Intel's Future Chips: News, Rumours & Reviews

[-Fran-]

Back in 2014--2015 Intel planed a 10nm Cannonlake with 4/6/8 cores for mainstream. Then found big difficulties with the 14nm node, which delayed the development of the 10nm node, forced a change on the roadmaps and introduced the new refresh cycles, with 4-core Kabylake and 6-core CoffeLake.

Scaling down to the smallest nodes is getting more and more difficult. Glofo canceled its 14nm node. TSMC and Samsung avoided the problems of dealing with a 14nm node by introducing hybrid '14nm' and '16nm' nodes based in a real 20nm substratum. Glofo canceled also the 10nm node, whereas Samsung and TSMC newest '10nm' are closer to Intel 14nm than to Intel 10nm.

Spoiler

10nm Cannonlake chips start shipping in second half of this year. And 10nm+ Icelake is coming in 2018.

Look, Intel using a useless "transistor / mm^2" metric in a slide! Like measuring cars per squared meter, right?

 

[juanrga]

Back in 2014--2015 Intel planed a 10nm Cannonlake with 4/6/8 cores for mainstream. Then found big difficulties with the 14nm node, which delayed the development of the 10nm node, forced a change on the roadmaps and introduced the new refresh cycles, with 4-core Kabylake and 6-core CoffeLake.

Scaling down to the smallest nodes is getting more and more difficult. Glofo canceled its 14nm node. TSMC and Samsung avoided the problems of dealing with a 14nm node by introducing hybrid '14nm' and '16nm' nodes based in a real 20nm substratum. Glofo canceled also the 10nm node, whereas Samsung and TSMC newest '10nm' are closer to Intel 14nm than to Intel 10nm.

Spoiler

10nm Cannonlake chips start shipping in second half of this year. And 10nm+ Icelake is coming in 2018.

Look, Intel using a useless "transistor / mm^2" metric in a slide! Like measuring cars per squared meter, right?

Right, it continues being an useless metric.

 

[8350rocks]

You are welcome Yuka,

Guys, I have a question because we talk about how close is Ryzen IPC to Broadwell-E chips, But according to Cpu-Z and various others benchmarks (that I have personally ran), my First generation core i7 has a bit better single core performance than Ryzen 7(I'm running @4.5 GHz)

Ryzen first generation seems to cap out about ~4.1-4.2 GHz on great silicon. The gap you are seeing is merely clock speed gap. Especially if you are comparing stock Ryzen to your chip.

A 3.6-3.7 GHz Ryzen is at a 20% clock speed deficit compared to your OC'ed chip, a 4 GHz Ryzen is at a 12% defict in clock speed. So, if the chip is 10% faster per core per clock, then you should see a 2-10% difference in single core from clockspeed.

Intel has a 10--20% IPC gap over Zen.

6% != 10-20% IPC gap.

Thanks for the effort though.

 

[8350rocks]

Back in 2014--2015 Intel planed a 10nm Cannonlake with 4/6/8 cores for mainstream. Then found big difficulties with the 14nm node, which delayed the development of the 10nm node, forced a change on the roadmaps and introduced the new refresh cycles, with 4-core Kabylake and 6-core CoffeLake.

Scaling down to the smallest nodes is getting more and more difficult. Glofo canceled its 14nm node. TSMC and Samsung avoided the problems of dealing with a 14nm node by introducing hybrid '14nm' and '16nm' nodes based in a real 20nm substratum. Glofo canceled also the 10nm node, whereas Samsung and TSMC newest '10nm' are closer to Intel 14nm than to Intel 10nm.

Spoiler

10nm Cannonlake chips start shipping in second half of this year. And 10nm+ Icelake is coming in 2018.

Look, Intel using a useless "transistor / mm^2" metric in a slide! Like measuring cars per squared meter, right?

Right, it continues being an useless metric.

Ultimately density only means the process will be more prone to failures under heat...like the whole 7700k "do not overclock it" fiasco.

In other words...useless.

 

[KirbysHammer]

Back in 2014--2015 Intel planed a 10nm Cannonlake with 4/6/8 cores for mainstream. Then found big difficulties with the 14nm node, which delayed the development of the 10nm node, forced a change on the roadmaps and introduced the new refresh cycles, with 4-core Kabylake and 6-core CoffeLake.

Scaling down to the smallest nodes is getting more and more difficult. Glofo canceled its 14nm node. TSMC and Samsung avoided the problems of dealing with a 14nm node by introducing hybrid '14nm' and '16nm' nodes based in a real 20nm substratum. Glofo canceled also the 10nm node, whereas Samsung and TSMC newest '10nm' are closer to Intel 14nm than to Intel 10nm.

Spoiler

10nm Cannonlake chips start shipping in second half of this year. And 10nm+ Icelake is coming in 2018.

Look, Intel using a useless "transistor / mm^2" metric in a slide! Like measuring cars per squared meter, right?

Right, it continues being an useless metric.

What I take away from that graph is that the nm markings companies give have pretty much zero meaning (heck this has been true for a while) but that;s part of the reason why Intel chips have better IPC despite bth being 14nm

 

[juanrga]

You are welcome Yuka,

Guys, I have a question because we talk about how close is Ryzen IPC to Broadwell-E chips, But according to Cpu-Z and various others benchmarks (that I have personally ran), my First generation core i7 has a bit better single core performance than Ryzen 7(I'm running @4.5 GHz)

Ryzen first generation seems to cap out about ~4.1-4.2 GHz on great silicon. The gap you are seeing is merely clock speed gap. Especially if you are comparing stock Ryzen to your chip.

A 3.6-3.7 GHz Ryzen is at a 20% clock speed deficit compared to your OC'ed chip, a 4 GHz Ryzen is at a 12% defict in clock speed. So, if the chip is 10% faster per core per clock, then you should see a 2-10% difference in single core from clockspeed.

Intel has a 10--20% IPC gap over Zen.

6% != 10-20% IPC gap.

Thanks for the effort though.

It is 10--20% higher for x86 code and about 300% higher for AVX512 code.

 

[juanrga]

Back in 2014--2015 Intel planed a 10nm Cannonlake with 4/6/8 cores for mainstream. Then found big difficulties with the 14nm node, which delayed the development of the 10nm node, forced a change on the roadmaps and introduced the new refresh cycles, with 4-core Kabylake and 6-core CoffeLake.

Scaling down to the smallest nodes is getting more and more difficult. Glofo canceled its 14nm node. TSMC and Samsung avoided the problems of dealing with a 14nm node by introducing hybrid '14nm' and '16nm' nodes based in a real 20nm substratum. Glofo canceled also the 10nm node, whereas Samsung and TSMC newest '10nm' are closer to Intel 14nm than to Intel 10nm.

Spoiler

10nm Cannonlake chips start shipping in second half of this year. And 10nm+ Icelake is coming in 2018.

Look, Intel using a useless "transistor / mm^2" metric in a slide! Like measuring cars per squared meter, right?

Right, it continues being an useless metric.

Ultimately density only means the process will be more prone to failures under heat...like the whole 7700k "do not overclock it" fiasco.

In other words...useless.

What fiasco? I hope you don't refer to invented news from certain media.

 

[goldstone77]

Spoiler

Back in 2014--2015 Intel planed a 10nm Cannonlake with 4/6/8 cores for mainstream. Then found big difficulties with the 14nm node, which delayed the development of the 10nm node, forced a change on the roadmaps and introduced the new refresh cycles, with 4-core Kabylake and 6-core CoffeLake.

Scaling down to the smallest nodes is getting more and more difficult. Glofo canceled its 14nm node. TSMC and Samsung avoided the problems of dealing with a 14nm node by introducing hybrid '14nm' and '16nm' nodes based in a real 20nm substratum. Glofo canceled also the 10nm node, whereas Samsung and TSMC newest '10nm' are closer to Intel 14nm than to Intel 10nm.

Spoiler

10nm Cannonlake chips start shipping in second half of this year. And 10nm+ Icelake is coming in 2018.

Look, Intel using a useless "transistor / mm^2" metric in a slide! Like measuring cars per squared meter, right?

Right, it continues being an useless metric.

Ultimately density only means the process will be more prone to failures under heat...like the whole 7700k "do not overclock it" fiasco.

In other words...useless.

What fiasco? I hope you don't refer to invented news from certain media.

