8700k benchmarks leaked.

[ravenshrike]

1 minute ago, Taf the Ghost said:

 

To me, what is probably the key, is that the Zen Package can be run in 4+0 mode and still be in dual-channel memory. It means the CCXs are actually separate from the memory infrastructure that's above them in the I/O chain. It means you can run two DDR4 controllers with however many CCXs you can run traces to the controllers. It should also mean that the CPU wouldn't need more pins, which lets you stay on the same socket. 

 

I get the feeling that AMD looked at their future plans and has extra pins unused by Ryzen/EPYC on the socket but which connect to the various MB systems that are just turned off at the moment. It's the only thing that really makes sense if they're actually planning on running the same socket for 4+ years and have it be future proof.

[MyName13]

6 minutes ago, Taf the Ghost said:

Technically, "Starship" is only one reference from a nearly 2 year old slide pointing to a 48c server CPU. It's possible they dropped it for Zen2. Just as a foregoing note.

 

To the design situation, I've been following along as people with a lot more CPU technical insight have been chatting about it. It appears the "easier" design approach with adding another CCX is that the calls between CCX are always +1 hop from within the CCX. This makes the CCX more like a 4-split Core as much as a 4c design. Putting a third CCX inside the package makes the design more simple as the calls from CCX0 to either CCX1 or CCX2 the same.

 

To me, what is probably the key, is that the Zen Package can be run in 4+0 mode and still be in dual-channel memory. It means the CCXs are actually separate from the memory infrastructure that's above them in the I/O chain. It means you can run two DDR4 controllers with however many CCXs you can run traces to the controllers. It should also mean that the CPU wouldn't need more pins, which lets you stay on the same socket. 

 

Zen is a really different design philosophy. It's very much like Lego bricks, but we're not quite to the final form of what that looks like. That comes with "Chiplet" in a few years.

What is this "chiplet"?

[AnonymousGuy]

8700k is probably going to a really good "sweet spot" chip for a decently future proof gaming box.  

 

I can't wait to see the performance matrix from max overclock vs. core count from CFL and SKX.

[Taf the Ghost]

1 hour ago, ravenshrike said:

I get the feeling that AMD looked at their future plans and has extra pins unused by Ryzen/EPYC on the socket but which connect to the various MB systems that are just turned off at the moment. It's the only thing that really makes sense if they're actually planning on running the same socket for 4+ years and have it be future proof.

3 years. It's supposed to run to 2020 on AM4. However, that's probably just Zen2. Zen3 should be on AM5 simply because DDR5 should show up around then.

1 hour ago, MyName13 said:

What is this "chiplet"?

Mix & match CPU/GPU parts "on die". Processors as Lego. Let's you keep producing I/O on, say, 14nm process, but the CPU cores are on 7nm. In theory, it could allow for massively cheaper CPUs, as only the cores & some parts of the memory system would need to be on the same process node. You could have other controllers on old, cheaper nodes because there's little reason to redesign the entire part just for a node shrink.

 

It's pretty much just applying Motherboard approach of add-in slots right onto the CPU die itself. However, it's going to be a while, still. There is a lot of technical aspects to work through, though a lot of money is being spent that way. In theory, you could double or triple yields from a specific process because your "die" need just went from 300 mm sq to 100 mm sq. This is part of the reason why Zen has come in so stupidly cheap for AMD to produce.

1 hour ago, AnonymousGuy said:

8700k is probably going to a really good "sweet spot" chip for a decently future proof gaming box.  

 

I can't wait to see the performance matrix from max overclock vs. core count from CFL and SKX.

I would expect the 8700k to beat the 7800X in most gaming applications. About the only place the SKL-X should be better is stuff that's AVX-dependent. Though I also expect the 8700k will use less power per clock speed than the 7800X. Though launch might be wonky as it is an actual new chipset, much like happened with the 7800X.

[Cartoons Plural]

I5 8600k will probably be the one to get for coffee lake.

[iiNNeX]

oh man I can't wait, here goes the Ryzen 7 build out the window ;d

[Jito463]

On 8/27/2017 at 6:33 PM, mr moose said:

Intel determine the TDP from the base clock rate with a predefined workload (whatever that is). 

 

No it's not, it is a defined spec. It has a specific meaning in EE,  That is until the Internet came along and turned it into a pissing contest.

