently to survive. At speeds like these a notional option might be a very large thin package, but then you'd hit problems due to propagation delays.
Yes, there's going to be an absolute practical upper limit and we may be getting near it already. But in any case why would I need speed like this in a laptop? For word processing? Oh sorry, I forgot about code bloat.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
