• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Core 9000 series

no good having extra cores if they not worth using. or slower than your competitor in actual world uses.

most of the intel chips we talking about will be going into gaming rigs. they make the product for whats needed. not try and make it then when fail give extra cores and try and market it for something else.
 
~>Dg<~ said:
no good having extra cores if they not worth using. or slower than your competitor in actual world uses.

most of the intel chips we talking about will be going into gaming rigs. they make the product for whats needed. not try and make it then when fail give extra cores and try and market it for something else.
Click to expand...

The IPC on Ryzen 2 is already give and take the same as Coffeelake, the core 9000 is renamed coffeelake, Ryzen 3000 (Zen 2) is rumoured to have 10 to 15% higher IPC, even if Ryzen 3000 can only manage 200Mhz / 300Mhz over ryzen 2000 the IPC gain will MORE than make up for the clock speed difference.
So at least on par with Intel with at least 50% more cores.

That's how AMD will play this one out, by matching if not beating Intel core for core and a lot more of them denying Intel a HEDT market, mind you they are effectively already doing that.

Dg, Intel are playing to AMD's tune, and they will still lose on the technology front.
 
I did put the disclaimer that it could be fud.

But if they're using an improved 14nm again, the it will most likely clock higher than the 6 core part. The 8700K came with a 100~200Mhz bump on the clocks you could achieve with the 7700K due to 14nm++ (higher gate pitch compared to 14nm+), if they further refined the process then it could very well clock higher than the 8700K. Remains to be seen though.

@humbug Give and take 5~7%? Pretty large give and take margin. And smaller nodes won't clock as high as larger ones, at least not initially. It's one of the reason why you won't see desktop parts at 10nm from Intel until they get to 10nm+.
 
AndreiD said:
I did put the disclaimer that it could be fud.

But if they're using an improved 14nm again, the it will most likely clock higher than the 6 core part. The 8700K came with a 100~200Mhz bump on the clocks you could achieve with the 7700K due to 14nm++ (higher gate pitch compared to 14nm+), if they further refined the process then it could very well clock higher than the 8700K. Remains to be seen though.

@humbug Give and take 5~7%? Pretty large give and take margin. And smaller nodes won't clock as high as larger ones, at least not initially. It's one of the reason why you won't see desktop parts at 10nm from Intel until they get to 10nm+.
Click to expand...

-3.5% per core, +4.5% with SMT.
 
Smaller nodes does tend to reduce clock speeds at first but remember AMD are on a "low power" node right now and are supposedly moving to a "high performance" node next year. We have even less information about IPC and clocks of Ryzen 3 than we do about Core 9000, it's not really worth discussing all the variable permutations yet.

I think the i7-9700K will be the go-to pure gaming CPU for a while but that may not last very long if Ryzen 3 delivers and Intel suffer supply shortages again like they did with Coffee Lake. If people can't really buy it at SRP until January, it might only be around for a few months before Ryzen 3. I'm not really sure who the i9-9900K is aimed at though.
 
DragonQ said:
Smaller nodes does tend to reduce clock speeds at first but remember AMD are on a "low power" node right now and are supposedly moving to a "high performance" node next year. We have even less information about IPC and clocks of Ryzen 3 than we do about Core 9000, it's not really worth discussing all the variable permutations yet.
Click to expand...

Right now AMD are on Mobile nodes, the sort of thing your ARM phone CPU's are on.

Peak efficiency is at 3Ghz 1.1v, that's the 8 core Ryzen 1700, 65 Watts, impressive, it even landed in some Laptops.

At the other end you have the 1800X at 3.7Ghz with 1.3v.

12nm (Ryzen 2000) is simply an improved version of that ^^^^ slightly more efficient, about 10% so you get about 10% higher clocks for the same volts.

Zen 2 (Ryzen 3000) is indeed on a proper node, they will clock properly.

The 10 - 15% IPC gain comes from CanardPC, they thus far have been bang on with every Ryzen prediction, the 12 to 16 core Mainstream also came from them and we now know because EPYC 2 has 4X 12 core and 4X 16 core dies that is indeed the case.
 