The TDP as a spec may have a predefined meaning, but it doesn't mean that the TDP numbers we get are actually of any use if we don't know the methodology they used to determine them.  AMD and Intel both put out very different TDP numbers, which are virtually worthless unless you plan to run with the stock cooler at stock speeds.

[ravenshrike]

47 minutes ago, Jito463 said:

AMD and Intel both put out very different TDP numbers, which are virtually worthless unless you plan to run with the stock cooler at stock speeds.

Hey now, Ryzen's TDP has been pretty accurate given their stock cooler performance. Course IIRC they changed how they calculated TDP at some point.

[mr moose]

1 hour ago, Jito463 said:

The TDP as a spec may have a predefined meaning, but it doesn't mean that the TDP numbers we get are actually of any use if we don't know the methodology they used to determine them.  AMD and Intel both put out very different TDP numbers, which are virtually worthless unless you plan to run with the stock cooler at stock speeds.

They do give you the methodology though, it's all in the tech data for each chip.  It's also in the white paper I linked earlier. The methodology will change between Intel and AMD and every other manufacturer.  That is the nature of TDP ratings.

 

The figures aren't worthless, they need to be interpreted with the data from the tech sheets so you can design a cooling solution that caters  for the conditions you plan to run the chip at. 

 

The real problem with TDP is that people only look at the number and ignore why it exists and how it is supposed to be used.  Enthusiasts really need to just ignore it, largely it is not their for enthusiasts anyway, it is a rating for system builders who have to develop custom cooling solutions, it is not for enthusiasts who will use aftermarket coolers. 

 

[MageTank]

18 minutes ago, mr moose said:

They do give you the methodology though, it's all in the tech data for each chip.  It's also in the white paper I linked earlier. The methodology will change between Intel and AMD and every other manufacturer.  That is the nature of TDP ratings.

 

The figures aren't worthless, they need to be interpreted with the data from the tech sheets so you can design a cooling solution that caters  for the conditions you plan to run the chip at. 

 

The real problem with TDP is that people only look at the number and ignore why it exists and how it is supposed to be used.  Enthusiasts really need to just ignore it, largely it is not their for enthusiasts anyway, it is a rating for system builders who have to develop custom cooling solutions, it is not for enthusiasts who will use aftermarket coolers. 

 

You are still only getting half the story though. Intel and AMD might provide their testing methodology and various instruction sets/workloads used to obtain that number, but how can we be certain that the manufacturers of various cooling solutions used the same testing methodology? I've seen coolers rated for 95w of heat, that cannot dissipate 95w of heat, even at extremely low ambients. ID Cooling has been guilty of this several times over in my personal tests, and I also saw a similar issue with Cryorig (who thankfully changed their TDP rating once a few users brought it to their attention). 

 

It's almost as if various cooling manufacturers take a CPU with X TDP, run a light stress test on it, and say "well, we kept this X TDP CPU cool, therefore our product must be capable of cooling X TDP" when it's not the case. No two levels of load are identical. I can run identical voltage and clock speeds, and still get more heat out of small FFT AVX loads than what I could using the same values on a game or even H.264 stress tests. You will be hard pressed to find any actual testing methodology from cooler manufacturers, even from the big name brands. It's also hard to trust third party reviewers as well, as there are a ton of variables that impact the results of those tests. Adaptive voltages, fluctuating ambients, load-scaling (depending on the stress test) and (for the reviewers that refuse to let their coolers heatsoak long enough to form a valid conclusion) time. 

 

This is all ignoring the fact that TDP is NOT the most power or heat a chip can produce, and that it's only used as a "normal workload" figure. Some even prefer to use trickier marketing of the term, and alternatives like Scenario Design Power, to get away with falling outside of that thermal envelope. Having spent years trying to tame thermals in an ITX environment, the current use of TDP just frustrates me. I wish we had a standardized way of testing a TDP, and have it be universally applicable to all components and cooling solutions, but I doubt we will have that any time soon. 

[mr moose]

4 hours ago, MageTank said:

You are still only getting half the story though. Intel and AMD might provide their testing methodology and various instruction sets/workloads used to obtain that number, but how can we be certain that the manufacturers of various cooling solutions used the same testing methodology? I've seen coolers rated for 95w of heat, that cannot dissipate 95w of heat, even at extremely low ambients. ID Cooling has been guilty of this several times over in my personal tests, and I also saw a similar issue with Cryorig (who thankfully changed their TDP rating once a few users brought it to their attention). 

