KarathKasun

7 minutes ago, Archyn said:

Nope, just the 1tb ssd NVME drive, so i don't think theres anything wrong with it

Any other USB devices other than the KB and mouse?

You may have a failing HDD/SSD or a failing USB device.  I have an old USB mic that has a broken USB connector, it can caause similaar things to happen.

Do you have any other storage devices connected with SATA or USB?

5 hours ago, ShadySocks said:

The larger SoC die (with iGPU)

What part is the image of?

Was asking about the OP image, sorry.

Found all of the info.

Source...

https://en.wikichip.org/wiki/intel/microarchitectures/ice_lake_(client)

Larger square die is CPU+GPU.

SoC die shot

Smaller rectangular die is the PCH.

PCH die shot

IGPU is not on a different die on any current parts.  Older Intel parts looking like this had an external eDRAM die for cache that could be used by the IGPU or CPU cores on the CPU die.  The IGP was still on the CPU die though.  These were known by the codename crystal well AFAIK, the marketing term was Iris Pro.  Im going to go out on a limb and guess that the seccond die in that shot is eDRAM and not the GPU itself.

Even further back on 1st gen i7 parts the memory controller and IGP was housed on a seperate die in the CPU package, but this was 10+ years ago now.

Neither.

What is that supposed to be a picture of?

On 9/24/2018 at 11:25 AM, Stefan Payne said:

That's a stupid idea because in a car the nominal voltage when the generator is running is aroudn 14V, not 12.

And its very noisy as well. so you need something that cleans the shit up that is thrown around in the car.

 

That is why you woulrn't recommend a "Pico PSU" for RV duty!!

Those seem like a viable option, though pricey...

And a bit heavy on the +5V rail...

At least they seem to clean up the shit from the primary side...

Apologies for necro quote, but Ive run across a bit of this during the research stage of setting up an automotive application PC.

Pico PSUs have been in use for CarPCs for YEARS, and are generally specced to handle ~14v input.  If you really feel that noise is a problem you can always just add a few huge bulk caps on the input side.

HDMI -> VGA is expensive, what outputs does your GPU have?  DP -> VGA stuff is generally cheaper, as is DVI -> VGA.

2 minutes ago, Mister Woof said:

Yeah, I'm just saying there's no improvement for real life users because usually only enthusiasts buy k chips and most enthusiasts overclock anyway so the stock clock unnoticed m improvements are lately meaningless.

Im going to say that 90% of K series chips never get overclocked beyond the auto max turbo setting on motherboards.  I do computer service and see TONS of K series chips in H/B series boards.

"It costs more money so its better, and that one streamer guy uses it"

Just now, aDoomGuy said:

I tried Minecraft RTX yesterday on my machine, I got 20-25 FPS because my GPU was in 2D mode. I put it in always 3D mode with GPU Tweak and I was up at 75 FPS. Minecraft doesn't induce much GPU load so cards don't see the need to go into 3D.

AFAIK you can do this in the nv control panel too, under the power management setting.

Just now, Mister Woof said:

In gaming I don't expect a real world difference between equally clocked 8700k and 10900k tbh

Nope.  They use the same core, if you can get equal clocks you get equal performance.

You shouldnt be looking at more than 8 cores for gaming anyway.

3 minutes ago, AshRiver said:

I wouldn't say 25% at minimum though. My bet is roughly 15% for the 14nm. In gaming probably 5% difference.

Added numbers from the leaks.  31%.  i9 isnt a gaming CPU TBQH, its a low end workstation CPU.

6 minutes ago, AshRiver said:

How fast are we talking? 10% fast for bleeding edge 14mn, a new motherboard, and $500ish cpu?

Well, 2 more cores... so 25% at minimum.  Clocks are higher too, so likely closer to 35%.

CB R15 scores...

9900k - 2030

10900k - 2670

Look at that, 31%.  Ive gotten pretty good at pulling numbers out of thin air over the last 25 years.

Just now, 5x5 said:

Well, the 250W stock i9 in the news section is reminiscent of the 9590

Except its actually significantly faster overall than the lower wattage chip it replaces.

1 minute ago, Mister Woof said:

It was better than FX though lol

 

I had both an AMD Phenom IIx4 965 and later an FX-6300.

 

The Phenom played games better.

 

Intel may be only moving at a snail's pace, but at least they aren't going BACKWARDS.

They are moving at the A64 -> Phenom II pace currently, but they are doing dev work on something new.  I'd put money on it.