Last edited:
~>Dg<~ said:
no good having extra cores if they not worth using. or slower than your competitor in actual world uses.

most of the intel chips we talking about will be going into gaming rigs. they make the product for whats needed. not try and make it then when fail give extra cores and try and market it for something else.
Click to expand...

This right here shows your utter lack of comprehension... gaming market is insignificant to the rest of the market that these chips will go into, HEDT in Enterprise will use a lot of these as a cheaper solution to the real HEDT Stuff, Corporate business still uses 2 core laptops (well until AMD kicked Intel up the rear and now CFL Stuff is 4 core) Corporate HEDT is often quad core Xeon which is cheaper than buying 79xx chips. These chips will be the perfect option for high end HEDT Corporate Workstations to pair with Enterprise GPU's without costing a ton.

Gaming Market is peanuts compared to the actual volume intel shifts elsewhere.
 
@humbug I too can make up numbers.

@DragonQ 12LP isn't a low power node though, it's GF's high performance version of Samsung's 14LPP (which despite the name was a higher performance version of 14LPE).
Look up Intel's 10nm sliders, even their initial 10nm won't be as high performance as 14nm.

But as for the 9000 series, if they're using 14nm++ then they'll probably clock as high as the 6 core parts, if they're using an improved processed then it might clock slightly higher.
 
AndreiD said:
@humbug I too can make up numbers.

@DragonQ 12LP isn't a low power node though, it's GF's high performance version of Samsung's 14LPP (which despite the name was a higher performance version of 14LPE).
Look up Intel's 10nm sliders, even their initial 10nm won't be as high performance as 14nm.

But as for the 9000 series, if they're using 14nm++ then they'll probably clock as high as the 6 core parts, if they're using an improved processed then it might clock slightly higher.
Click to expand...

This is true, Intels 10nm wont be nearly as good as their 14nm+++ stuff, afterall that has had the benefit of many revisions. It will probably take a couple of revisions on 10nm to bypass 14nm.
 
I don't even expect 10nm+ to have higher clocks than their most mature 14nm, but they're supposedly increasing L1D and L2 cache sizes with Icelake to increase IPC, so single core performance won't take a dive due to lower clocks. Remains to be seen what happens on that front, but I do think people should be tempered in their expectations of higher clocks on denser nodes, it never really worked like that.
 
With what's said in the couple of posts above ^^^^

If Intel can't get 10nm to surpass their current 14nm+++ they will use it anyway and they will not care much about your Mega Gigahertz gaming performance, they will be marketed as efficient and they will sell by the ton because yes actually gamers are a tiny tiny fraction of an already tiny fraction of Intel's market.

You do not matter to them, the instant you become inconvenient to their wider plans you're left out in the cold.

Don't get so invested in one company, they don't care about you at all.
 
Sucks that they are disabling HT on the i7 just to carve out a higher price bracket in the i9.

The 2700X is already closer to the i5 than i7 in price so i'm not convinced these will improve the value proposition over AMD's existing line (for productivity).
 
Although people may call the 9700k/9900k just rebranded Coffeelake, there are clear improvements over the i7-8700k.

People are right when they say intel should have always refrained from using the pigeon poop TIM under the IHS, but even with this paste, it was and still is a clear winner for IPC vs IPC, stock clocks vs stock clocks and OC potential vs OC potential against absolutely anything AMD has to offer and that also applies prior to any delidding.
Even with the Thermal limitations of the poor TIM, the i7-8700k comes out on top in all these specific tests in pretty much any benchmark you'll find on the web. The positive differences we have this time around is that rather than the need for delidding to push the boundaries and to try to achieve that 5ghz or greater OC, we now have a soldered IHS that will significantly lower thermal limitations and we also have an Intel 8 core 16 thread mainstream cpu that is capable of 5ghz out of the box, no OC, delidding or special requirements needed, it will hit 5ghz over 2 cores without touching a single bios setting.

