Core 9000 series

[Doobedoo]

So 95deg flatout overclock at 5ghz 1.4v on all 8cores or 70deg 4.7/4.8?1.2v on all 8. (On that particular chip)

Going to be a pretty good performer. I wonder how much the 9900k will suffer with ht?

 

[SiDeards73]

So 95deg flatout overclock at 5ghz 1.4v on all 8cores or 70deg 4.7/4.8?1.2v on all 8. (On that particular chip)

Going to be a pretty good performer. I wonder how much the 9900k will suffer with ht?

Interesting comment at the end about cherry picked CPU though

 

[pete910]

That i9 9900K geekbench result looks pretty beefy, ~30% gain in multicore vs 8700K,
~5% higher single and ~30% higher multi vs 8700K, no IPC improvements, the single bump is due to clocks
~30% higher single and ~25% higher multi vs 2700X

The main factor on how much they can snuff AMD is going to be pricing.

A 2700 must surely crack 32K, Hell my my humble 1700x @ 3.9 does a tad over 30k. singles crap @ 5k.

 

[jigger]

Intel are not selling many HEDT chips, they are not very good, at least if the rainforest best sellers list is anything to go by they are constantly lagging well behind the competition, even Bulldozer is consistently outselling them.

This is their attempt to reinvigorate some of that mindshare, to a limited extent it probably will work, few people will be buying the 9900K as gaming CPU's but some of those looking at the 7820-X and thinking "nope i think i'll go for this Threadripper" will go back to Intel.

Short term that's great for Intel, but in about 6 months with Ryzen 3000 the 9900K is going to be obsolete, very obsolete.

Unless Intel had a new design up it's sleeves the 9900K was obsolete 12 months ago.

 

[Gerard]

https://elchapuzasinformatico.com/2018/09/intel-core-i7-9700k-review/

Alleged 9700k review, probably posted already, no idea.

 

[IT Troll]

Either they are salvaging broken 9900s and gimping them making them 9700 that work, or...

I think this is quite likely. Would make sense that 9900s which don’t meet the boost frequencies are turned into 9700s without HT and with a slight boost/clock reduction. Doesn’t make sense to intend to make a non-HT part.

They couldn’t really bin these as slower 8/16 CPUs because fast 6/12 CPUs would start to encroach.

 

[BTVA]

So 95deg flatout overclock at 5ghz 1.4v on all 8cores or 70deg 4.7/4.8?1.2v on all 8. (On that particular chip)

Going to be a pretty good performer. I wonder how much the 9900k will suffer with ht?

5Ghz @ 1.3v

 

[IT Troll]

A 2700 must surely crack 32K, Hell my my humble 1700x @ 3.9 does a tad over 30k. singles crap @ 5k.

A stock 2700X scores 26421. Looks like it needs an overclock of 4.3 to break 32K on Windows. We don’t know yet what an overclocked 9900 will score.

 

[Phil2008]

I dont know why but if I disable the HT the 6 cores run slightly hotter on my 8086k?

 

[pete910]

A stock 2700X scores 26421. Looks like it needs an overclock of 4.3 to break 32K on Windows. We don’t know yet what an overclocked 9900 will score.

Is windows that bad with ryzen? Here's mine

https://browser.geekbench.com/v4/cpu/9782552

 

[AndreiD]

Are you using any kind of performance bias option in BIOS? I know ASUS has one for geekbench explicitly which gives you a few K extra points.

 

[youdsy]

im itching to upgrade my 4790K to back up my 1080ti a bit more.

worth going for a 9700K or should i wait longer for the next die shrink?

 

[humbug]

I got near 22K on my 1600. https://browser.geekbench.com/v4/cpu/9789609

26K on a 2700X also seems a bit low to me, i think Worzel's 30K seems more realistic.

 

[SiDeards73]

Interesting results...

https://www.3dmark.com/compare/spy/4374217/spy/4386839/spy/4044761#

 

[subbytna]

Nobody, literally nobody runs them at stock? Bold statement........

First rule of vega club is not to talk about vega cl.. to undervolt them. I dont' know anyone who doesn't undervolt it? Of course there are people who run at stock levels. More a figure of speech I think...

 

[subbytna]

Anyway enough of vega in this thread. Is there a confirmed release date for the 9*** series? Is it this month? Is it a paper launch or full proper "old school" launch with chips to buy?

 

[humbug]

To re-irritate what Panos said a few pages back, and to put my own spin on it i think if you're spending £400+ on a GPU you're not an average joe user, you are an enthusiast in which case you probably understand your GPU well enough to tweak it to get the most out of it.

 

[IT Troll]

Is windows that bad with ryzen? Here's mine

https://browser.geekbench.com/v4/cpu/9782552

I got near 22K on my 1600. https://browser.geekbench.com/v4/cpu/9789609

26K on a 2700X also seems a bit low to me, i think Worzel's 30K seems more realistic.

If you browse the 2700X scores all the high scores are either running Linux or have a heavy overclock on Windows. Not sure if this is a Windows issue or just the Windows version of Geekbench.
https://browser.geekbench.com/v4/cpu/search?dir=desc&page=1&q=2700x&sort=multicore_score

The 26K score is the overall average across all user results.
https://browser.geekbench.com/processors/2080

 

[orbitalwalsh]

And

Anyway enough of vega in this thread. Is there a confirmed release date for the 9*** series? Is it this month? Is it a paper launch or full proper "old school" launch with chips to buy?

just week 40 launch for z390 for sure

 

[humbug]

If you browse the 2700X scores all the high scores are either running Linux or have a heavy overclock on Windows. Not sure if this is a Windows issue or just the Windows version of Geekbench.
https://browser.geekbench.com/v4/cpu/search?dir=desc&page=1&q=2700x&sort=multicore_score

The 26K score is the overall average across all user results.
https://browser.geekbench.com/processors/2080

So very little difference between a 1'st gen 6 core and a 2'nd gen 8 clocked higher.

I think its probably Windows yet again unable to deal with anything with more than four cores

It must be because the Linux score is hugely higher.

7.zip is the same, compression it doesn't scale at all in Windows, yet in Linux it scales perfectly.