Intel's latest 11th Gen Processor Brings 5.0GHz Speeds To Thin and Light Laptops (theverge.com) 51
Intel made a splash earlier in May with the launch of its first 11th Gen Tiger Lake H-series processors for more powerful laptops, but at Computex 2021, the company is also announcing a pair of new U-series chips -- one of which marks the first 5.0GHz clock speed for the company's U-series lineup of lower voltage chips. From a report: Specifically, Intel is announcing the Core i7-1195G7 -- its new top of the line chip in the U-series range -- and the Core i5-1155G7, which takes the crown of Intel's most powerful Core i5-level chip, too. Like the original 11th Gen U-series chips, the new chips operate in the 12W to 28W range. Both new chips are four core / eight thread configurations, and feature Intel's Iris Xe integrated graphics (the Core i7-1195G7 comes with 96 EUs, while the Core i5-1155G7 has 80 EUs.)
The Core i7-1195G7 features a base clock speed of 2.9GHz, but cranks up to a 5.0GHz maximum single core speed using Intel's Turbo Boost Max 3.0 technology. The Core i5-1155G7, on the other hand, has a base clock speed of 2.5GHz and a boosted speed of 4.5GHz. Getting to 5GHz out of the box is a fairly recent development for laptop CPUs, period: Intel's first laptop processor to cross the 5GHz mark arrived in 2019.
The Core i7-1195G7 features a base clock speed of 2.9GHz, but cranks up to a 5.0GHz maximum single core speed using Intel's Turbo Boost Max 3.0 technology. The Core i5-1155G7, on the other hand, has a base clock speed of 2.5GHz and a boosted speed of 4.5GHz. Getting to 5GHz out of the box is a fairly recent development for laptop CPUs, period: Intel's first laptop processor to cross the 5GHz mark arrived in 2019.
4C/8T configuration is a bit disappointing (Score:4, Interesting)
Re: (Score:3)
4/8 has only been standard on laptops in the past few years, but core counts have been a problem with the "other 11th gen" processors using rocket lake on desktop.
I find this announcement odd, since much of the problems with rocket lake has been the regression on core count from the previous gen, which was justified by 10nm not being mature enough to get high clock speeds. If they're still getting 5ghz on the laptop silicon on this 10nm process the question we have to ask is whether 8cores at 5.3ghz is bet
Re: (Score:1)
What's really a shame is that AMD isn't supporting Thunderbolt/USB 4 yet. Their CPUs are better in every way, as are their GPUs. A Flawless Victory, as some called it. Except, no Thunderbolt. USB 3 isn't bad but as someone who likes to keep laptops for a long time I'm not sure I want to invest in one that doesn't have Thunderbolt.
Re: (Score:2)
I'm not sure this is an AMD issue - I have a Gigabyte motherboard with an AM4 socket on it, and it has a Thunderbolt header that I can presumably attach to some kind of cable somewhere to get thunderbolt ports on the back of the PC.
Now it's probably not integrated into the chipset - that would be on AMD to get done, but apparently it's plausible to add a controller as an on-board PCI bridge in the Zen3 / PCIe 4.0 architecture they've been shipping.
Re: (Score:1)
I will not call 4/8 as "standard" on laptops for the pass few years. It has been changing at least in the pass 1 or 2 years, since AMD's Zen has been in the market.
I myself am using an AMD 4800U laptop with 8 cores, 16 threads the pass year.
It's just that Intel has difficulty increasing the core counts on both mobile and desktop CPUs.
Re: (Score:2)
Yeah, I wouldn't disagree with that, but the context of a standard config from intel is what stands out. Intel did 4/8 for their typical desktop processor from 2008-2018 from the introduction of the i7, not ending until coffee lake. At the same time the typical laptop processor was a 2/4. You could pay extra for a 4/8, but it also typically had a 20W higher TDP and wouldn't be practical for the 13-14in standard laptop, but targeted more at the mobile workstation class.
