Confirmed: Core i9 CPU Series will get soldered heatspreader

That 40 lanes is for the platform (16 from CPU + 24 from chipset). That's the same as their current consumer products.

 

20% is pretty significant.

The percentage difference is even bigger on somewhat older games or titles that aren't well-threaded.

Take CSGO for example.

AVG for 8700k is 835 vs 2700x 550.

That's like right at 50% faster.
Similar story for popular older titles like TES oblivion/skyrim, old UE3 games like borderlands 2 etc

I also see that the min frames are pretty significantly different too, as much as 30-50% in Hardware Unboxed's 35 game review when GPU is not the restriction.

It'll be like this for a while until game devs can distribute game logic across many cores, which is pretty complicated.

I wish AMD was faster but It still stands that there is still a pretty significant gap when GPU isn't the limit.

I can only imagine if 2080Ti actually exists and is like 50% faster than the 1080Ti

 

I'm using a 60 Hz UHD TV (49").
Some, maybe most, prefer speed (Hz), but I will not go back to anything smaller (in fact, I'm waiting for a 65" LCD with HDMI 2.1 next year).
The 60 Hz limit is the reason why I'm still using my 4790k@4,7Ghz and waiting for Zen 2.

 

Are you sure you quoted the right person/answer to the thread? Makes more sense you responded to @Borys' answer, just my opinion.

 

Thanks AMD/Ryzen. Not that I will purchase this anyway but..

 

Good Job. You basically used as example 3rd category. One where fps is well above screen capability. (Wait what? Outside of G/Freesync range?)
Now bit of education. With 1000 fps, motion to photon time is not 1ms. Even with 1000Hz mouse Input still has to go through driver, be processed by windows, handled down to game and processed for upcoming frame, as you can not alter currently rendered frame to be based on latest state of input. Then Frame prepared and sent to rendering, it is likely 2nd or third frame in queue. Then once this frame is rendered it is going to screen. That's at best 240Hz and if we ignore all internal processing (which is done especially on gaming screens), you still have wait time in between 0,5~4,2ms there. Entire chain takes like 25ms at best. In most competitive setups around 30ms. And on regular screens 45ms.

Now, what's difference in delay caused by game engine outputting 500fps vs. 1000fps? Oh Yes, one produces result ever 2 ms and other every 1 ms. That's right going from 500fps to 1000fps reduces input lag by 1ms.
Your example improves input lag chain by mere 0.62ms.

I see quite some difference between fluidity of 120fps and 240fps. But I would have to make fairy tales to even begin persuading anyone that I can tell difference between 25 and 35ms input lag.
I even wonder if you could tel which is which in following scenario:
- Two 240Hz screens
- Both playing same 240 fps movie
- One is displaying 1 frame behind (4,2ms)

What about 2 frames difference there? (8,3ms) And so on...

 

My point has nothing do with super high frame rates

Try oblivion using an ancient engine where 2700x gets 40 fps in one spot and 8700k gets double that.

The point is performance is pretty significant for many games where play-ability is greatly increased.

 

Do you have a link where older titles, like Oblivion, have been reviewed with both? The only concerning difference I see in the HW unboxed benches is Total War Saga.

 

Just like they optimized for GCN and Bulldozer and its successor when AMD got the console contracts.. Don't beat a dead horse, the meme is dead, let it rest in peace.

 

I saw a page on a UK famous website that said this below but it's gone now been taken down.
Core I9 9900K Pre-Binned 5.4Ghz

 

I still use 60hz cause I wont pay those premiums for those 120hz monitors or the premium to feed those monitors those framerates. if all i did was play 15 year old games then it would be no issue

so stock would probably say well under 55c under full load, I would love those temps! I not gona delid though, there bigger chance of me OC which people that know me around here know I will not do haha

 

The last soldered mainstream Intel cpus were Sandy Bridge.

The last soldered HEDT cpus were Broadwell-E/Haswell-E.

Those mainstream cpus were always better OC-ers than the HEDT ones.

The culprit wasn't really the solder vs the paste but processes cpus were made on. Soldered 6900k vs 7820x with the paste. 7820x can reach 4.8 without deliding, 5.0ghz after delid, forget about it with 6900k. Yes, the paste makes cpu run hotter but it is not the sole reason behind lower clocks of previous generations. The current lineup of Intel cpu is better at OC-ing than the SB.

32nm seems like the best process Intel have ever made. Leaked ES could do 4.7ghz on air in 2009 and the retail ones like i5 6xx could reach 5.0ghz, best ones up to 5.2ghz which was pretty insane in 2010. One year later you got Sandy Bridge on the 32nm. Then Ivy, Haswell, Broadwell, Skylake etc. They coundn't be OC-ed as much as the 32nn ones because new processes of Intel had a higher density and were much less mature.