3 minutes ago, 5x5 said:

12? You mean 6. Phenom II was competitive up until 2011. Also my point is intel have been resting on their laurals for a long time because they were the market leader and now they're getting punished for selling the same old CPU under a new IHS

12.

Phenom II was not considerably better than Phenom I outside of much better clocks.  Phenom I was A64 with some tweaks and four cores per package.  BD was slower than PH I/II per clock.  AMD coasted on the A64 design for nearly 15 years in total.

Check real benchmarks.  Minecraft is EXTREMELY sensitive to things like render distance, shader packs, and map location.

2 minutes ago, 5x5 said:

Skylake came out in 2015, though, and intel haven't been able to raise IPC since Haswell in 2013. The Skylake performance came from DDR4 and memory management, not a CPU innovation on the chip itself. The ringbus design is even older dating back to 2011 with Sandy. They hVent done anything meaningful for around 7 years at least

And AMD hadnt done anything of real merit for 12 years before Ryzen (excepting that they blew a ton of cash on hookers and blow).  Whats your point?

2 minutes ago, 5x5 said:

Intel have been vocal that the earliest we can expect a new uarch is 2022. Meaning they were literally circlejerking themselves while Zen+ was already here

2017 to 2022 sounds a lot like ~5 years.  Which is about the time needed for a new ground-up architecture.

1 minute ago, 5x5 said:

And yet, 4 years after the launch of Zen, intel are still on Skylake+++++++++. I think their workforce is slacking quite a bit. Maybe if the CEO wasn't busy with his secretary and I sider trading, intel wouldn't be in a hole (Kraznich is a meme at this point)

Quote

Intel has problems that stem from basic architecture design concepts that can only be addressed with a VERY low level redesign.  That kind of thing takes the better part of 5 years to push out.  Ryzen wasn't something that just happened, it was the reason that AMD dropped any work on BD/PD core designs.  It took 5+ years of work to get 1st gen Ryzen to market, and it still had major issues when it launched.

This is why they are still using Kaby Lake cores with no real improvement.  They are 100% working on something from scratch new while the minor dev teams just copy and paste cores onto existing designs.

What are you powering with PoE?

Nevermind, that should work fine.

Just now, GDRRiley said:

It’s not always low hanging fruit. It’s just a push to make the chips actually faster so they can gain market share. they also though have 2 teams working on chips that leapfrog. Intel has at least 4 right now. (Cove, lake, and 2 others I can’t remember)

Not really as much as you think.  Intel has problems that stem from basic architecture design concepts that can only be addressed with a VERY low level redesign.  That kind of thing takes the better part of 5 years to push out.  Ryzen wasn't something that just happened, it was the reason that AMD dropped any work on BD/PD core designs.  It took 5+ years of work to get 1st gen Ryzen to market, and it still had major issues when it launched.

AFAIK Intel has a central development team for the basic architecture and other teams that take that and twiddle it for specific markets.  Intel also has 10x the workforce.

1 minute ago, GDRRiley said:

It’s not really IPC most of that was other changes and you also had a memory swap in there. 

Each generation of ryzen is looking like ~15% 

(not series but proper generation)

We saw that on early Core series CPUs too (Core 2 -> Nehalem -> Sandy Bridge), its just flatlined since the Kabylake.  Ryzen is still a new architecture with lots of low hanging fruit.

9 minutes ago, GDRRiley said:

given intel was yelling look we got you 2% it is big compared to most gains from 4770k to a 9900k

Err, no.  4770k to 7700k (same basic core as 9900k) was actually quite significant in a lot of places.  A 15% gain was not unheard of at similar clocks.

I have chips from most intel generations from 2600k to 7600k, the gains were there.  Its just that games didnt always see the largest part of them.

Ryzen was such a big jump for AMD because BD/FX was such crap when it was new and didnt get any updates for ~10 years.

2 minutes ago, GDRRiley said:

it isn't a tiny amount a 4.3ghz amd chip is very close to a 5.0ghz+ intel chip on single core.

given zen 3/4000 is expecting higher IPC and clocks intel has 0

 Its 7% give or take a little.  That is tiny.

4.3ghz -> 5ghz is a 16% increase.  AMD is still ~10% behind in many workloads.

Going to point out that architecture itself isnt what makes emulation better this time around.  Its the clocks.

Ryzen actually pushes more ops/clock than Intel by a tiny amount.

Sandisk has one problem.  They make cheap trash all the way up to outstanding products.  Just because it says Sandisk on it does not mean it will be great.