It's most certainly not a coincidenc
Re:4C/8T configuration is a bit disappointing (Score:5, Insightful)
I think those of us who carry thin and light laptops (aka ultrabooks) are more about portability than actual number crunching power. After all, that was the appeal of laptops like the Macbook Air - it was super thin and light and you can slip it in your bag and carry it around all day without your arm hurting. It would have limited connectivity due to size (you were lucky if you got more than a couple of USB ports). And since the battery is limited, you'd want something not so high powered so you can work several hours on it.
I'm sure they could do 8 core units, but for the market, it probably wasn't particularly useful.
The people who need an 6 or 8 core laptop will probably need more in the way of RAM, ports, storage, screen, etc.
Re: (Score:2)
They're likely also chasing the M1/ARM at this point, not just their existing lines, with one big caveat. The M1 effectively doubled the thin-and-light MacBook Air's multi-core performance while extending battery time and eliminating a fan. There was likely a "why can't we do this" session in the C-suite. The caveat is when Windows on ARM is mature enough for a majority of apps and games, then Intel would need to worry much more about which chips will be going into PC rigs.
Re: (Score:2)
Nevertheless my MacBook Air is > 90% of its usage time plugged in.
Re: (Score:2)
Processor Bugs? (Score:4, Insightful)
I still have yet to see a webpage or document that lists which processor flaws (spectre, meltdown, other sidechannel attacks, etc.) are fixed with O.S. patches, microcode updates, or in hardware for each processor.
While we're at it, how about some ECC ram support across all product lines?
Re: (Score:1)
Buy Xeon. That's your only real option in Intelland.
Re: (Score:3)
Not good enough. Even Xeons refuse to make use of ECC memory if the chipset that they are paired with isn't "approved" for its use. This despite the little detail that the IMC is on the CPU, not the chipset, and it's the exact same IMC they use on non-xeon.
Re: (Score:1)
Actually 5GHz? (Score:3)
How fast is it when all the spectre patches are applied? Intel's likes to fudge the truth so I know that we'll never get a real answer out of them.
Re: (Score:2)
Re: (Score:3)
Re: (Score:2)
All the manufacturers need spectre patches. New Intel chips are not particularly hurt more by that than others. Older Intel and Apple chips though...
Re: (Score:2)
You're asking the wrong question. The question you really want to know the answer to is: "How many of those 5 billion cycles per second are being wasted on mitigating Spectre and Meltdown attack vectors?"
Thermal Throttling Kills the U-Series (Score:5, Informative)
Re: (Score:3)
Re: (Score:2)
Is this as good as or the same as being able to run at a constant 5ghz and do some real number-crunching on your ultrabook? No,
Re: Thermal Throttling Kills the U-Series (Score:3)
I'ma go on a limb here and claim that essentially everyone who's in their 40s or 50s now, and "into computers", in particular chip engineers designing electronics for a living, have seen Knight Rider in the '80s and loved it... *shrug*
Re: (Score:3)
It's not the Thermal Throttling that kills it, it's the Power Management throttling. Often times, it will throttle the processor because it is using too many watts, and it doesn't even get enough for thermal throttling.
You can actually turn off Power Management throttling, then thermal throttling will kick in, and keeping the processor cool becomes the issue.
Re: (Score:2)
In a test I managed to stabilize the clock by turning off the turbo boost in the BIOS and I was even able to play a game at minimum settings, but then the processor I paid for 4.9GHz achieved a maximum sustained 1.8GHz. I returned the notebook and dug around until I g
Re: (Score:2)
This is why you don't buy based on clock frequency, which has been a bad indicator for performance since the Pentium 3 days, but rather on review benchmarks where they give the thing actual work to do, and compare it against other systems that have run the same benchmark.