Intel tells Core i7-7700K owners to stop overclocking to avoid high temps
By Paul Lilly May 04, 2017

Some owners of Intel's Core i7-7700 and 7700K processors have been complaining for the past three months of unusual temperature increases. According to user accounts, temps sometimes spike up to 90C, close to the Core i7-7700's maximum 100C threshold, The Register reports. Intel finally offered up a response to the complaints, though the answer is not sitting well with affected users.

In an ongoing thread on Intel's community forums, a spokesman for Intel offered up the following response:

"We appreciate the feedback you have provided, and your patience as we investigated this behavior. The reported behavior of the 7th Generation Intel Core i7-7700K Processor, showing momentary temperature changes from the idle temperature, is normal while completing a task (like opening a browser or an application or a program).

In our internal investigation, we did not observe temperature variation outside of the expected behavior and recommended specifications. For processor specifications, please refer to the Intel Core i7-7700K Processor Product Specifications.

Most motherboard manufacturers offer customizable fan speed control settings that may allow for smoother transition of fan revolutions per minute (rpm). Please consult your motherboard manufacturer’s manual or website for instructions on how to change default fan speed control settings.

We do not recommend running outside the processor specifications, such as by exceeding processor frequency or voltage specifications, or removing of the integrated heat spreader (sometimes called 'de-lidding'). These actions will void the processor warranty."

The TL;DR version is that everything checks out on Intel's end and users should not overclocked their Core i7-7700K processors, which have unlocked multipliers specifically for overclocking.

As you might imagine, Intel's response did not sit well with users. The first response to Intel's post is from a ticked off customer who has sworn off Intel products.

"Three months waiting for Intel to come out with a solution, and now this?! This is all you can say?! We know already what you've just said... You know what, never mind, this would be my last product from Intel," the user wrote.

Another user called Intel's response "BS," noting that "some 7700K [processors] run even hotter than an AMD Bulldozer overclocked at 5GHz."

"I don't even have the major issue like everyone else is having. However, after Intel's response just now, they are not getting another penny out of me. I'm going to sell my Intel stuff and go to Ryzen," yet another user wrote.

To Intel's credit, some users experiencing the issue have admitted to de-lidding their processors. One reason this is done so that cooling solutions can be applied directly to the CPU die, but it's a risky procedure that can result in a dead chip. Removing the IHS can also render certain coolers incompatible, as they were designed with the height of the IHS in mind. Another motivating factor is to replace the stock thermal compound that Intel uses between the die and IHS.

Intel has also never stated that it would warranty processors that have been overclocked or over-volted (though it does offer an overclocking warranty as a separate purchase). However, that isn't the part that has users all riled up. They're ticked because Intel basically shrugged off the temperature spikes as being normal, and telling them to run their unlocked CPUs at stock settings rubs salt in the wound.

 

[juanrga]

Spoiler

Back in 2014--2015 Intel planed a 10nm Cannonlake with 4/6/8 cores for mainstream. Then found big difficulties with the 14nm node, which delayed the development of the 10nm node, forced a change on the roadmaps and introduced the new refresh cycles, with 4-core Kabylake and 6-core CoffeLake.

Scaling down to the smallest nodes is getting more and more difficult. Glofo canceled its 14nm node. TSMC and Samsung avoided the problems of dealing with a 14nm node by introducing hybrid '14nm' and '16nm' nodes based in a real 20nm substratum. Glofo canceled also the 10nm node, whereas Samsung and TSMC newest '10nm' are closer to Intel 14nm than to Intel 10nm.

Spoiler

10nm Cannonlake chips start shipping in second half of this year. And 10nm+ Icelake is coming in 2018.

Look, Intel using a useless "transistor / mm^2" metric in a slide! Like measuring cars per squared meter, right?

Right, it continues being an useless metric.

Ultimately density only means the process will be more prone to failures under heat...like the whole 7700k "do not overclock it" fiasco.

In other words...useless.

What fiasco? I hope you don't refer to invented news from certain media.

Intel tells Core i7-7700K owners to stop overclocking to avoid high temps
By Paul Lilly May 04, 2017

Some owners of Intel's Core i7-7700 and 7700K processors have been complaining for the past three months of unusual temperature increases. According to user accounts, temps sometimes spike up to 90C, close to the Core i7-7700's maximum 100C threshold, The Register reports. Intel finally offered up a response to the complaints, though the answer is not sitting well with affected users.

In an ongoing thread on Intel's community forums, a spokesman for Intel offered up the following response:

"We appreciate the feedback you have provided, and your patience as we investigated this behavior. The reported behavior of the 7th Generation Intel Core i7-7700K Processor, showing momentary temperature changes from the idle temperature, is normal while completing a task (like opening a browser or an application or a program).

In our internal investigation, we did not observe temperature variation outside of the expected behavior and recommended specifications. For processor specifications, please refer to the Intel Core i7-7700K Processor Product Specifications.

Most motherboard manufacturers offer customizable fan speed control settings that may allow for smoother transition of fan revolutions per minute (rpm). Please consult your motherboard manufacturer’s manual or website for instructions on how to change default fan speed control settings.

We do not recommend running outside the processor specifications, such as by exceeding processor frequency or voltage specifications, or removing of the integrated heat spreader (sometimes called 'de-lidding'). These actions will void the processor warranty."

The TL;DR version is that everything checks out on Intel's end and users should not overclocked their Core i7-7700K processors, which have unlocked multipliers specifically for overclocking.

As you might imagine, Intel's response did not sit well with users. The first response to Intel's post is from a ticked off customer who has sworn off Intel products.

"Three months waiting for Intel to come out with a solution, and now this?! This is all you can say?! We know already what you've just said... You know what, never mind, this would be my last product from Intel," the user wrote.

Another user called Intel's response "BS," noting that "some 7700K [processors] run even hotter than an AMD Bulldozer overclocked at 5GHz."

"I don't even have the major issue like everyone else is having. However, after Intel's response just now, they are not getting another penny out of me. I'm going to sell my Intel stuff and go to Ryzen," yet another user wrote.

To Intel's credit, some users experiencing the issue have admitted to de-lidding their processors. One reason this is done so that cooling solutions can be applied directly to the CPU die, but it's a risky procedure that can result in a dead chip. Removing the IHS can also render certain coolers incompatible, as they were designed with the height of the IHS in mind. Another motivating factor is to replace the stock thermal compound that Intel uses between the die and IHS.

Intel has also never stated that it would warranty processors that have been overclocked or over-volted (though it does offer an overclocking warranty as a separate purchase). However, that isn't the part that has users all riled up. They're ticked because Intel basically shrugged off the temperature spikes as being normal, and telling them to run their unlocked CPUs at stock settings rubs salt in the wound.

That is just what I mean. The media invented the news, because he just said that (i) you overclock at your own risk because it voids warrranty and (ii) Intel cannot warrant that chips will work on a given way when overclocked because the chip is running outside specs. Overclocking is silicon lottery.

Of course the same media that invented the news against Intel forgot to mention that AMD makes the same recommendations:

WARNING: AMD processors, including chipsets, CPUs, APUs and GPUs (collectively and
individually “AMD processor”), are intended to be operated only within their associated
specifications and factory settings. Operating your AMD processor outside of official AMD
specifications or outside of factory settings, including but not limited to the conducting of
overclocking (including use of this overclocking software, even if such software has been directly or
indirectly provided by AMD or an entity otherwise affiliated in any way with AMD), may damage
your processor, affect the operation of your processor or the security features therein and/or lead to
other problems, including but not limited to damage to your system components (including your
motherboard and components thereon (e.g., memory)), system instabilities (e.g., data loss and
corrupted images), reduction in system performance, shortened processor, system component and/or
system life, and in extreme cases, total system failure. It is recommended that you save any important
data before using the tool. AMD does not provide support or service for issues or damages related
to use of an AMD processor outside of official AMD specifications or outside of factory settings.
You may also not receive support or service from your board or system manufacturer. Please make
sure you have saved all important data before using this overclocking software. DAMAGES
CAUSED BY USE OF YOUR AMD PROCESSOR OUTSIDE OF OFFICIAL AMD
SPECIFICATIONS OR OUTSIDE OF FACTORY SETTINGS ARE NOT COVERED UNDER
ANY AMD PRODUCT WARRANTY AND MAY NOT BE COVERED BY YOUR BOARD OR
SYSTEM MANUFACTURER’S WARRANTY.

Still no site wrote any news about how AMD tell users to stop overclocking because AMD chips "are intended to be operated only within their associated specifications and factory settings." The usual media bias once again...

