Intel Skylake-U For Laptops Posts Solid Gains In Testing, Especially Graphics (hothardware.com) 104
MojoKid writes: Intel's 6th Generation Skylake family of Core processors has been available for some time now for desktops. However, the mobile variant of Skylake is perhaps Intel's most potent incarnation of the new architecture that has been power-optimized on 14nm technology with a beefier graphics engine for notebooks. In late Q3, Intel started rolling out Skylake-U versions of the chip in a 15 Watt TDP flavor. This is the power envelope that most "ultrabooks" are built with and it's likely to be Intel's highest volume SKU of the processor. The Lenovo Yoga 900 tested here was configured with an Intel Core i7-6500U dual-core processor that also supports Intel HyperThreading for 4 logical processing threads available. Its base frequency is 2.5GHz, but the chip will Turbo Boost to 3GHz and down clocks way down to 500MHz when idle. The chip also has 4MB of shared L3 cache and 512K of L2 and 128K of data cache, total. In the benchmarks, the new Skylake-U mobile chip is about 5 — 10 faster than Intel's previous generation Broadwell platform in CPU-intensive tasks and 20+ percent faster in graphics and gaming, at the same power envelope, likely with better battery life, depending on the device.
Well.... (Score:2)
Optimized code for Intel chips runs really well on Intel benchmarks
Re: Well.... (Score:3, Insightful)
Meh (Score:4, Insightful)
Less than 10% is a "solid improvement" these days?
Re:Meh (Score:5, Insightful)
Sadly, yes. Since AMD does not put any pressure on Intel on the CPU front, 5-10% CPU performance increase per year become the norm.
Re: (Score:1)
The truth is, we're simply past the stage of low hanging fruit and easy gains for traditional silicon tech. That was obvious to me around over a dozen years back when clock frquencies stopped increasing that it was just a matter of time. And yes, yes, you can't compare clock frequencies ONLY to make a speed determination between two completely different CPUs... but it does greatly affect speed just like RPM in a motor. It was then I knew they'd work on other stuff to make up for lack of easy frequency ra
Re: (Score:1)
Re:Meh (Score:5, Informative)
Since AMD does not put any pressure on Intel on the CPU front, 5-10% CPU performance increase per year become the norm.
The Intel of 2015 still has a very solid competitor eating into its profits: the Intel of 2010-13. I am typing this on a 2600K I bought in 2011, and I have no intention of upgrading any time soon. I have went from 8 GB of RAM to 16 GB, from a 128 GB SSD to a 480 GB SSD, and I upgraded my monitor setup. But my desktop processor is still more than twice as fast as my 4300U work laptop, which I never worry about being slow. I wouldn't be that surprised if this processor lasts me until 2020, unless it stops working before then.
Re: (Score:2)
>The Intel of 2015 still has a very solid competitor eating into its profits: the Intel of 2010-13. I am typing this on a 2600K I bought in 2011, and I have no intention of upgrading any time soon. I have went from 8 GB of RAM to 16 GB, from a 128 GB SSD to a 480 GB SSD, and I upgraded my monitor setup. But my desktop processor is still more than twice as fast as my 4300U work laptop, which I never worry about being slow. I wouldn't be that surprised if this processor lasts me until 2020, unless it stops
Re: (Score:2)
Re: Meh (Score:1)
Due to the lack of competition, Intel has a strict conservative policy to not use performance enhancing modifications unless the performance increases by more than 2% for every 1% increase in power consumption. There are still quite a lot of gains to be had in a desktop/server environment, but Intel are not using them so they can reuse the same architecture across their mobile and workstation markets.
Re:Meh (Score:4, Informative)
1985 - 2 MHz
1990 - 33 MHz
1995 - 300 MHz
2000 - 1.2 GHz
2005 - 3.5 GHz
2010 - 3.7 GHz
2015 - 4.0 GHz
At about 3-4 GHz,we reached a point where power leakage made higher frequencies completely impractical. AMD used a more power-thrifty architecture at that time which allowed them to briefly take the CPU lead from Intel, who was completely committed to ramping up clock speed with Prescott. Intel had to abandon netburst and later Intel CPUs were based on the mobile Pentium M, which eschewed high clock speeds to instead concentrate on lower power consumption (it was designed for laptops).
Ever since then, both Intel and AMD have kept clock speed about the same, and focused instead on redesigning CPUs for more efficient parallel processing, increasing the number of cores, and reducing power consumption. Unless there's some earthshattering technological breakthrough, the days of CPU performance increasing 10x every 5 years are over. 5%-10% a year (about 1.5x increase every 5 years, which is about the performance delta between Sandy Bridge and Skylake) is the new norm. Get used to it.