All the "boost clock" in the world doesn't make any difference if you never actually get there and stay there. Sure, there can be two thin and light notebooks with the same processor in them, but if one does 20% more work in the same test
Meh (Score:4, Interesting)
Re: (Score:2)
i7 950 to i5 10400 is a big difference after a decade. :O
Re: (Score:2)
Re: (Score:2)
Well, web browsing got really slow like on bloated web sites. :(
Re: (Score:2)
What do you think is 'artificially "locked" into Windows 10', unless you're talking about wanting to stick with an unsupported previous version of Windows that modern hardware isn't getting drivers written for, or some ARM-based hardware? Sure seems like Linux runs great on all the latest hardware from both Intel and AMD.
Guess what, you probably can't find Windows 2000 drivers for brand new hardware either. Time moves on, and companies don't have endless resources to write and compile drivers for every OS
Re: (Score:2)
Re: (Score:2)
No, not a bad day, there's just a whole lot of entitlement going on out there where people would like large organizations to conform to their snowflake needs. If you choose to continue to run Win7 knowing full well the limitations and decide to go that way, good on you. You're doing it better than most who don't analyze all the important aspects and then get pissed when something doesn't work.
However, characterizing hardware that came out last month as "locked into" a specific operating system isn't remot
Re: (Score:2)
Re: (Score:2)
Let me be clear: I don't feel you are acting entitled - as I said before, you are making a choice based on knowing how the industry works. I was referring to others that turn up around here or on Reddit who don't, or actively refuse to accept the way the industry works.
And until recently, I was running an i7-5930K with dual channel DDR4 and upgraded to an AMD Ryzen 9 5900X and there is a profound increase in performance in applications that actually make use of CPU. Feeling "snappy" in Windows of any ver
They actually, kinda work (Score:5, Interesting)
I like the "ultrathin" Intel processors for their market. But this particular one concerns me.
I have a ultra thin Samsung Chromebook with an i5 processor. And it actually works nicely without a fan. Not the 11th generation of course ( i5-10210U). For office work and web browsing the CPU is more than powerful enough, but of course sustained performance is not possible. Even a full metal body cannot dissipate heat fast enough to run extended times at turbo.
This new Core i7-1195G7 seems to require 28-35 W TDP. That is pretty much impossible to go without a fan, and that kills the purpose. For a comparison, the i5 I have can run between 10-25W, and can still get hot to touch.
Ho-hum (Score:2)
Re: (Score:2)
Oh, we've been able to crank a clock rate for decades. The question is what useful work per clock you're getting out of it. And Pentium 4 was garbage at high clock rates, which is why the Core-series stepped way back on the clock rates, yet had far more performance.
Turns out if you don't have a 30+ stage execution pipeline and your branch prediction fails, you don't have to spend a shit-ton of clock cycles to clear that pipeline out in order to get the right instruction in and continue. The "Prescott" Pe
What does the Average user do with an i7? (Score:1)
just FYI... (Score:1)
Re: (Score:2)
This. I have a i5-1035G1 in my current laptop and it is great for non-intensive work like writing text, browsing etc. But, try something trivial like playing minecraft and it overheats and throttles. In theory, it is more powerful than my old Asus Zenbook from 2014 (which broke) but in practice it is much worse since I can't use it for the same things. There are small benefits such as browsers starting faster, short compiles taking less time, the fan is off while writing text etc but it just can't handle a
Re: (Score:2)
They follow *minimum* recommendations from Intel. Lenovo has been spectacularly bad about this kind of thing starting with the X1 Carbon, like 10 years ago - I remember getting two units in on a demo program, one with an i5 and another with an i7. I benchmarked the hell out of both of them, and the i5 was basically always faster over any meaningful amount of work, because the i7 was throttling.
In the "ultraportable" market, save some money and get better performance - buy the midrange processor spec becau
Re: (Score:2)
(I returned the notebook to the manufacturer when I realized my mistake and then bought one from another manufacturer without a dedicated video card and with a Ryzen 4800U processor, this one works)
Overheating pile of.. (Score:1)
I've had to undervolt and cut the turbo multipliers anyways so I think 5 GHz that is usable is just a daydream..