When you think that an 8 core 16 thread cpu with a higher IPC will go 5ghz out of the box, there is no reason for anyone to suggest that it wont take some serious beating. This 5ghz is also a very reserved and a very safe turbo boost that is designed for people to run these clocks under the poorest of coolers, making sure that they keep their RMAs down to a minimum.

I personally feel i'm not going to be too far away when i suggest that we will be probably be looking at a 5.2ghz chip over all 8 cores with 16 threads for the I9-9900k (when under decent cooling of course). The 8700k had a max boost of 4.7ghz but has an average overclock of around 5ghz with some cases higher OC over all 6 available cores, this means it had a 300mhz overclock boost over its max turbo clock on all cores, if you had a 300mhz boost on top of the i9-9900k's max boost over all of its available cores, then we'd be at 5.3ghz.
You could even say 5.2ghz is being reserved what with a now soldered IHS and refinements to 14nm++....This I9 is going to be one hell of a fast cpu and surely nobody can deny this?

I personally think you could subtract the small performance percentage hits of any windows 10 patch/update/code for as many security flaws as you wish, there still wont be anything come remotely close to it in terms of actual speed.
 
Jamie Archer said:
Although people may call the 9700k/9900k just rebranded Coffeelake, there are clear improvements over the i7-8700k.

People are right when they say intel should have always refrained from using the pigeon poop TIM under the IHS, but even with this paste, it was and still is a clear winner for IPC vs IPC, stock clocks vs stock clocks and OC potential vs OC potential against absolutely anything AMD has to offer and that also applies prior to any delidding.
Even with the Thermal limitations of the poor TIM, the i7-8700k comes out on top in all these specific tests in pretty much any benchmark you'll find on the web. The positive differences we have this time around is that rather than the need for delidding to push the boundaries and to try to achieve that 5ghz or greater OC, we now have a soldered IHS that will significantly lower thermal limitations and we also have an Intel 8 core 16 thread mainstream cpu that is capable of 5ghz out of the box, no OC, delidding or special requirements needed, it will hit 5ghz over 2 cores without touching a single bios setting.

When you think that an 8 core 16 thread cpu with a higher IPC will go 5ghz out of the box, there is no reason for anyone to suggest that it wont take some serious beating. This 5ghz is also a very reserved and a very safe turbo boost that is designed for people to run these clocks under the poorest of coolers, making sure that they keep their RMAs down to a minimum.

I personally feel i'm not going to be too far away when i suggest that we will be probably be looking at a 5.2ghz chip over all 8 cores with 16 threads for the I9-9900k (when under decent cooling of course). The 8700k had a max boost of 4.7ghz but has an average overclock of around 5ghz with some cases higher OC over all 6 available cores, this means it had a 300mhz overclock boost over its max turbo clock on all cores, if you had a 300mhz boost on top of the i9-9900k's max boost over all of its available cores, then we'd be at 5.3ghz.
You could even say 5.2ghz is being reserved what with a now soldered IHS and refinements to 14nm++....This I9 is going to be one hell of a fast cpu and surely nobody can deny this?

I personally think you could subtract the small performance percentage hits of any windows 10 patch/update/code for as many security flaws as you wish, there still wont be anything come remotely close to it in terms of actual speed.
Click to expand...

If security was not a big issue you could always stay with the older Bios for your motherboard and avoiding some of the slow downs but you will not be able to with the Z390 boards.
 
Jamie Archer said:
This I9 is going to be one hell of a fast cpu and surely nobody can deny this?
Click to expand...
Yes, if you want pure speed up to 16 threads at any cost it is a no-brainer, unless you need other HEDT features like PCIe lanes or quad channel RAM.

shamus21 said:
If security was not a big issue you could always stay with the older Bios for your motherboard and avoiding some of the slow downs but you will not be able to with the Z390 boards.
Click to expand...
The new chips' default microcode will already include a bunch of security patches so that won't help much. I imagine the 9000 series will have worse "out of the box" IPC than the 8000 series due to this (although it won't be very noticeable in gaming).
 
You must log in or register to reply here.
Back
Top Bottom