14nm wasn't any different, Intel had similar problems as now with 10nm, Broadwell got delayed and it was just the 14nm die shrink of Haswell. OC-ing improved with Skylake, further with 14nm+ and Coffee Lake /8700k/ and now 14nm++ with 9900k.

Broadwell, Haswell wouldn't often even POST past 4.6ghz and that's something you would call ordinary OC under Coffe Lake.


Why would Intel do something about OC-ing? Their entire desktop business is built around that and they even offer you OC protection. If you damage/kill your cpu by OC, they will send you a new one. If you had paid.

https://click.intel.com/tuningplan/purchase-a-plan

https://click.intel.com/tuningplan/faq

What is the Performance Tuning Protection Plan?
The Performance Tuning Protection Plan is an additional plan that a customer can purchase to cover boxed processor failures caused by operating the eligible boxed processor outside of Intel’s published specifications.

"Why is Intel offering the Performance Tuning Protection Plan?
Intel has received feedback from customers who want to implement overclocking on eligible boxed processors, that because of the lack of any replacement coverage for the eligible boxed processors, the risk of overclocking is too great. We understand this position, and while we cannot endorse overclocking, we want to provide a limited remedy if issues arise as a result of their decision to enable overclocking. The limited remedy consists of a replacement processor (i.e., replacing the processor damaged solely as a result of the overclocking)."

The offer has been up for years.

 

@Agent-A01
And once you compare to equivalent amount of cores on both sides, we can talk. Oh wait, intel doesnt have consumer cpus with 32 cores...
Stating that an intel (with a lot less cores) is faster in gamin, is a given, as thats the same comparing TR to R.

@cowie
Not since FX days.
Google amd and select "images". How many logos/icons are red...

 

Uhm, what's your point?

HEDT 28c is around the corner.
But that's not even close to the point I made.

Are you trying to say a 2700x is not comparable to a 8700k or i9 9900k?

You must be smoking something as I never made a comparison between a 6core and a 32core.

No general consumer is buying a 32c for games.
2700x or less is what they'll buy.

Since I mentioned 2700x multiple times in previous posts, maybe you don't understand that it's an 8 core.

 

Cooking out of water much? Because ES: Oblivion s 12 years old. What was top dog CPU at time? Something like Core 2 Duo @2,6GHz?
I remember playing it on AthlonXP 2200+ @1,8GHz and X800. No 40fps issue either.

So, do you want to tell me, that CPUs which have like 1/5th performance per core played it at 60fps and modern chips do 40/80fps?

Maybe you are right. Maybe Oblivion used some decade obsolete instruction set which is now emulated. But are you really basing your estimation on that?
More you want to prove some point, more nonsense you post. Maybe you can return to era where intel's compiler checked CPU and if it was not intel, it just crippled performance. That's your dreamland of arguments anyway.

 

@Agent-A01
Is was referring to the fact that TR can run in game mode 8/16) similar to i9, but once the intel has similar core count, performance in games will drop as well.

Sure vs the 2700 is a bit slower, but only for ST/1080p/high fps gaming, which isn't the case for 90% of ppl buying ryzen.

 

Parts of oblivion have a lot of draw calls which makes game thread choke.

Considering they support the same instruction set it's not that.
But nobody plays stock oblivion, there are many mods.

Anyways, I play a lot of older games so it's just a fact that you'll have to accept that ryzen is much slower in older titles.

I haven't spewed any nonsense, you just can't accept facts is all I'm getting from you.
Bringing back ancient things intel has done which has nothing do with the topic just proves that.

 

Nope, just your artificial scenario which suits your agenda (not necessarily truthful or real). While I can say that old poorly threaded game overloaded with mods can bring to its knees one core on any CPU.
I still say that user who decides to go into unrealistic situation by loading all those mods and then complains does not have all five together.

Crippled by decision. It is almost like those people who get their arms/legs amputated because they have fetish for being taken care of. Or beggars by occupation who make their situation look much worse than it actually is.

So for now, you just make stories.

 

Just a short reminder:
This forum is what we make of it. Let's keep it civil folks, I sense a disturbance in this thread and in overall politeness over discussions.
Let's thinks that all world can read our comments, don't we all want to keep this forum at the highest standards as far goes tech knowledge, politeness, and fair play?
I am no mod, nor do I want to be one. I just respect what Hilbert gives us, this forum for free and I intend to remind everyone what he always said:
This forum is what we make of it.
Gentlemen, enjoy your day!

 

It's not an artificial scenario.
Especially when millions of players play that game engine.

There's a huge difference in performance in even the latest Fallout 4.

Even 'old' games like crysis 1 shows 8700k almost 2x faster than Ryzen.

Just shows they have a ways to go for gaming performance.

It'll probably take a total redesign to increase older games performance.

You can believe what you want but there's already plenty of data to prove my point.