Most of the CPU improvements are instead going into reducing power consumption (Skylake uses about 1/3 to 1/4 the power of Sandy Bridge). My phone is more powerful than the computer I was using in 2000 and lasts 36+ hours on a single charge of a battery smaller than a Kit Kat bar. That is mind-boggling if you think about it.
Re: (Score:2)
The P4 got very high clock rates by having a lot of very short pipeline stages. You can push clock rates a lot higher than they are today if you split your pipelines into more stages, each of which does less. The P4 had some piplelines that were over 30 stages long. A 30-stage pipeline means that if you put an operation in at one end, you get the result 30 cycles later. If you want to saturate the execution units, then you need to find 29 more operations to start that don't depend on the results of the
Re: (Score:2)
Try this:
http://www.amazon.com/gp/produ... [amazon.com]
$800 for a true quad core Intel notebook, Skylake and all:
Intel i5-6300HQ 2.3 GHz Quad-Core
NVIDIA GeForce GTX 960M 4GB GDDR5
8 GB DDR3L / 256 GB Solid-State Drive
15.6-Inch FHD IPS, Wide-Angle, Anti Glare Screen
I own one of these, preordered it the day I saw it on Amazon, it is fast and the screen is very nice.
Re: (Score:1)
AMD is stuck with 28nm fabs.
AMD has much smaller budget.
AMD is rapidly losing GPU market share despite having excellent products (both R9 380 and R9 390 rock).
So yeah, shame on AMD.
Re: (Score:3)
AMD is rapidly losing GPU market share despite having excellent products (both R9 380 and R9 390 rock).
What good is the hardware if the drivers suck and have virtually no settings? It would be so easy to fix and instead they make another wacky huge application with in-built social media, built-in advertisements and other nonsense. It seems to me that AMD don't understand and have never really understood what their buyers want.
Re: (Score:2)
So people can wait 10 years instead of 5 years before they replace their computer?
Sounds good to me.
Re: (Score:1)
You do realize there's more then just pure power improvements in processors nowadays right?
Re: (Score:2)
But hurr durr, Moore's law! Singularity!!!1!
Re: Meh (Score:1)
Humph. At this rate it seems I'll finally retire my Sandy Bridge i7 by 2020
Re: (Score:2)
Re: (Score:2)
If that ever happens you change the thermal paste : wipe the old one and put like 10 cents worth of new one.
You can also undervolt and/or underclock it.
If the motherboard doesn't fail this should be usable or useful for way more than 10 years.
Re: (Score:2)
Re: (Score:3)
In a power-constrained scenario, then yes. Leakage is eating up any power efficiency you get from smaller features. Smaller chips are cheaper to produce - meaning we should get getting them cheaper or more of them if competition was good - but it still wouldn't make a processor do in 10W what last year's processor did with 15W. We're not quite at the end of the road yet but the fat lady is warming up, Intel missing their tick-tock with Kaby Lake is just the first sign.
Re: (Score:1)
The smaller chips aren't really even cheaper to produce anymore. There are reasons that companies aren't jumping down to new nodes as fast as they come out, and it's not just availability.
Re: (Score:3)
That said, i agree Moore's law seems to be on the ropes...
I want to see 11. (Score:4, Funny)
the new Skylake-U mobile chip is about 5 â" 10 faster than Intel's previous generation Broadwell platform in CPU-intensive tasks...
Yeah, well, I'll be impressed when it goes to 11.
"about 5 — 10 faster"? 5-10 WHAT? (Score:3)
Pounds or kilograms? Don't leave us hanging!
If only Moore's Law described the cost and performance of competent editing.
Re: (Score:2)
Re: (Score:1)
Haikus is what you know 'em's.
And this is one too.
Intel needs to simplify (Score:2)
Waiting for secure version without Intel vPro/AMT (Score:5, Interesting)
For some reason I get very nervous with an out of band remote proprietary management system baked into recent Intel chips, which operates below the OS, and has not been independently audited and reviewed by trusted 3rd parties (such as those not associated with mass surveillance).
https://en.wikipedia.org/wiki/... [wikipedia.org]
Note that AMT is also in all Intel chips with vPro:
https://en.wikipedia.org/wiki/... [wikipedia.org]
This posting from the FSF (Free Software Foundation) has a decent writeup about it:
https://fsf.org/blogs/communit... [fsf.org]
It seems that we are now in the age of hardware backdoors.
