Intel's Future Chips: News, Rumours & Reviews

[goldstone77]

INTEL|Posted 22 hours ago
Intel's Coffee Lake compatibility block is broken by a Z170 BIOS hack
By
Jacob Ridley

Intel’s Coffee Lake finally brought six-cores to the mainstream processor market, but if you wanted to make use of them, you’d need to purchase the very latest Z370 chipset. Any old 100- or 200-series motherboard just won’t do - but with a little tinkering behind the scenes, some adventurous modders have rendered Intel’s backwards compatibility block obsolete.

Unfortunately, getting Coffee Lake up and running on older boards is not as easy as a simple motherboard BIOS transfusion. The modders discovered that you need to perform some Intel Management Engine and microcode trickery to get the chip up and running - but it works.

The Baidu user was tweaking an MSI Z170A Gaming Titanium Edition motherboard, intended for Skylake processors at its inception, paired with a Coffee Lake i3-8350K. The system isn’t flawless just yet, however, with PCIe not being recognised by the system whatsoever. Other than that pretty substantial loss, the system apparently ran exactly to spec.

Asus have already pointed towards Intel’s ME as the main blockade to backwards compatibility on 300-series motherboards. Andrew Wu, Asus ROG product manager, believes they could have supported the hex-core chips on at least last generation Z270 chipset had Intel not ‘locked the compatibility’ within their ME.

Intel’s ME already seems to be on the brink of extinction due to some major security flaws in Intel’s unpopular hidden OS. Intel have even been under duress from the Minix OS creator, Andrew Tenenbaum, who deems it unsafe for mass use (it’s only the tiny CPU manufacturer Intel, after all), but now even a few system OEMs have decided to ditch the system entirely - including system super-manufacturer, Dell. It may not be long before other manufacturers follow suit and also ditch the insecure Minix-based OS for the safety of their customers.
With this latest escapade proving that these chips are mostly functional once Intel’s ME limitations have been bypassed, it will be interesting to see how modding is affected now that the latest OEM’s builds have disabled Intel’s backend OS in its entirety - potentially offering a more permanent solution to the system limitations for users.

The next challenge for enterprising modders will be to get Intel’s flagship i7 8700K with all its six-cores and 12-threads running smoothly. Power delivery is supposedly the main limitation to running Intel’s flagship Coffee Lake K-series on older motherboards, so the real test will be to see if these chips are really as power hungry as they are said to be.

This mod is still in early stages, but hopefully this will eventually offer a solution to users who aren’t too hot on the idea of replacing their high-end motherboard in the near future. A pre-packaged BIOS for your Z170/Z270 board of choice would be ideal, but maybe a simplified version of the working solution, in bite-sized chunks, would be easy for users who don’t fiddle with microcode often. Not me, of course, I know everything there is to know about the inner workings of Intel’s Minix ME… I practically wrote the book on the matter.

 

[goldstone77]

Dell Now Shipping Laptops With Intel’s Management Engine Disabled
By Joel Hruska on December 4, 2017 at 4:10 pm

Several years ago, security researchers began sounding the alarm about Intel’s Management Engine (IME). IME support allows the laptop to run an entirely separate OS with what amounts to backdoor access to the entire system. It’s invisible to the main operating system, which means there’s no way for Windows or Linux to tell you anything about what IME is doing on your machine. Intel has defended the system as being both secure and a useful way to perform various remote administration functions. Researchers, however, have found several high-profile bugs that punch holes in that argument. It seems OEMs agree — Dell just became the first major OEM to ship systems that don’t use IME at all.

Dell’s willingness to turn IME off is significant, because normally that’s impossible to do. IME can be disabled once the OS has loaded, according to security researchers, but it can’t be turned off at boot. Also, there’s apparently a risk of bricking your hardware (there’s no information on how common this is). Turning IME off in UEFI in a system configured to use it will result in a boot failure until the system is re-enabled.

Two Linux vendors, System76 and Purism, have already announced they’ll disable IME going forward, but neither is a top-tier manufacturer. Dell, on the other hand, commands a significant percentage of the PC market, and it’s offering three systems without IME enabled: the Dell Latitude 14 Rugged, the Latitude 15 E5570, and the Latitude 12 Rugged Tablet.