 

[-Fran-]

I don't remember Sandy Bridge making the rounds for that specifically. I do remember from Ivy on-wards, when they swapped to toothpaste, they've been making the rounds for OC problems.

Cheers!

 

[juanrga]

I don't remember Sandy Bridge making the rounds for that specifically. I do remember from Ivy on-wards, when they swapped to toothpaste, they've been making the rounds for OC problems.

It doesn't have anything to do with using thermal paste or soldier tech. AMD is making similar claims despite using soldier.

 

[juanrga]

Intel 10nm news

https://newsroom.intel.com/news/intel-technology-manufacturing-day-china/

Not only Intel did show a waffer of 10nm CannonLake chips, but also showed a waffer of 10nm ARM chips!!!

 

[-Fran-]

I don't remember Sandy Bridge making the rounds for that specifically. I do remember from Ivy on-wards, when they swapped to toothpaste, they've been making the rounds for OC problems.

It doesn't have anything to do with using thermal paste or soldier tech. AMD is making similar claims despite using soldier.

We were talking about "making the news" about having OC issues, weren't we?

But meh; I CBA.

 

[KirbysHammer]

Spoiler

Back in 2014--2015 Intel planed a 10nm Cannonlake with 4/6/8 cores for mainstream. Then found big difficulties with the 14nm node, which delayed the development of the 10nm node, forced a change on the roadmaps and introduced the new refresh cycles, with 4-core Kabylake and 6-core CoffeLake.

Scaling down to the smallest nodes is getting more and more difficult. Glofo canceled its 14nm node. TSMC and Samsung avoided the problems of dealing with a 14nm node by introducing hybrid '14nm' and '16nm' nodes based in a real 20nm substratum. Glofo canceled also the 10nm node, whereas Samsung and TSMC newest '10nm' are closer to Intel 14nm than to Intel 10nm.

Spoiler

10nm Cannonlake chips start shipping in second half of this year. And 10nm+ Icelake is coming in 2018.

Look, Intel using a useless "transistor / mm^2" metric in a slide! Like measuring cars per squared meter, right?

Right, it continues being an useless metric.

Ultimately density only means the process will be more prone to failures under heat...like the whole 7700k "do not overclock it" fiasco.

In other words...useless.

What fiasco? I hope you don't refer to invented news from certain media.

Intel tells Core i7-7700K owners to stop overclocking to avoid high temps
By Paul Lilly May 04, 2017

Some owners of Intel's Core i7-7700 and 7700K processors have been complaining for the past three months of unusual temperature increases. According to user accounts, temps sometimes spike up to 90C, close to the Core i7-7700's maximum 100C threshold, The Register reports. Intel finally offered up a response to the complaints, though the answer is not sitting well with affected users.

In an ongoing thread on Intel's community forums, a spokesman for Intel offered up the following response:

"We appreciate the feedback you have provided, and your patience as we investigated this behavior. The reported behavior of the 7th Generation Intel Core i7-7700K Processor, showing momentary temperature changes from the idle temperature, is normal while completing a task (like opening a browser or an application or a program).

In our internal investigation, we did not observe temperature variation outside of the expected behavior and recommended specifications. For processor specifications, please refer to the Intel Core i7-7700K Processor Product Specifications.

Most motherboard manufacturers offer customizable fan speed control settings that may allow for smoother transition of fan revolutions per minute (rpm). Please consult your motherboard manufacturer’s manual or website for instructions on how to change default fan speed control settings.

We do not recommend running outside the processor specifications, such as by exceeding processor frequency or voltage specifications, or removing of the integrated heat spreader (sometimes called 'de-lidding'). These actions will void the processor warranty."

The TL;DR version is that everything checks out on Intel's end and users should not overclocked their Core i7-7700K processors, which have unlocked multipliers specifically for overclocking.

As you might imagine, Intel's response did not sit well with users. The first response to Intel's post is from a ticked off customer who has sworn off Intel products.

"Three months waiting for Intel to come out with a solution, and now this?! This is all you can say?! We know already what you've just said... You know what, never mind, this would be my last product from Intel," the user wrote.

Another user called Intel's response "BS," noting that "some 7700K [processors] run even hotter than an AMD Bulldozer overclocked at 5GHz."

"I don't even have the major issue like everyone else is having. However, after Intel's response just now, they are not getting another penny out of me. I'm going to sell my Intel stuff and go to Ryzen," yet another user wrote.

To Intel's credit, some users experiencing the issue have admitted to de-lidding their processors. One reason this is done so that cooling solutions can be applied directly to the CPU die, but it's a risky procedure that can result in a dead chip. Removing the IHS can also render certain coolers incompatible, as they were designed with the height of the IHS in mind. Another motivating factor is to replace the stock thermal compound that Intel uses between the die and IHS.

Intel has also never stated that it would warranty processors that have been overclocked or over-volted (though it does offer an overclocking warranty as a separate purchase). However, that isn't the part that has users all riled up. They're ticked because Intel basically shrugged off the temperature spikes as being normal, and telling them to run their unlocked CPUs at stock settings rubs salt in the wound.

That is just what I mean. The media invented the news, because he just said that (i) you overclock at your own risk because it voids warrranty and (ii) Intel cannot warrant that chips will work on a given way when overclocked because the chip is running outside specs. Overclocking is silicon lottery.

Of course the same media that invented the news against Intel forgot to mention that AMD makes the same recommendations:

WARNING: AMD processors, including chipsets, CPUs, APUs and GPUs (collectively and
individually “AMD processor”), are intended to be operated only within their associated
specifications and factory settings. Operating your AMD processor outside of official AMD
specifications or outside of factory settings, including but not limited to the conducting of
overclocking (including use of this overclocking software, even if such software has been directly or
indirectly provided by AMD or an entity otherwise affiliated in any way with AMD), may damage
your processor, affect the operation of your processor or the security features therein and/or lead to
other problems, including but not limited to damage to your system components (including your
motherboard and components thereon (e.g., memory)), system instabilities (e.g., data loss and
corrupted images), reduction in system performance, shortened processor, system component and/or
system life, and in extreme cases, total system failure. It is recommended that you save any important
data before using the tool. AMD does not provide support or service for issues or damages related
to use of an AMD processor outside of official AMD specifications or outside of factory settings.
You may also not receive support or service from your board or system manufacturer. Please make
sure you have saved all important data before using this overclocking software. DAMAGES
CAUSED BY USE OF YOUR AMD PROCESSOR OUTSIDE OF OFFICIAL AMD
SPECIFICATIONS OR OUTSIDE OF FACTORY SETTINGS ARE NOT COVERED UNDER
ANY AMD PRODUCT WARRANTY AND MAY NOT BE COVERED BY YOUR BOARD OR
SYSTEM MANUFACTURER’S WARRANTY.

Still no site wrote any news about how AMD tell users to stop overclocking because AMD chips "are intended to be operated only within their associated specifications and factory settings." The usual media bias once again...

I agree, Intel has gotten so much biased press. Unfortunately that is the way the media has gone. It's nearly impossible to get unbiased news even on tech these days.... I'm getting pretty fed up of the fact that some people think that you have to be either an AMD or Intel fanboy. Ryzen no doubt is a good product, but in certain non-gaming tasks even an Intel 4c can equal AMD's 8c chip.

What I'm trying to say is even if you are a productivity user it's not a black and white choice.

 

[goldstone77]

Spoiler

Spoiler

Back in 2014--2015 Intel planed a 10nm Cannonlake with 4/6/8 cores for mainstream. Then found big difficulties with the 14nm node, which delayed the development of the 10nm node, forced a change on the roadmaps and introduced the new refresh cycles, with 4-core Kabylake and 6-core CoffeLake.

Scaling down to the smallest nodes is getting more and more difficult. Glofo canceled its 14nm node. TSMC and Samsung avoided the problems of dealing with a 14nm node by introducing hybrid '14nm' and '16nm' nodes based in a real 20nm substratum. Glofo canceled also the 10nm node, whereas Samsung and TSMC newest '10nm' are closer to Intel 14nm than to Intel 10nm.

Spoiler

10nm Cannonlake chips start shipping in second half of this year. And 10nm+ Icelake is coming in 2018.

Look, Intel using a useless "transistor / mm^2" metric in a slide! Like measuring cars per squared meter, right?

Right, it continues being an useless metric.