Maybe AMD which cannot seem to compete with Intel on performance and low-power, can make a niche for itself as a secure (backdoorless) alternative.
These days, I would value my privacy over performance.
Re:Waiting for secure version without Intel vPro/A (Score:4, Insightful)
Yes processors run microcode.
But that is no reason to connect it to an antenna which allows a pc which is turned off to still be able to run wireless remote management commands.
In security one of the most critical consideration is to reduce the attack surface.
Intel vPro/AMT has such a large attack surface, that if we can assume there are no deliberate back doors, it is a safe bet that having it still introduces a wide range of new attack methods against us.
And for what? Just to help make corporate IT's job a bit easier? And remember those extra gates to support it does increase the chip's die size, power consumption, and cost.
Why not have AMT/vPro only in corporate PC's on request, and not have it in anything else.
Re: (Score:1)
And for what? Just to help make corporate IT's job a bit easier? And remember those extra gates to support it does increase the chip's die size, power consumption, and cost.
Why not have AMT/vPro only in corporate PC's on request, and not have it in anything else.
AMT is not enabled on most consumer hardware, however the underlying technology (Intel ME) is present in every chipset since it has other functions like *cough* DRM *cough* and fast-reacting power management. Basically the CPU can't function efficiently without it, but that's kind of beside the point. My point is that there are a lot of insecurities in the most popular architectures that are they are there by malicious (tinfoil hat NSA etc.) or performance-dictated (DMA in everything without IOMMU) reasons.
Re: (Score:2)
For some reason I get very nervous with an out of band remote proprietary management system baked into recent Intel chips, which operates below the OS, and has not been independently audited and reviewed by trusted 3rd parties (such as those not associated with mass surveillance).
What is known isn't good either. All you need is a valid certificate purchasable from any CA in AMTs root list to totally own any system with default configuration if your ever in a position to broadcast DHCP...oh and the computer doesn't even need to be turned on to do it.
It seems that we are now in the age of hardware backdoors.
It can be disabled from bios in some systems and effectively nerf'd in others by disabling I/O virtualization needed to "share" hardware such as your NIC with the operating system.
Maybe AMD which cannot seem to compete with Intel on performance and low-power, can make a niche for itself as a secure (backdoorless) alternative.
These days, I would value my privacy over performance.
It is a good idea to check with vendor to make sure AMT c
AMD Carrizo, ever heard of it? (Score:1)
AMD does fine at low power, at least after Carrizo release:
http://www.anandtech.com/show/... [anandtech.com]
E.g. in HP EliteBook 725 G3, 7+ hours on battery:
http://store.hp.com/us/en/pdp/... [hp.com]
Re: (Score:2)
Maybe AMD which cannot seem to compete with Intel on performance and low-power, can make a niche for itself as a secure (backdoorless) alternative.
Don't bet on it; I recall credible reports of backdoors in AMD chips over a decade ago...
Re: (Score:2)
If only the previous microarch were done as well! Had a friend buy a new Asus laptop (one of those ridiculously light and thin ones, no discrete graphics) with a Broadwell-U i5. About 2 weeks after he got it, it randomly decided to permanently power down all of the sudden while being used (and while connected to the A/C charger). Attempts to make the system do *something at all* (including hours of googling and trying various things) yielded no results; the system instantaneously transitioned from being a f
percent (Score:4, Informative)
>"In the benchmarks, the new Skylake-U mobile chip is about 5 - 10 faster than Intel's previous generation Broadwell platform in CPU-intensive tasks"
That is 5 to 10 *PERCENT* faster. Not a huge whoop. Of course, any improvement is an improvement. (At first I was reading it as "5 to 10 times faster")
Better Graphics (Score:2)
Re: (Score:2)
running at 1080p.
on low settings.
a game that also runs on xbox 360.
a game that you describe as 'new'
sure, it'll run shovelknight etc.. but duh. the point is pretty much that it's as fast as a budget card from 3 years back.
Re: (Score:1)
That's actually impressive for a graphics chip that doesn't require a PCIe card, a dedicated power connection and its own cooling.
Re: (Score:2)
Intel GPUs used to not suck, actually. Seriously, stop laughing!
The GMA945 was actually capable despite not having hardware T&L. What was Intel's fuck up was the original reference speed was 400 MHz, and when they shipped it, every single one was either 133 or 166 MHz, which killed its performance dramatically.
I popped GMABooster on a machine with a GMA945 and cranked it back to reference speeds, and what do you know, despite not having hardware T&L, if your CPU was beefy enough you were hanging wit
Re: (Score:3)
Ain't nobody bitchslapping me.