 

It's almost as if various cooling manufacturers take a CPU with X TDP, run a light stress test on it, and say "well, we kept this X TDP CPU cool, therefore our product must be capable of cooling X TDP" when it's not the case. No two levels of load are identical. I can run identical voltage and clock speeds, and still get more heat out of small FFT AVX loads than what I could using the same values on a game or even H.264 stress tests. You will be hard pressed to find any actual testing methodology from cooler manufacturers, even from the big name brands. It's also hard to trust third party reviewers as well, as there are a ton of variables that impact the results of those tests. Adaptive voltages, fluctuating ambients, load-scaling (depending on the stress test) and (for the reviewers that refuse to let their coolers heatsoak long enough to form a valid conclusion) time. 

 

This is all ignoring the fact that TDP is NOT the most power or heat a chip can produce, and that it's only used as a "normal workload" figure. Some even prefer to use trickier marketing of the term, and alternatives like Scenario Design Power, to get away with falling outside of that thermal envelope. Having spent years trying to tame thermals in an ITX environment, the current use of TDP just frustrates me. I wish we had a standardized way of testing a TDP, and have it be universally applicable to all components and cooling solutions, but I doubt we will have that any time soon. 

 

That is an issue with the cooler and how the manufacturer rates it rather than the TDP rating.

 

Also I have already pointed out that TDP is not a roof top figure.   It really seems like people need to stop referring to the TDP because no one understands it's purpose.

[mr moose]

TDP is just like horse power,  It is a defined figure and there is a difference (big) between a company selling a product to the end user and inflating that figure for marketing purposes and a manufacturer giving a dodgy figure in their data sheet (which as of yet remains to be the case).  

 

Again, TDP is not for end users, it is for engineers to design cooling solutions.  If you buy a cooler of the shelf then it is up to the manufacturer of that cooler to define its suitability for a specific use,  looking at the Intel or AMD TDP specs will not help. 

[MageTank]

6 minutes ago, mr moose said:

 

That is an issue with the cooler and how the manufacturer rates it rather than the TDP rating.

 

Also I have already pointed out that TDP is not a roof top figure.   It really seems like people need to stop referring to the TDP because no one understands it's purpose.

To be fair, a lot of websites and reviewers are guilty of using TDP as an estimation of power consumption, when it's certainly not the case. PCPartpicker does this as well. Pretty misleading, as it makes people think they can get away with using weaker power supplies, but that's a rant for another day.

 

My point is, TDP as a general concept, is difficult to use across multiple sectors of the computer industry due to the fact that people get their numbers in different ways. A TDP for an AMD and Intel CPU are certainly going to be very different and incomparable to each other, and the same is said when comparing cooling solutions from competing manufacturers. It's not standardized, nor can it be used as a blanket term in that regard. Unless you have evidence that it's derived from a standardized formula/methodology that must be adhered to by anyone that uses "TDP" as a means to show the average amount of power that needs to be dissipated/ that their product can dissipate, I am left to conclude that it's still as ambiguous as it was back then. 

[XenosTech]

3 minutes ago, MageTank said:

To be fair, a lot of websites and reviewers are guilty of using TDP as an estimation of power consumption, when it's certainly not the case. PCPartpicker does this as well. Pretty misleading, as it makes people think they can get away with using weaker power supplies, but that's a rant for another day.

 

My point is, TDP as a general concept, is difficult to use across multiple sectors of the computer industry due to the fact that people get their numbers in different ways. A TDP for an AMD and Intel CPU are certainly going to be very different and incomparable to each other, and the same is said when comparing cooling solutions from competing manufacturers. It's not standardized, nor can it be used as a blanket term in that regard. Unless you have evidence that it's derived from a standardized formula/methodology that must be adhered to by anyone that uses "TDP" as a means to show the average amount of power that needs to be dissipated/ that their product can dissipate, I am left to conclude that it's still as ambiguous as it was back then. 

I always looked at tdp as how good your cooling should be at minimum (being stock config). I must have been look at this through the wrong eye

[MageTank]

Just now, XenosTech said:

I always looked at tdp as how good your cooling should be at minimum. I must have been look at this through the wrong eye

I kinda do the same, except I go for "worst possible stress" then adjust my CPU/Cooling solution until I hit the thermal results I wanted to see. For example: Using my delidded 6700k on a 45mm vapor chamber, I refused to allow it to go beyond 80C, even under 48k FFT Prime95 loads. So because of that limitation imposed upon myself, I never overclocked it on that cooling solution in that case. Now that I have a bigger case and a 280mm AIO, I run a 7700k at 5ghz under the exact same levels of stress. If for some reason, my AIO did not suffice, or I needed a bigger case, I would either upgrade coolers/cases, or adjust my CPU voltages/clocks until I met my goal again. 