Ultimately density only means the process will be more prone to failures under heat...like the whole 7700k "do not overclock it" fiasco.

In other words...useless.

What fiasco? I hope you don't refer to invented news from certain media.

Intel tells Core i7-7700K owners to stop overclocking to avoid high temps
By Paul Lilly May 04, 2017

Some owners of Intel's Core i7-7700 and 7700K processors have been complaining for the past three months of unusual temperature increases. According to user accounts, temps sometimes spike up to 90C, close to the Core i7-7700's maximum 100C threshold, The Register reports. Intel finally offered up a response to the complaints, though the answer is not sitting well with affected users.

In an ongoing thread on Intel's community forums, a spokesman for Intel offered up the following response:

"We appreciate the feedback you have provided, and your patience as we investigated this behavior. The reported behavior of the 7th Generation Intel Core i7-7700K Processor, showing momentary temperature changes from the idle temperature, is normal while completing a task (like opening a browser or an application or a program).

In our internal investigation, we did not observe temperature variation outside of the expected behavior and recommended specifications. For processor specifications, please refer to the Intel Core i7-7700K Processor Product Specifications.

Most motherboard manufacturers offer customizable fan speed control settings that may allow for smoother transition of fan revolutions per minute (rpm). Please consult your motherboard manufacturer’s manual or website for instructions on how to change default fan speed control settings.

We do not recommend running outside the processor specifications, such as by exceeding processor frequency or voltage specifications, or removing of the integrated heat spreader (sometimes called 'de-lidding'). These actions will void the processor warranty."

The TL;DR version is that everything checks out on Intel's end and users should not overclocked their Core i7-7700K processors, which have unlocked multipliers specifically for overclocking.

As you might imagine, Intel's response did not sit well with users. The first response to Intel's post is from a ticked off customer who has sworn off Intel products.

"Three months waiting for Intel to come out with a solution, and now this?! This is all you can say?! We know already what you've just said... You know what, never mind, this would be my last product from Intel," the user wrote.

Another user called Intel's response "BS," noting that "some 7700K [processors] run even hotter than an AMD Bulldozer overclocked at 5GHz."

"I don't even have the major issue like everyone else is having. However, after Intel's response just now, they are not getting another penny out of me. I'm going to sell my Intel stuff and go to Ryzen," yet another user wrote.

To Intel's credit, some users experiencing the issue have admitted to de-lidding their processors. One reason this is done so that cooling solutions can be applied directly to the CPU die, but it's a risky procedure that can result in a dead chip. Removing the IHS can also render certain coolers incompatible, as they were designed with the height of the IHS in mind. Another motivating factor is to replace the stock thermal compound that Intel uses between the die and IHS.

Intel has also never stated that it would warranty processors that have been overclocked or over-volted (though it does offer an overclocking warranty as a separate purchase). However, that isn't the part that has users all riled up. They're ticked because Intel basically shrugged off the temperature spikes as being normal, and telling them to run their unlocked CPUs at stock settings rubs salt in the wound.

That is just what I mean. The media invented the news, because he just said that (i) you overclock at your own risk because it voids warrranty and (ii) Intel cannot warrant that chips will work on a given way when overclocked because the chip is running outside specs. Overclocking is silicon lottery.

Of course the same media that invented the news against Intel forgot to mention that AMD makes the same recommendations:

WARNING: AMD processors, including chipsets, CPUs, APUs and GPUs (collectively and
individually “AMD processor”), are intended to be operated only within their associated
specifications and factory settings. Operating your AMD processor outside of official AMD
specifications or outside of factory settings, including but not limited to the conducting of
overclocking (including use of this overclocking software, even if such software has been directly or
indirectly provided by AMD or an entity otherwise affiliated in any way with AMD), may damage
your processor, affect the operation of your processor or the security features therein and/or lead to
other problems, including but not limited to damage to your system components (including your
motherboard and components thereon (e.g., memory)), system instabilities (e.g., data loss and
corrupted images), reduction in system performance, shortened processor, system component and/or
system life, and in extreme cases, total system failure. It is recommended that you save any important
data before using the tool. AMD does not provide support or service for issues or damages related
to use of an AMD processor outside of official AMD specifications or outside of factory settings.
You may also not receive support or service from your board or system manufacturer. Please make
sure you have saved all important data before using this overclocking software. DAMAGES
CAUSED BY USE OF YOUR AMD PROCESSOR OUTSIDE OF OFFICIAL AMD
SPECIFICATIONS OR OUTSIDE OF FACTORY SETTINGS ARE NOT COVERED UNDER
ANY AMD PRODUCT WARRANTY AND MAY NOT BE COVERED BY YOUR BOARD OR
SYSTEM MANUFACTURER’S WARRANTY.

Still no site wrote any news about how AMD tell users to stop overclocking because AMD chips "are intended to be operated only within their associated specifications and factory settings." The usual media bias once again...

I agree, Intel has gotten so much biased press. Unfortunately that is the way the media has gone. It's nearly impossible to get unbiased news even on tech these days.... I'm getting pretty fed up of the fact that some people think that you have to be either an AMD or Intel fanboy. Ryzen no doubt is a good product, but in certain non-gaming tasks even an Intel 4c can equal AMD's 8c chip.

What I'm trying to say is even if you are a productivity user it's not a black and white choice.

bi·as
noun
1.
prejudice in favor of or against one thing, person, or group compared with another, usually in a way considered to be unfair.
"there was evidence of bias against foreign applicants"
synonyms: prejudice, partiality, partisanship, favoritism, unfairness, one-sidedness; More
2.
in some sports, such as lawn bowling, the irregular shape given to a ball.
verb
1.
cause to feel or show inclination or prejudice for or against someone or something.

Bias is the tendency to have an opinion, or view, that is often without considering evidence and other information.
https://en.wikipedia.org/wiki/Bias

Confirmation bias is the tendency people have to favor facts or arguments that confirm the beliefs and positions they already hold. The extreme form of this bias is referred to as “belief perseverance” when people hold onto their beliefs even after they've been proven false.

Because you expect to hear good things about a company or product you like, and instead hear bad things from multiple different respected reviewers and media outlets. One might want to think deeply about what bias is and it's meaning, or just remain blissful about the subject!

bliss
noun
1.
perfect happiness; great joy.
"she gave a sigh of bliss"
synonyms: joy, happiness, pleasure, delight, ecstasy, elation, rapture, euphoria
"she gave a sigh of bliss"
verb
1.
informal
reach a state of perfect happiness, typically so as to be oblivious of everything else.
"blissed-out hippies"

 

[goldstone77]

I received my i7-3770K in the mail yesterday! I spent a little work trying to overclock it!

30% more multi-threading, I can feel the difference! Slight bump in FPS score 8-10%.
http://www.cpu-world.com/Compare/575/Intel_Core_i5_i5-2500K_vs_Intel_Core_i7_i7-3770K.html

I'm happy with it!

 

[KirbysHammer]

Spoiler

Spoiler

Back in 2014--2015 Intel planed a 10nm Cannonlake with 4/6/8 cores for mainstream. Then found big difficulties with the 14nm node, which delayed the development of the 10nm node, forced a change on the roadmaps and introduced the new refresh cycles, with 4-core Kabylake and 6-core CoffeLake.

Scaling down to the smallest nodes is getting more and more difficult. Glofo canceled its 14nm node. TSMC and Samsung avoided the problems of dealing with a 14nm node by introducing hybrid '14nm' and '16nm' nodes based in a real 20nm substratum. Glofo canceled also the 10nm node, whereas Samsung and TSMC newest '10nm' are closer to Intel 14nm than to Intel 10nm.

Spoiler

10nm Cannonlake chips start shipping in second half of this year. And 10nm+ Icelake is coming in 2018.

Look, Intel using a useless "transistor / mm^2" metric in a slide! Like measuring cars per squared meter, right?

Right, it continues being an useless metric.

Ultimately density only means the process will be more prone to failures under heat...like the whole 7700k "do not overclock it" fiasco.

In other words...useless.

What fiasco? I hope you don't refer to invented news from certain media.

Intel tells Core i7-7700K owners to stop overclocking to avoid high temps
By Paul Lilly May 04, 2017

Some owners of Intel's Core i7-7700 and 7700K processors have been complaining for the past three months of unusual temperature increases. According to user accounts, temps sometimes spike up to 90C, close to the Core i7-7700's maximum 100C threshold, The Register reports. Intel finally offered up a response to the complaints, though the answer is not sitting well with affected users.