The disablement, however, isn’t free. On all three systems, scroll down to Systems Management, and you’ll see the following:

We’ve reached out to Dell to understand what the difference is between “No out-of-band systems management” and “Intel vPro – ME Inoperable, Custom Order” and will update this story when we hear back. Either way, Dell is joining a group of companies that are offering systems without the IME enabled. Oddly, however, the price to disable it varies by system, from $17 to $30 depending on which of the three available options you choose.

It’s not clear if Dell will start offering this option with later CPU families. Right now, the only chips supported are from the Core M and Skylake (Core 6th generation) lines. Since the bug affects all Skylake, Kaby Lake, and Coffee Lake CPUs, we’ll hopefully see updated SKUs take this step as well.

It’s never been clear how many customers use IME in the first place, but finding multiple serious security bugs in the standard suggests Intel’s complete secrecy approach is the wrong fit for this capability. All companies are going to want to obfuscate some details of their security implementations, for obvious reasons, but Intel has faced increasing calls to disclose more about how IME operates so it can be checked for bugs. In the wake of these disclosures, those calls are only going to get louder.

 

[8350rocks]

"The 8700k is 60% faster in Hitman and 55% faster in Ashes"
you are kidding yourself with these figures. there are so many tests which do not agree with this bs.

the same goes for rendering. you're only fooling yourself there.
----

not only cherry picked. but also bogus.

in the same class as the 'glued together cores' spin

your stated logic is based on non representational and questionable results. results which are atypical when compared to the majority of results i have viewed.

According to Hardware Unboxed, the 7700K beats the 1400 by ~47% in Hitman, and ~31.5% in Ashes of the Singularity: Escalation, when both are at 4GHz using DDR4-3200.

That is garbage, in order to be that much faster at those games, the 7700k has to be overclocked by about 25% to get those figures. Which would mean that the 7700k is running 5 GHz while the 1400 is running 4 GHz

 

[juanrga]

He is doing real-world testing. But he is not testing the laptops, but the chips inside. That is the reason why he had to isolate the laptop screen size effect by using the same external monitor for both laptops.

He gives both direct measurement numbers and normalized battery numbers. The number of cells is irrelevant.

He is testing battery life, but wants to plug in a monitor and someone simulate web browsing to as a test for battery life. But he fails to include the information on the batteries being used in each laptop. This test is far from real world testing(which he admits in the quotes I used), when he decided to plug in a monitor, and not plug in the laptop! Batteries for laptop have different size batteries comprised of cells. These batteries range from ~6-12 cells. 12 cells(generally depending on mAh) being twice the capacity of a 6 cell battery, and thus extremely relevant in determining the power consumption on longevity of a battery in this test unless you already know that Intel has a bigger battery in their laptop, and are purposely limiting test results by not making that fact known. Not releasing the battery information used in this testing makes find this type of testing useless for the point he is trying to make.

He is not testing battery life. He is testing efficiency of the SOCs. Aka how much battery power consumes this SoC for doing this work.

12-cell battery has twice the longevity of a 6-cell battery if and only if the cells are the same in both batteries. The reviewer knows the longevity of each battery. It is irrelevant if a given battery achieves its rated longevity with 6 big cells or with 12 smaller cells or with some other combination of cells. What matters is the longevity of the battery. The reviewer uses this longevity for each battery in the normalized-battery graphs posted before.

So he is giving all the information needed and his test shows Intel 8th gen is clearly superior and this will be evidently seen in sales.

 

[Gon Freecss]

"The 8700k is 60% faster in Hitman and 55% faster in Ashes"
you are kidding yourself with these figures. there are so many tests which do not agree with this bs.

the same goes for rendering. you're only fooling yourself there.
----

not only cherry picked. but also bogus.

in the same class as the 'glued together cores' spin

your stated logic is based on non representational and questionable results. results which are atypical when compared to the majority of results i have viewed.

According to Hardware Unboxed, the 7700K beats the 1400 by ~47% in Hitman, and ~31.5% in Ashes of the Singularity: Escalation, when both are at 4GHz using DDR4-3200.

That is garbage, in order to be that much faster at those games, the 7700k has to be overclocked by about 25% to get those figures. Which would mean that the 7700k is running 5 GHz while the 1400 is running 4 GHz

 

[juanrga]

Spoiler

I'm so sick of this thread being derailed by the ridiculous IPC discussion. Who cares which CPU has better IPC and whether it's a 10% gap or a 75% gap in IPC? What matters is real world performance, and, more specifically, real world performance per $$$.