Not in that thread, given APK thinks HOSTs works when it's constantly demonstrated that it's fucking useless with programs and OSes having their own hardcoded DNS bypasses.
HOSTs is shit as my fiance learned. 5TB of dead data now on APK's ass, and counting.
Real hardware solutions or GTFO, n00b.
The bitch slapping of Khyber (Score:1)
"NOD32 detects a trojan in APK's HOSTS bullshit." - by Khyber (864651) on Saturday August 22, 2015 @01:02PM (#50370415)
VirusTotal & NOD32 shows it COMPLETELY CLEAN
https://www.virustotal.com/en/... [virustotal.com]
&
https://www.virustotal.com/en/... [virustotal.com]
+
http://f.virscan.org/APKHostsF... [virscan.org]
There's only 2 exe's & 5 text files in it - The exe's are proven clean as shown above in the 2 links from VirusTotal, the installer's a SFX rar (keeps it 2mb smaller on download) - that's NO virus!
(Unless YOU know of a way that .txt files are "viruses")
---
"he's tying to get your fucking information." - by Khyber (864651) on Saturday August 22, 2015 @01:02PM (#50370415)
My program doesn't transmit outward ONLY intake of data from 10 reputable sources in the security
Re: (Score:3)
No, you eat your words at 5TB of lost data because your HOSTs doesn't protect against SHIT when a program bypasses HOSTs with its own hardcoded shit.
You complete utter fucking out of date moron, with your insecure out of date almost TWO DECADES OLD USELESS JUNK.
Get the fuck off here and go to Reddit where your bullshit is tolerated by the uneducated masses.
And any AV company espousing your bullshit needs to be avoided as well. It's quite obvious they don't know the first thing about security.
Re: (Score:2)
You seem to be confusing DNS with the IP stack, which is not surprising, as you clearly have no idea what you're doing. I mean, no one would purposefully spam the ever-loving shit out of a website they regularly use, expecting anything positive to come out of it.
Hint: Browsers can make requests to IP addresses. You really should know this.
Re: (Score:2)
Ahh, the delightful sound of a simple yet thorough bitch-slap to the face.
Better graphics is way more than FPS (Score:2)
Not everyone plays games. The graphics matter because it means running every day applications with reasonable responsiveness on laptops which don't have space for a spare card. A 2 pound laptop running both an internal screen and an external desktop at 7680x2160 with a 60Hz refresh and it still gets 6 hours of useful battery life? Fucking fabulous, let me tell you. FPS in games doesn't even make the top ten in 90% of laptop users want lists.
Re: (Score:2)
Yes, at least for the desktop line. Because for my needs, the integrated graphics are more than enough. Very low power requirements and not having to have a bulky card keeping the overall volume of the case minimal are more important.
"i7"-6500U dual-core (Score:3)
mobile i7 = desktop i3, gotta love lack of competition :(
The end is nigh ! (Score:2)
So now, after 18 months in development, a 15% gain is "solid". Not worth changing your laptop over this.
Intel software needs to get their shit together (Score:3)
Oh, sure, the graphics on these chips is worlds better than previous generations, and the power savings is great. BUT, if you can run their drivers without constant crashes and kernel panics it's not really a step forward. Most of the U series laptops and tablets our there are having a myriad of problems - hue shifts, sleep power drain, failure to wake up, driver crash/restarts and - yes - straight up kernel panics/BSOD that require a reboot. It looks like they hires a bunch of amateurs to code this round of drivers.
Re: (Score:1)
So true. Their X79 boards were plagued with UEFI and chipset driver problems, never fixed--they actually dropped their Mobo line completely after that boondoggle. Drivers for NUC have suffered years of neglect. Integrated products still have some kind of interference between USB 3 ports and ethernet, that you can't use a wireless keyboard more than two feet away from the NUC (same was true of the X79, never fixed). I couldn't believe it when I had to connect an extender to my NUC to use a wireless keyboard
Re: (Score:2)
Everyone who bought a MS Surface Pro / Book at launch, especially those who switched from Macs, are collectively losing their shit all over the internet. Now that other Skylake U chips are showing up in other mfrs laptops, it's quickly becoming obvious that it's an Intel problem rather than a MS one.
Why is this worth noting? (Score:1)
"U" About as Fast as a 4-Year Old "M" (Score:2)
Yes, the graphics are likely better, but sadly if you want a powerful dual-core, there isn't much to cheer about.
Re: (Score:2)
Some Numbers: http://www.notebookcheck.net/I... [notebookcheck.net]