In an ongoing thread on Intel's community forums, a spokesman for Intel offered up the following response:

"We appreciate the feedback you have provided, and your patience as we investigated this behavior. The reported behavior of the 7th Generation Intel Core i7-7700K Processor, showing momentary temperature changes from the idle temperature, is normal while completing a task (like opening a browser or an application or a program).

In our internal investigation, we did not observe temperature variation outside of the expected behavior and recommended specifications. For processor specifications, please refer to the Intel Core i7-7700K Processor Product Specifications.

Most motherboard manufacturers offer customizable fan speed control settings that may allow for smoother transition of fan revolutions per minute (rpm). Please consult your motherboard manufacturer’s manual or website for instructions on how to change default fan speed control settings.

We do not recommend running outside the processor specifications, such as by exceeding processor frequency or voltage specifications, or removing of the integrated heat spreader (sometimes called 'de-lidding'). These actions will void the processor warranty."

The TL;DR version is that everything checks out on Intel's end and users should not overclocked their Core i7-7700K processors, which have unlocked multipliers specifically for overclocking.

As you might imagine, Intel's response did not sit well with users. The first response to Intel's post is from a ticked off customer who has sworn off Intel products.

"Three months waiting for Intel to come out with a solution, and now this?! This is all you can say?! We know already what you've just said... You know what, never mind, this would be my last product from Intel," the user wrote.

Another user called Intel's response "BS," noting that "some 7700K [processors] run even hotter than an AMD Bulldozer overclocked at 5GHz."

"I don't even have the major issue like everyone else is having. However, after Intel's response just now, they are not getting another penny out of me. I'm going to sell my Intel stuff and go to Ryzen," yet another user wrote.

To Intel's credit, some users experiencing the issue have admitted to de-lidding their processors. One reason this is done so that cooling solutions can be applied directly to the CPU die, but it's a risky procedure that can result in a dead chip. Removing the IHS can also render certain coolers incompatible, as they were designed with the height of the IHS in mind. Another motivating factor is to replace the stock thermal compound that Intel uses between the die and IHS.

Intel has also never stated that it would warranty processors that have been overclocked or over-volted (though it does offer an overclocking warranty as a separate purchase). However, that isn't the part that has users all riled up. They're ticked because Intel basically shrugged off the temperature spikes as being normal, and telling them to run their unlocked CPUs at stock settings rubs salt in the wound.

That is just what I mean. The media invented the news, because he just said that (i) you overclock at your own risk because it voids warrranty and (ii) Intel cannot warrant that chips will work on a given way when overclocked because the chip is running outside specs. Overclocking is silicon lottery.

Of course the same media that invented the news against Intel forgot to mention that AMD makes the same recommendations:

WARNING: AMD processors, including chipsets, CPUs, APUs and GPUs (collectively and
individually “AMD processor”), are intended to be operated only within their associated
specifications and factory settings. Operating your AMD processor outside of official AMD
specifications or outside of factory settings, including but not limited to the conducting of
overclocking (including use of this overclocking software, even if such software has been directly or
indirectly provided by AMD or an entity otherwise affiliated in any way with AMD), may damage
your processor, affect the operation of your processor or the security features therein and/or lead to
other problems, including but not limited to damage to your system components (including your
motherboard and components thereon (e.g., memory)), system instabilities (e.g., data loss and
corrupted images), reduction in system performance, shortened processor, system component and/or
system life, and in extreme cases, total system failure. It is recommended that you save any important
data before using the tool. AMD does not provide support or service for issues or damages related
to use of an AMD processor outside of official AMD specifications or outside of factory settings.
You may also not receive support or service from your board or system manufacturer. Please make
sure you have saved all important data before using this overclocking software. DAMAGES
CAUSED BY USE OF YOUR AMD PROCESSOR OUTSIDE OF OFFICIAL AMD
SPECIFICATIONS OR OUTSIDE OF FACTORY SETTINGS ARE NOT COVERED UNDER
ANY AMD PRODUCT WARRANTY AND MAY NOT BE COVERED BY YOUR BOARD OR
SYSTEM MANUFACTURER’S WARRANTY.

Still no site wrote any news about how AMD tell users to stop overclocking because AMD chips "are intended to be operated only within their associated specifications and factory settings." The usual media bias once again...

I agree, Intel has gotten so much biased press. Unfortunately that is the way the media has gone. It's nearly impossible to get unbiased news even on tech these days.... I'm getting pretty fed up of the fact that some people think that you have to be either an AMD or Intel fanboy. Ryzen no doubt is a good product, but in certain non-gaming tasks even an Intel 4c can equal AMD's 8c chip.

What I'm trying to say is even if you are a productivity user it's not a black and white choice.

bi·as
noun
1.
prejudice in favor of or against one thing, person, or group compared with another, usually in a way considered to be unfair.
"there was evidence of bias against foreign applicants"
synonyms: prejudice, partiality, partisanship, favoritism, unfairness, one-sidedness; More
2.
in some sports, such as lawn bowling, the irregular shape given to a ball.
verb
1.
cause to feel or show inclination or prejudice for or against someone or something.

Bias is the tendency to have an opinion, or view, that is often without considering evidence and other information.
https://en.wikipedia.org/wiki/Bias

Confirmation bias is the tendency people have to favor facts or arguments that confirm the beliefs and positions they already hold. The extreme form of this bias is referred to as “belief perseverance” when people hold onto their beliefs even after they've been proven false.

Because you expect to hear good things about a company or product you like, and instead hear bad things from multiple different respected reviewers and media outlets. One might want to think deeply about what bias is and it's meaning, or just remain blissful about the subject!

bliss
noun
1.
perfect happiness; great joy.
"she gave a sigh of bliss"
synonyms: joy, happiness, pleasure, delight, ecstasy, elation, rapture, euphoria
"she gave a sigh of bliss"
verb
1.
informal
reach a state of perfect happiness, typically so as to be oblivious of everything else.
"blissed-out hippies"

It's pretty obvious they're biased though. They have downplayed Intel's advantage in gaming and exageratted AMD's advantage in productivity. 10C Intel is actually impressively competitive with Ryzen 16C. in tasks that can't use all 32 of Ryzen's threads intel tends to win out. Yes, while many tasks are heavily threaded they sometimes cannot use all 32.

TR is still a good product and if SC performance isn't a factor it's absolutely the best choice.

 

[-Fran-]

It's pretty obvious they're biased though. They have downplayed Intel's advantage in gaming and exageratted AMD's advantage in productivity. 10C Intel is actually impressively competitive with Ryzen 16C. in tasks that can't use all 32 of Ryzen's threads intel tends to win out. Yes, while many tasks are heavily threaded they sometimes cannot use all 32.

TR is still a good product and if SC performance isn't a factor it's absolutely the best choice.

Woah, woah, wait a moment there.

When you put a CPU that is aimed at tasks that are not primarily gaming or "single threaded", it is *obvious* you will test it in such disciplines as an addition. Then, you will not care how the CPU actually achieves the performance, only that is achieves a particular threshold at the stipulated price range it will sell. In this particular case, TR and the 7900X were both put into a shizzle ton of tests across 50+ disciplines (if not more) and in all they were trading blows. There is no clear winner and the answer in every single review (as far as I have read) is that TR justifies its price with no issues facing the 7900X.

If you want to de-merit AMD for achieving a milestone just because "it doesn't game as well as the Intel", that is on you, not on the data gathered nor the tasks at hand. If you can think of a test that actually is needed that should be added, why not make the suggestion for a next round of tests here at Toms or another site that is good at reviewing stuff?

Also, if you have doubts on methodology, that is another story. Not all sites are trustworthy or draw conclusions in a weird manner, but as long as they expose their testing methodology and you can reproduce their figures, you will know why those numbers come to be and have more information to take in and make a decision.

Hell, I still haven't seen a *SINGLE* site that has put OBS (or any streaming software) while testing games to see how the CPUs behave. People has whined about 4K SO FRIGGIN' MUCH, but I haven't seen a single tear shed about streaming. So, I'll have to shed those manly tears from now on.

Cheers!

 

[goldstone77]

Spoiler

Spoiler

Spoiler

Back in 2014--2015 Intel planed a 10nm Cannonlake with 4/6/8 cores for mainstream. Then found big difficulties with the 14nm node, which delayed the development of the 10nm node, forced a change on the roadmaps and introduced the new refresh cycles, with 4-core Kabylake and 6-core CoffeLake.

Scaling down to the smallest nodes is getting more and more difficult. Glofo canceled its 14nm node. TSMC and Samsung avoided the problems of dealing with a 14nm node by introducing hybrid '14nm' and '16nm' nodes based in a real 20nm substratum. Glofo canceled also the 10nm node, whereas Samsung and TSMC newest '10nm' are closer to Intel 14nm than to Intel 10nm.

Spoiler

10nm Cannonlake chips start shipping in second half of this year. And 10nm+ Icelake is coming in 2018.

Look, Intel using a useless "transistor / mm^2" metric in a slide! Like measuring cars per squared meter, right?

Right, it continues being an useless metric.

Ultimately density only means the process will be more prone to failures under heat...like the whole 7700k "do not overclock it" fiasco.

In other words...useless.

What fiasco? I hope you don't refer to invented news from certain media.

Intel tells Core i7-7700K owners to stop overclocking to avoid high temps
By Paul Lilly May 04, 2017

Some owners of Intel's Core i7-7700 and 7700K processors have been complaining for the past three months of unusual temperature increases. According to user accounts, temps sometimes spike up to 90C, close to the Core i7-7700's maximum 100C threshold, The Register reports. Intel finally offered up a response to the complaints, though the answer is not sitting well with affected users.

In an ongoing thread on Intel's community forums, a spokesman for Intel offered up the following response:

"We appreciate the feedback you have provided, and your patience as we investigated this behavior. The reported behavior of the 7th Generation Intel Core i7-7700K Processor, showing momentary temperature changes from the idle temperature, is normal while completing a task (like opening a browser or an application or a program).

In our internal investigation, we did not observe temperature variation outside of the expected behavior and recommended specifications. For processor specifications, please refer to the Intel Core i7-7700K Processor Product Specifications.

Most motherboard manufacturers offer customizable fan speed control settings that may allow for smoother transition of fan revolutions per minute (rpm). Please consult your motherboard manufacturer’s manual or website for instructions on how to change default fan speed control settings.

We do not recommend running outside the processor specifications, such as by exceeding processor frequency or voltage specifications, or removing of the integrated heat spreader (sometimes called 'de-lidding'). These actions will void the processor warranty."

The TL;DR version is that everything checks out on Intel's end and users should not overclocked their Core i7-7700K processors, which have unlocked multipliers specifically for overclocking.

As you might imagine, Intel's response did not sit well with users. The first response to Intel's post is from a ticked off customer who has sworn off Intel products.

"Three months waiting for Intel to come out with a solution, and now this?! This is all you can say?! We know already what you've just said... You know what, never mind, this would be my last product from Intel," the user wrote.

Another user called Intel's response "BS," noting that "some 7700K [processors] run even hotter than an AMD Bulldozer overclocked at 5GHz."

"I don't even have the major issue like everyone else is having. However, after Intel's response just now, they are not getting another penny out of me. I'm going to sell my Intel stuff and go to Ryzen," yet another user wrote.

To Intel's credit, some users experiencing the issue have admitted to de-lidding their processors. One reason this is done so that cooling solutions can be applied directly to the CPU die, but it's a risky procedure that can result in a dead chip. Removing the IHS can also render certain coolers incompatible, as they were designed with the height of the IHS in mind. Another motivating factor is to replace the stock thermal compound that Intel uses between the die and IHS.

Intel has also never stated that it would warranty processors that have been overclocked or over-volted (though it does offer an overclocking warranty as a separate purchase). However, that isn't the part that has users all riled up. They're ticked because Intel basically shrugged off the temperature spikes as being normal, and telling them to run their unlocked CPUs at stock settings rubs salt in the wound.

That is just what I mean. The media invented the news, because he just said that (i) you overclock at your own risk because it voids warrranty and (ii) Intel cannot warrant that chips will work on a given way when overclocked because the chip is running outside specs. Overclocking is silicon lottery.

Of course the same media that invented the news against Intel forgot to mention that AMD makes the same recommendations:

WARNING: AMD processors, including chipsets, CPUs, APUs and GPUs (collectively and
individually “AMD processor”), are intended to be operated only within their associated
specifications and factory settings. Operating your AMD processor outside of official AMD
specifications or outside of factory settings, including but not limited to the conducting of
overclocking (including use of this overclocking software, even if such software has been directly or
indirectly provided by AMD or an entity otherwise affiliated in any way with AMD), may damage
your processor, affect the operation of your processor or the security features therein and/or lead to
other problems, including but not limited to damage to your system components (including your
motherboard and components thereon (e.g., memory)), system instabilities (e.g., data loss and
corrupted images), reduction in system performance, shortened processor, system component and/or
system life, and in extreme cases, total system failure. It is recommended that you save any important
data before using the tool. AMD does not provide support or service for issues or damages related
to use of an AMD processor outside of official AMD specifications or outside of factory settings.
You may also not receive support or service from your board or system manufacturer. Please make
sure you have saved all important data before using this overclocking software. DAMAGES
CAUSED BY USE OF YOUR AMD PROCESSOR OUTSIDE OF OFFICIAL AMD
SPECIFICATIONS OR OUTSIDE OF FACTORY SETTINGS ARE NOT COVERED UNDER
ANY AMD PRODUCT WARRANTY AND MAY NOT BE COVERED BY YOUR BOARD OR
SYSTEM MANUFACTURER’S WARRANTY.

Still no site wrote any news about how AMD tell users to stop overclocking because AMD chips "are intended to be operated only within their associated specifications and factory settings." The usual media bias once again...

I agree, Intel has gotten so much biased press. Unfortunately that is the way the media has gone. It's nearly impossible to get unbiased news even on tech these days.... I'm getting pretty fed up of the fact that some people think that you have to be either an AMD or Intel fanboy. Ryzen no doubt is a good product, but in certain non-gaming tasks even an Intel 4c can equal AMD's 8c chip.

What I'm trying to say is even if you are a productivity user it's not a black and white choice.

bi·as
noun
1.
prejudice in favor of or against one thing, person, or group compared with another, usually in a way considered to be unfair.
"there was evidence of bias against foreign applicants"
synonyms: prejudice, partiality, partisanship, favoritism, unfairness, one-sidedness; More
2.
in some sports, such as lawn bowling, the irregular shape given to a ball.
verb
1.
cause to feel or show inclination or prejudice for or against someone or something.

Bias is the tendency to have an opinion, or view, that is often without considering evidence and other information.
https://en.wikipedia.org/wiki/Bias

Confirmation bias is the tendency people have to favor facts or arguments that confirm the beliefs and positions they already hold. The extreme form of this bias is referred to as “belief perseverance” when people hold onto their beliefs even after they've been proven false.

Because you expect to hear good things about a company or product you like, and instead hear bad things from multiple different respected reviewers and media outlets. One might want to think deeply about what bias is and it's meaning, or just remain blissful about the subject!

bliss
noun
1.
perfect happiness; great joy.
"she gave a sigh of bliss"
synonyms: joy, happiness, pleasure, delight, ecstasy, elation, rapture, euphoria
"she gave a sigh of bliss"
verb
1.
informal
reach a state of perfect happiness, typically so as to be oblivious of everything else.
"blissed-out hippies"

It's pretty obvious they're biased though. They have downplayed Intel's advantage in gaming and exageratted AMD's advantage in productivity. 10C Intel is actually impressively competitive with Ryzen 16C. in tasks that can't use all 32 of Ryzen's threads intel tends to win out. Yes, while many tasks are heavily threaded they sometimes cannot use all 32.

TR is still a good product and if SC performance isn't a factor it's absolutely the best choice.

Performance is all relative. Media attention is relative. Intel had great single thread performance for ~10 years, which has been talked about to about that long. And it's still talked about. It's only recently that it's being down played in gaming, because Ryzen has really good competitive single thread performance. The comparison shows that single thread performance between Intel and AMD is at a point of diminishing returns in gaming. You need a $500+ video card to see noticeable FPS differences between the two. And then you need to have a monitor capable of utilizing higher than 60 FPS. Once you add this evidence into the conversation single thread performance in gaming doesn't really matter anymore except for a very small niche group using 144HZ@1080p. Ryzen has a single thread score of ~150+ depending on clock speed. Adjust your own frequency to see how much of a difference you see in your Intel chip with Rzyen like single thread performance. Now add 50% or 100% the cores, and look at the price tags! You will come to the realization really quickly that for the vast majority of consumers Ryzen is just the much better buy, and that is why it's being talked about so much. Now, that being said Ryzen has gone through growing pains, and until it is adopted completely and programs are optimized for it there will be performance deficits. When you buy Intel you have over a decade of program optimizations, which add to the peace of mind when buying one. But there isn't anymore bias in the media than their has been over Intel for the last 10 years or so. Intel is just sitting in the back of the bus at the moment, and people are going to talk about the "HOT" new thing. I have a feeling that is about to change when Intel mainstreams 4, and 6 core products at what was 4 core and 2 core prices. This is what I've been talking about with TV's in commercial electronics market. Many manufactures, cost keep getting reduced, and we get cheaper, bigger, and better quality TVs. Microprocessor exact opposite. Intel's monopoly has had a strangle hold on prices. Even though every node it gets cheaper to make they have raised the prices, because of lack of competition. Last years 10 core Intel was $1,800. Because of AMD this year it's $962.89(which is under the retail now). So, why is everyone talking about AMD? Is it because they are bias? Or is it because there is a lot to talk about?

 

[KirbysHammer]

Spoiler

Spoiler

Spoiler

Back in 2014--2015 Intel planed a 10nm Cannonlake with 4/6/8 cores for mainstream. Then found big difficulties with the 14nm node, which delayed the development of the 10nm node, forced a change on the roadmaps and introduced the new refresh cycles, with 4-core Kabylake and 6-core CoffeLake.

Scaling down to the smallest nodes is getting more and more difficult. Glofo canceled its 14nm node. TSMC and Samsung avoided the problems of dealing with a 14nm node by introducing hybrid '14nm' and '16nm' nodes based in a real 20nm substratum. Glofo canceled also the 10nm node, whereas Samsung and TSMC newest '10nm' are closer to Intel 14nm than to Intel 10nm.

Spoiler

10nm Cannonlake chips start shipping in second half of this year. And 10nm+ Icelake is coming in 2018.

Look, Intel using a useless "transistor / mm^2" metric in a slide! Like measuring cars per squared meter, right?

Right, it continues being an useless metric.

Ultimately density only means the process will be more prone to failures under heat...like the whole 7700k "do not overclock it" fiasco.

In other words...useless.

What fiasco? I hope you don't refer to invented news from certain media.

Intel tells Core i7-7700K owners to stop overclocking to avoid high temps
By Paul Lilly May 04, 2017

Some owners of Intel's Core i7-7700 and 7700K processors have been complaining for the past three months of unusual temperature increases. According to user accounts, temps sometimes spike up to 90C, close to the Core i7-7700's maximum 100C threshold, The Register reports. Intel finally offered up a response to the complaints, though the answer is not sitting well with affected users.

In an ongoing thread on Intel's community forums, a spokesman for Intel offered up the following response:

"We appreciate the feedback you have provided, and your patience as we investigated this behavior. The reported behavior of the 7th Generation Intel Core i7-7700K Processor, showing momentary temperature changes from the idle temperature, is normal while completing a task (like opening a browser or an application or a program).

In our internal investigation, we did not observe temperature variation outside of the expected behavior and recommended specifications. For processor specifications, please refer to the Intel Core i7-7700K Processor Product Specifications.

Most motherboard manufacturers offer customizable fan speed control settings that may allow for smoother transition of fan revolutions per minute (rpm). Please consult your motherboard manufacturer’s manual or website for instructions on how to change default fan speed control settings.

We do not recommend running outside the processor specifications, such as by exceeding processor frequency or voltage specifications, or removing of the integrated heat spreader (sometimes called 'de-lidding'). These actions will void the processor warranty."

The TL;DR version is that everything checks out on Intel's end and users should not overclocked their Core i7-7700K processors, which have unlocked multipliers specifically for overclocking.

As you might imagine, Intel's response did not sit well with users. The first response to Intel's post is from a ticked off customer who has sworn off Intel products.

"Three months waiting for Intel to come out with a solution, and now this?! This is all you can say?! We know already what you've just said... You know what, never mind, this would be my last product from Intel," the user wrote.

Another user called Intel's response "BS," noting that "some 7700K [processors] run even hotter than an AMD Bulldozer overclocked at 5GHz."

"I don't even have the major issue like everyone else is having. However, after Intel's response just now, they are not getting another penny out of me. I'm going to sell my Intel stuff and go to Ryzen," yet another user wrote.

To Intel's credit, some users experiencing the issue have admitted to de-lidding their processors. One reason this is done so that cooling solutions can be applied directly to the CPU die, but it's a risky procedure that can result in a dead chip. Removing the IHS can also render certain coolers incompatible, as they were designed with the height of the IHS in mind. Another motivating factor is to replace the stock thermal compound that Intel uses between the die and IHS.

Intel has also never stated that it would warranty processors that have been overclocked or over-volted (though it does offer an overclocking warranty as a separate purchase). However, that isn't the part that has users all riled up. They're ticked because Intel basically shrugged off the temperature spikes as being normal, and telling them to run their unlocked CPUs at stock settings rubs salt in the wound.

That is just what I mean. The media invented the news, because he just said that (i) you overclock at your own risk because it voids warrranty and (ii) Intel cannot warrant that chips will work on a given way when overclocked because the chip is running outside specs. Overclocking is silicon lottery.

Of course the same media that invented the news against Intel forgot to mention that AMD makes the same recommendations:

WARNING: AMD processors, including chipsets, CPUs, APUs and GPUs (collectively and
individually “AMD processor”), are intended to be operated only within their associated
specifications and factory settings. Operating your AMD processor outside of official AMD
specifications or outside of factory settings, including but not limited to the conducting of
overclocking (including use of this overclocking software, even if such software has been directly or
indirectly provided by AMD or an entity otherwise affiliated in any way with AMD), may damage
your processor, affect the operation of your processor or the security features therein and/or lead to
other problems, including but not limited to damage to your system components (including your
motherboard and components thereon (e.g., memory)), system instabilities (e.g., data loss and
corrupted images), reduction in system performance, shortened processor, system component and/or
system life, and in extreme cases, total system failure. It is recommended that you save any important
data before using the tool. AMD does not provide support or service for issues or damages related
to use of an AMD processor outside of official AMD specifications or outside of factory settings.
You may also not receive support or service from your board or system manufacturer. Please make
sure you have saved all important data before using this overclocking software. DAMAGES
CAUSED BY USE OF YOUR AMD PROCESSOR OUTSIDE OF OFFICIAL AMD
SPECIFICATIONS OR OUTSIDE OF FACTORY SETTINGS ARE NOT COVERED UNDER
ANY AMD PRODUCT WARRANTY AND MAY NOT BE COVERED BY YOUR BOARD OR
SYSTEM MANUFACTURER’S WARRANTY.

Still no site wrote any news about how AMD tell users to stop overclocking because AMD chips "are intended to be operated only within their associated specifications and factory settings." The usual media bias once again...

I agree, Intel has gotten so much biased press. Unfortunately that is the way the media has gone. It's nearly impossible to get unbiased news even on tech these days.... I'm getting pretty fed up of the fact that some people think that you have to be either an AMD or Intel fanboy. Ryzen no doubt is a good product, but in certain non-gaming tasks even an Intel 4c can equal AMD's 8c chip.

What I'm trying to say is even if you are a productivity user it's not a black and white choice.

bi·as
noun
1.
prejudice in favor of or against one thing, person, or group compared with another, usually in a way considered to be unfair.
"there was evidence of bias against foreign applicants"
synonyms: prejudice, partiality, partisanship, favoritism, unfairness, one-sidedness; More
2.
in some sports, such as lawn bowling, the irregular shape given to a ball.
verb
1.
cause to feel or show inclination or prejudice for or against someone or something.

Bias is the tendency to have an opinion, or view, that is often without considering evidence and other information.
https://en.wikipedia.org/wiki/Bias

Confirmation bias is the tendency people have to favor facts or arguments that confirm the beliefs and positions they already hold. The extreme form of this bias is referred to as “belief perseverance” when people hold onto their beliefs even after they've been proven false.