------------------------------------------------------------------------------------------
I don't care how strong your "horses" are. Here's what you're telling me:

I can buy 6 *REALLY STRONG* horses in a bundle package for $400 and hook them up to a compatible wagon that costs $150-200. If those 6 horses pulling that wagon can travel 10 miles in 60 minutes then the figure that matters is the fact that I'm able to go 10 miles in 60 minutes and it will cost me $550-600 to do so.

OR

I can buy 8 weaker horses in a bundle package from a different dealer for $300 and hook them up to a compatible wagon that costs $100-150. If those 8 horses pulling that wagon can travel 9 miles in 60 minutes then the figure that matters is the fact that I'm able to go 9 miles in 60 minutes and it will cost me $400-450 to do so.



Sure, the 6 horse package might be faster. But you tell me what's the better buy: transportation at 10 miles per hours that costs me $550 or transportation at 9 miles per hour that costs me $450?
--------------------------------------------------------------------------------------------


Moral of the story: Let's stop talking about who "has stronger horses" and focus on real processor NEWS in this thread. An intelligent custom build doesn't even start with "just give me the fastest CPU" anyway. An intelligent build begins with budgeting and determining purpose of the system, followed by logically allocating a percentage of funds to each part, and THEREFORE most builders will find themselves asking the question like this: I have $200-250 allocated to my CPU, so what CPU gives me the best performance at my price point given current prices RIGHT NOW?"

Inproper analogy. The correct analogy would be "I can sometimes hook up eight slower horses, but sometimes I have to send them over one at a time rather then side by side". Software doesn't always scale, that's why FX was junk.

Oh I know the analogy is flawed. I used that one due in deference to someone else trying to use it earlier. Ah, but you are missing my main point here. It isn't about how strong an individual "horse" is, but rather the discussion should be about how fast you can get your horses and wagon from point A to point B at a given price point. Performance per dollar should be the focus. Intel has corner that market for years and we are now finally starting to see something resembling competition, which is good for everyone, especially those not locked in to a certain brand.

Performance per dollar is a flawed metric when total costs go beyond acquisition costs. Join AMD support forums and read comments of all that angry people that purchased broken RyZens for work and couldn't work because of bugs. They are angry because are loosing money. Similar comments about working CPUs. A faster CPU can be more expensive initially, but does the work in less time, saving working-hours (aka money).

Performance per dollar ignores that price is not a linear function of performance, and so this metric doesn't adequately measures the technology is behind the chip and which its real cost.

Performance per dollar is a relevant metric only when one is on a budget or when the performance is not really needed.

 

[-Fran-]

Performance per dollar is a relevant metric only when one is on a budget or when the performance is not really needed.

Oh, right. So you mean every single person that doesn't really need to have an Intel CPU running at 5Ghz all the time nor even Ryzen at 4Ghz. Which is like... 80% of the people out there? XD

Also, you're spouting nonesense by still arguing with the TechReport laptop comparison with the monitors off. You can't be testing "efficiency" of the SoC when you're not isolating the SoC and just turning off the monitor; you have a whole PCB behind the SoC with different battery types behind it. Plus, it was already proven that it was a problem with HP and not the SoC from AMD. Stop being obtuse, please.

 

[juanrga]

Also, you're spouting nonesense by still arguing with the TechReport laptop comparison with the monitors off. You can't be testing "efficiency" of the SoC when you're not isolating the SoC and just turning off the monitor; you have a whole PCB behind the SoC with different battery types behind it. Plus, it was already proven that it was a problem with HP and not the SoC from AMD. Stop being obtuse, please.

As mentioned a pair of times, the battery element was isolated on the normalized battery results. The rest of the PCB is a minor fraction of the total consumption. So one can do the approximation for the SoC only.

 

[goldstone77]

https://en.wikichip.org/wiki/intel/microarchitectures/ice_lake#All_Ice_Lake_Chips

 

[goldstone77]

10nm, and 10nm+ will not perform as good as 14nm++, but will be more power efficient.

 

[juanrga]

10nm, and 10nm+ will not perform as good as 14nm++, but will be more power efficient.