 

I should note that Prime95 48k FFT is not "worst possible stress", but let's be real. Linpack MKL is far too scary to use when you use insanely overclocked ram, and god knows I am not going to underclock my ram to prevent my CPU from dying. I'd sooner let it go up in flames, lol. 

[hammer3339]

On 8/28/2017 at 3:17 AM, DrMacintosh said:

at this point Intel's IPC is useless unless they can price it right and I don't think they can. 

if they come in at the same, traditional 4 core i7 pricing that this is slated to replace it will be fine ? it will slaughter the 1700s out there

[XenosTech]

2 minutes ago, MageTank said:

I kinda do the same, except I go for "worst possible stress" then adjust my CPU/Cooling solution until I hit the thermal results I wanted to see. For example: Using my delidded 6700k on a 45mm vapor chamber, I refused to allow it to go beyond 80C, even under 48k FFT Prime95 loads. So because of that limitation imposed upon myself, I never overclocked it on that cooling solution in that case. Now that I have a bigger case and a 280mm AIO, I run a 7700k at 5ghz under the exact same levels of stress. If for some reason, my AIO did not suffice, or I needed a bigger case, I would either upgrade coolers/cases, or adjust my CPU voltages/clocks until I met my goal again. 

 

I should note that Prime95 48k FFT is not "worst possible stress", but let's be real. Linpack MKL is far too scary to use when you use insanely overclocked ram, and god knows I am not going to underclock my ram to prevent my CPU from dying. I'd sooner let it go up in flames, lol. 

Burn baby burn !!! Ummm I mean really shouldn't let the poor cpu become a george foreman grill

[mr moose]

3 minutes ago, MageTank said:

To be fair, a lot of websites and reviewers are guilty of using TDP as an estimation of power consumption, when it's certainly not the case. PCPartpicker does this as well. Pretty misleading, as it makes people think they can get away with using weaker power supplies, but that's a rant for another day.

 

My point is, TDP as a general concept, is difficult to use across multiple sectors of the computer industry due to the fact that people get their numbers in different ways. A TDP for an AMD and Intel CPU are certainly going to be very different and incomparable to each other, and the same is said when comparing cooling solutions from competing manufacturers. It's not standardized, nor can it be used as a blanket term in that regard. Unless you have evidence that it's derived from a standardized formula/methodology that must be adhered to by anyone that uses "TDP" as a means to show the average amount of power that needs to be dissipated/ that their product can dissipate, I am left to conclude that it's still as ambiguous as it was back then. 

 

TDP should never be used across sectors of the industry, it was never intended for that and has always only been relevant to the one product and definition it accompanies.  You are right though, tech media and websites are probably to blame for this confusion.

 

I don't see how it can be ambiguous when they define the conditions in enough detail to project cooling requirements for any load.

 

7 minutes ago, XenosTech said:

I always looked at tdp as how good your cooling should be at minimum (being stock config). I must have been look at this through the wrong eye

 

 

That's pretty much where the TDP comes from,  you don't really need to be concerned with the details if the manufacturer of your cooler has done their homework.

[XenosTech]

Just now, mr moose said:

That's pretty much where the TDP comes from,  you don't really need to be concerned with the details if the manufacturer of your cooler has done their homework.

I'm less worried about them doing their homework and more worried about what crazy idea I may get from certain users on this forum *cough* @MageTank *cough*

[mr moose]

6 minutes ago, XenosTech said:

I'm less worried about them doing their homework and more worried about what crazy idea I may get from certain users on this forum *cough* @MageTank *cough*

 

If the idea he gives you is to upgrade and overclock like your life depends on it, then that ain't that crazy. 

[XenosTech]

3 minutes ago, mr moose said:

 

If the idea he gives you is to upgrade and overclock like your life depends on it, then that ain't that crazy. 

lol it is when you're as bad of an impulse spender as I am.... Today my gf had to talk me out of buying a 1080Ti

[mr moose]

3 minutes ago, XenosTech said:

lol it is when you're as bad of an impulse spender as I am.... Today my gf had to talk me out of buying a 1080Ti

yesterday I spent $400 on an air compressor and new hoses for my oxy set.   I feel your pain.

[XenosTech]

7 minutes ago, mr moose said:

yesterday I spent $400 on an air compressor and new hoses for my oxy set.   I feel your pain.

Sigh the struggles of quitting your job and keeping money so you don't starve lol

[mr moose]

1 minute ago, XenosTech said:

Sigh the struggles of quitting your job and keeping money so you don't starve lol

you only live once, there's always food in the bins near McDonald's.  A 1080ti is a very important purchase.

[MageTank]

4 minutes ago, hammer3339 said:

if they come in at the same, traditional 4 core i7 pricing that this is slated to replace it will be fine ? it will slaughter the 1700s out there

Honestly, they can't really afford to price it any higher without stepping into X299's domain. If they price their 8700k too close to the 7800X (which is $383-$389 MSRP, currently priced for $395 on Newegg), you might as well buy an X299 board at that point and reap the benefits the platform has to offer. This is all under the assumption that you bought the 6 core, 12t CPU because you actually needed more thread power, not because you wanted a gaming CPU. At $340-$350, there is still a enough of a buffer to separate the CPU's (along with the difference in platform/cooling costs). 

 

2 minutes ago, mr moose said:

 

TDP should never be used across sectors of the industry, it was never intended for that and has always only been relevant to the one product and definition it accompanies.  You are right though, tech media and websites are probably to blame for this confusion.

 

I don't see how it can be ambiguous when they define the conditions in enough detail to project cooling requirements for any load.

I've read the Intel whitesheets, and I cannot find any disclosed methodology or information that offers us a means to validate the advertised TDP's as being "sufficient". While I know it's not my place to question their engineers, I do question their choice of words here:

Quote

The processor TDP is the maximum sustained power that should be used for the design of the processor thermal solution. TDP is a power dissipation and component temperature operating condition limit, specified in this documnent, that is validated during manufacturing for the base configuration when executing a near worst case commercially available workload as specified by Intel for the SKU segment. TDP may be exceeded for short periods of time or if running a very high power workload.

What is this workload, and where exactly is it specified? This is the source of that quote: https://www.intel.com/content/www/us/en/processors/core/7th-gen-core-family-desktop-s-processor-lines-datasheet-vol-1.html

 

But I've looked through several generations of Intel's whitesheets, and have yet to find an actual testing methodology that they use to arrive to the numbers they get. If they say "you need a cooler capable of dissipating 95w of heat under near worst case commercial load" for a 7700k, I'd like to know what test they used to achieve that load. We know what Intel's temperature tolerance range is, as they very clearly list tCase and tJunction, but there are still plenty left to be desired when speaking about TDP. Perhaps I just don't get it, but I feel like I should be able to know how they arrived at that number, and that I should be able to test their methodology as a means to validate those claims. I ask this out of every reviewer I watch, is it asking too much for the manufacturers to offer the same information?

 

The ambiguity for me, stems from the fact that as long as your CPU is not falling below base clocks, it's perfectly fine by Intel's standards. Not everyone else feels this way, and would consider their cooling solution inadequate if it could not maintain all-core boost ratios (as defined by the default boost table, not by advanced turbo). Intel also lists anything under tJunction as "okay", but in the very same document, mention prolonged temperatures above 76C being harmful to longevity. I've seen plenty of i7's go beyond 76C, even in just gaming workloads, at stock clocks on the Haswell DC stock cooler (the one with the copper slug). 

 

It becomes even more obfuscated to me, when you have i5's and i7's with identical TDP's, but i7's run a clear 15C hotter due to HT. No rhyme or reason as to why the TDP's are identical, just a static "X TDP for both of these CPU's". All of this completely ignoring the fact that turbo boost and variable voltage completely throws away the accuracy of any TDP rating anyways. 

 

Sorry for this rant, but man, I really hate the concept of TDP as a whole, lol. If you could somehow simplify it for a dummy like me to comprehend, I'd greatly appreciate it. Part of me feels I am overthinking it entirely, and the other part still doesn't even know if I am on the right track at all when it comes to understanding it. The more I look into it, and cross reference various coolers to the actual advertised loads (130w of heat being produced, cooled by a cooler rated for 130w), the further I get from actually making any sense of it.