Because you expect to hear good things about a company or product you like, and instead hear bad things from multiple different respected reviewers and media outlets. One might want to think deeply about what bias is and it's meaning, or just remain blissful about the subject!

bliss
noun
1.
perfect happiness; great joy.
"she gave a sigh of bliss"
synonyms: joy, happiness, pleasure, delight, ecstasy, elation, rapture, euphoria
"she gave a sigh of bliss"
verb
1.
informal
reach a state of perfect happiness, typically so as to be oblivious of everything else.
"blissed-out hippies"

It's pretty obvious they're biased though. They have downplayed Intel's advantage in gaming and exageratted AMD's advantage in productivity. 10C Intel is actually impressively competitive with Ryzen 16C. in tasks that can't use all 32 of Ryzen's threads intel tends to win out. Yes, while many tasks are heavily threaded they sometimes cannot use all 32.

TR is still a good product and if SC performance isn't a factor it's absolutely the best choice.

Performance is all relative. Media attention is relative. Intel had great single thread performance for ~10 years, which has been talked about to about that long. And it's still talked about. It's only recently that it's being down played in gaming, because Ryzen has really good competitive single thread performance. The comparison shows that single thread performance between Intel and AMD is at a point of diminishing returns in gaming. You need a $500+ video card to see noticeable FPS differences between the two. And then you need to have a monitor capable of utilizing higher than 60 FPS. Once you add this evidence into the conversation single thread performance in gaming doesn't really matter anymore except for a very small niche group using 144HZ@1080p. Ryzen has a single thread score of ~150+ depending on clock speed. Adjust your own frequency to see how much of a difference you see in your Intel chip with Rzyen like single thread performance. Now add 50% or 100% the cores, and look at the price tags! You will come to the realization really quickly that for the vast majority of consumers Ryzen is just the much better buy, and that is why it's being talked about so much. Now, that being said Ryzen has gone through growing pains, and until it is adopted completely and programs are optimized for it there will be performance deficits. When you buy Intel you have over a decade of program optimizations, which add to the peace of mind when buying one. But there isn't anymore bias in the media than their has been over Intel for the last 10 years or so. Intel is just sitting in the back of the bus at the moment, and people are going to talk about the "HOT" new thing. I have a feeling that is about to change when Intel mainstreams 4, and 6 core products at what was 4 core and 2 core prices. This is what I've been talking about with TV's in commercial electronics market. Many manufactures, cost keep getting reduced, and we get cheaper, bigger, and better quality TVs. Microprocessor exact opposite. Intel's monopoly has had a strangle hold on prices. Even though every node it gets cheaper to make they have raised the prices, because of lack of competition. Last years 10 core Intel was $1,800. Because of AMD this year it's $962.89(which is under the retail now). So, why is everyone talking about AMD? Is it because they are bias? Or is it because there is a lot to talk about?

I'm not talking about business practices here. I want the best chip.

When I'm referring to Intel having an edge in gaming i'm referring of course to if you were to pair it in a configuration where the CPU is the bottleneck, like you mentioned 1080p gaming.

I'm not saying Ryzen is a bad chip. But for top of the line gaming rigs it still loses to Intel's 7700k.

I think we're all sick of Intel's price gouging, and hopefully AMD can start a price war or something of that nature.

 

[juanrga]

Woah, woah, wait a moment there.

When you put a CPU that is aimed at tasks that are not primarily gaming or "single threaded", it is *obvious* you will test it in such disciplines as an addition. Then, you will not care how the CPU actually achieves the performance, only that is achieves a particular threshold at the stipulated price range it will sell. In this particular case, TR and the 7900X were both put into a shizzle ton of tests across 50+ disciplines (if not more) and in all they were trading blows. There is no clear winner and the answer in every single review (as far as I have read) is that TR justifies its price with no issues facing the 7900X.

If you want to de-merit AMD for achieving a milestone just because "it doesn't game as well as the Intel", that is on you, not on the data gathered nor the tasks at hand. If you can think of a test that actually is needed that should be added, why not make the suggestion for a next round of tests here at Toms or another site that is good at reviewing stuff?

Also, if you have doubts on methodology, that is another story. Not all sites are trustworthy or draw conclusions in a weird manner, but as long as they expose their testing methodology and you can reproduce their figures, you will know why those numbers come to be and have more information to take in and make a decision.

Hell, I still haven't seen a *SINGLE* site that has put OBS (or any streaming software) while testing games to see how the CPUs behave. People has whined about 4K SO FRIGGIN' MUCH, but I haven't seen a single tear shed about streaming. So, I'll have to shed those manly tears from now on.

My complain is not about reviews whose tone shows an evident bias towards a given brand, where when Intel wins by a huge margin such as 45%, this is dismissed with a "Zen gives enough performance" or a similar claim, but when AMD wins by a margin of 30%, the tone changes to something as "the 1950X completely smashes the i9-7900X".

My main complain is when reviews use dirty tricks to favor one brand over other. Since Zen launch, and including AMD pre-release demos, we have seen the next dirty tricks:

• 
  ■ quad-channel disabled on Intel chips
  ■ turbo disabled on Intel chips
  ■ testing with workloads that have a bug favoring Zen
  ■ most of tests being workloads where Zen shines such as rendering, rendering, rendering, and rendering
  ■ testing at 4K to generate huge GPU-bottlenecks and favor Zen on games
  ■ testing memory-bound workloads with memory subsystem crippled on Intel side
  ■ testing with compiler/flags combinations that reduce Intel chips performance by 40--60%
  ■ testing a concrete model of Intel CPU on motherboards with known compatibility issues with that concrete model of CPU
  ■ testing Intel engineering samples instead retails chips, and label the chip on graphs as if was retail
  ■ testing overclocked AMD chips vs stock Intel chips, and label chips on the graphs as if were both stock
  ■ testing custom workload that favor Zen, instead existent official workloads
  ■ test Intel platforms with beta BIOS, and don't retest latter with final BIOS
  ■ And so on

 

[manleysteele]

I've gotten to the point that I expect bias one way or the other according to what site it is and who is doing the testing or the writing. It doesn't bother me so much, because I only give any real weight to reviews that test the retail processors and boards on reputable sites. I read all the others to get a feel for where things are going to land, but I don't actually believe them.

I've also arrived at the point where I won't buy a new architecture until it's been on the street for about 6 months.

 

[juanrga]

I've gotten to the point that I expect bias one way or the other according to what site it is and who is doing the testing or the writing. It doesn't bother me so much, because I only give any real weight to reviews that test the retail processors and boards on reputable sites. I read all the others to get a feel for where things are going to land, but I don't actually believe them.

I've also arrived at the point where I won't buy a new architecture until it's been on the street for about 6 months.

Yes, in the past we could expect bias towards one brand or towards other, and one simply read different reviews to get the overall picture. But those days most reviews are biased towards a brand and it is difficult to find serious and reliable reviews.

Also in the past, when reviews tested engineering samples they added a label "[ES]" to graphs or when tested chips overclocked they added a label "[OC]" to the graphs. In those days of biased journalism one finds graphs as the next where AMD overclocked chips are labeled as stock and Intel engineering samples are labeled as stock

 

[aldaia]

10nm Cannonlake chips start shipping in second half of this year. And 10nm+ Icelake is coming in 2018.

I would noy bet my honor on that. Looks like Intel cannon-lake has been delayed to late-2018 reportedly due to 10nm issues.
http://www.digitimes.com/news/a20170920PD207.html
https://www.theinquirer.net/inquirer/news/3017743/intel-reportedly-delays-10nm-cannon-lake-cpus-until-late-2018
https://www.tweaktown.com/news/59209/intel-delays-10nm-cpu-tech-third-time-late-2018/index.html
https://www.neowin.net/news/intels-10nm-cannonlake-processors-delayed-to-late-2018
https://www.dvhardware.net/article67284.html

 

[manleysteele]

10nm Cannonlake chips start shipping in second half of this year. And 10nm+ Icelake is coming in 2018.

I would noy bet my honor on that. Looks like Intel cannon-lake has been delayed to late-2018 reportedly due to 10nm issues.
http://www.digitimes.com/news/a20170920PD207.html
https://www.theinquirer.net/inquirer/news/3017743/intel-reportedly-delays-10nm-cannon-lake-cpus-until-late-2018
https://www.tweaktown.com/news/59209/intel-delays-10nm-cpu-tech-third-time-late-2018/index.html
https://www.neowin.net/news/intels-10nm-cannonlake-processors-delayed-to-late-2018
https://www.dvhardware.net/article67284.html

If the slide shown by neowin.net is the source of the rumor, the rumor is wrong. It's an old Intel slide that shows when Intel introduced a process technology above the time bar and when their competitors introduced the same technology below the time bar. It has been mischaracterized by neowin.