10nm+ = 14nm++ for transistor performance. The key is in the second graph showing that all 10, 10+ and 10++ have much lower capacitance, which would, in principle allow to get higher frequencies for a given voltage.

 

[Gon Freecss]

Any guesses on how much faster Icelake will be than Skylake?

 

[goldstone77]

Intel Core i7-8700K Coffee Lake 6-Core 3.7 GHz (4.7 GHz Turbo) LGA 1151 (300 Series) 95W BX80684I78700K Desktop Processor Intel UHD Graphics 630 $404.99

Price drop and availability on Newegg

 

[goldstone77]

Intel Core i5-8600K Coffee Lake 6-Core 3.6 GHz (4.3 GHz Turbo) LGA 1151 (300 Series) 95W BX80684I58600K Desktop Processor Intel UHD Graphics 630 $279.99

 

[goldstone77]

Intel Announces Celeron, Pentium Gold And Silver Processors
by Paul Alcorn December 11, 2017 at 8:30 AM

 

[jdwii]

Once Intel actually provides mainstream boards and they have their CPU's at their MSRP i will say Coffee-lake is the winner but until then B350+R5 1600 has my win of the best CPU for the money pick!

Intel Release your mainstream boards and make enough coffee-lake CPU's so they are priced at their MSRP. Once that happens the I5 8400 will be my go to CPU for my future gaming rigs on a budget.

 

[logainofhades]

Pentium gold and silver? Sounds like pokemon titles.

 

[-Fran-]

Pentium gold and silver? Sounds like pokemon titles.

Pentiumon, I choose you!

No?

Cheers!

 

[goldstone77]

Welcome to the Editor Press Center. Check back periodically for photo and caption updates. The following press materials may be downloaded from this site for pre-conference publicity for the IEDM:
http://btbmarketing.com/iedm/

Spoiler

Intel researchers will present a 10nm logic technology platform with excellent transistor and interconnect performance and aggressive design-rule scaling. They demonstrated its versatility by building a 204Mb SRAM having three different types of memory cells: a high-density 0.0312µm2 cell, a low voltage 0.0367µm2 cell, and a high-performance 0.0441µm2 cell. The platform features 3rd-generation FinFETs fabricated with self-aligned quadruple patterning (SAQP) for critical layers, leading to a 7nm fin width at a 34nm pitch, and a 46nm fin height; a 5th-generation high-k metal gate; and 7th-generation strained silicon. There are 12 metal layers of interconnect, with cobalt wires in the lowest two layers that yield a 5-10x improvement in electromigration and a 2x reduction in via resistance. NMOS and PMOS current is 71% and 35% greater, respectively, compared to 14nm FinFET transistors. Metal stacks with four or six workfunctions enable operation at different threshold voltages, and novel self-aligned gate contacts over active gates are employed.

The graph on the left shows that the new platform maintains traditional scaling trends, while the photomicrograph on the right shows the platform’s 12-layer interconnect stack.

(Paper 29.1, “A 10nm High Performance and Low-Power CMOS Technology Featuring 3rd-Generation FinFET Transistors, Self-Aligned Quad Patterning, Contact Over Active Gate and Cobalt Local Interconnects,” C. Auth et al, Intel)

 

[goldstone77]

Once Intel actually provides mainstream boards and they have their CPU's at their MSRP i will say Coffee-lake is the winner but until then B350+R5 1600 has my win of the best CPU for the money pick!

Intel Release your mainstream boards and make enough coffee-lake CPU's so they are priced at their MSRP. Once that happens the I5 8400 will be my go to CPU for my future gaming rigs on a budget.

I agree, the 8400 is very compelling choice even now if you are looking for a gaming rig with budget restrictions.

 

[goldstone77]

Pentium gold and silver? Sounds like pokemon titles.

Pentiumon, I choose you!

No?

Cheers!

Pentiumon, bwahahaha!

 

[adamsleath]

wait for pentium platinum and pentium titanium. bound to be much better.

 

[8350rocks]

wait for pentium platinum and pentium titanium. bound to be much better.

inb4 new architecture called Pitanium that tanks in 24 months and loses billions in R&D.

 

[Isaiah4110]

wait for pentium platinum and pentium titanium. bound to be much better.

Pentium Ruby and Pentium Sapphire?

 

[adamsleath]

wait for diamond pro (II)

but the III version wont be worth it :lol: