Humbug

 

 

https://www.techradar.com/news/life-in-the-queue-meet-the-gamers-camping-out-for-an-rtx-3080-ti-across-the-us

 

 

 

 

 

7 minutes ago, GDRRiley said:

I've heard some rumors say its already better than DLSS 1.0 or lower resolutions with AA (which it better be) as they are low bars
but also that its a lot easier to add in

Ya. The rumors I heard are that the image quality is better than DLSS 1.0... but that DLSS 2.0 quality mode still has an edge

I suspect they are going to continuously improve their implementation over time. So even if the launch version does not have machine learning AI as detailed in this patent, I bet it would come down the line as a way of improving the image quality further....

Practical example. After an upgrade a 1660 TI getting worse fps than an old R9 390. CPU ivy bridge i7-3770K.

https://youtu.be/NWmConRMbD4

So few games tested, too few to get a representative sample.

LTT kinda falling behind on GPU reviews...

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

https://live.staticflickr.com/65535/50527558942_8c0f0235be_o.jpg[/img

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Don't cap the framerate, and do not use vsync.

 

Just turn on Enhanced Sync in the Radeon Settings along with Freesync for maximum smoothness. That allows the game engine to run unlocked with maximum fps. Even if it exceeds the monitors maximum refresh rate there will not be tearing as the GPU only sends the display the newest completed frame.

 

 

Just now, ShrimpBrime said:

Go to Trc control enable it and then it should allow you to change the timing to what you select or type in the number pad.

 

Just now, TofuHaroto said:

^

Ok I did that, below is what I did in the BIOS. But look at what it says about the range of possible values on the right. btw when I enabled that control it defaults to 39 before I change it. Anyway I changed it to 58.

 

 

After I set it to 58 and log back into windows below is what I get. tRC has actually gone up further.

 

23 minutes ago, ShrimpBrime said:

Yes he can.,

Just needs to click the "advanced timings configuration tab. You can see the option on the second screen shot in the first post.

 

21 minutes ago, TofuHaroto said:

Didn't see that lol

 

When i click on Advanced DRAM configuration I get the below. I can only change the items under Main Timing Configuration.

From Sub Timing Configuration downwards everything is greyed out. Regardless of whether I am in XMP mode or not.

 

 

 

If I go the advanced overclocking DRAM section I can see the below which looks like greek to me. Can I change the tRC to 58 from here? If so how?

 

15 minutes ago, TofuHaroto said:

Try more than one monitoring system

For cl latency it could happen if the imc needed to up the cl latency and couldn't get to the advertised speeds with the cl19 but the trc is really really weird Maybe try hwinfo64 to monitor the ram speeds

Verified with Ryzen Master. It confirmed all the CPU-Z readings.

 

15 minutes ago, TofuHaroto said:

And update the bios 

BIOS is up to date.

 

Also as mentioned in OP my system runs fine even at 16-19-19-19-39, which is even a lower CL than what my modules are rated for, yet it's stable outside the XMP profile...

But the tRC value is still really high.

Hello

 

I would like some assistance in getting my RAM to match the XMP spec exactly.

 

The RAM modules I own are these

2 x 8GB DDR4 3600 T-Force VULCAN TUF Gaming Alliance DDR4

CL19-19-19-39

https://www.teamgroupinc.com/en/product/vulcan-tuf-ddr4

 

It shows up correctly under the SPD tab of CPU-Z as the only XMP profile.

 

 

 

So then I enable XMP in the BIOS and it looks like the correct option in the BIOS as shown below.

 

 

 

However when I log back into Windows and check CPU-Z I can see the below issues.

CL is 20 instead of 19

tRC is 82 instead of 58

 

 

 

In the BIOS I cannot figure out how to tweak the TRC, I only have below options under DRAM configuration.

 

 

 

If I go to the advanced overclocking DRAM section I get this which I cannot make head or tail out of cause I am a noob to RAM.

 

 

 

Now there is an option under DRAM Overclocking called "Memmory Try It !" where you can pick from a list of presets. So for example I chose the below which works fine.

 

 

 

It worked fine but as shown in CPU-Z below my tRC is still at 82 even though the other timings look good.

 

 

 

System specs

Ryzen 5 3600

MSI B450 Gaming Pro Carbon AC WiFi

2 x 8GB DDR4 3600 VULCAN TUF Gaming Alliance DDR4 GAMING MEMORY

AMD Radeon R9 290 Sapphire Vapor-x

 

Seems quite likely. AMD is known to use the new architecture for the high end SKUs and rebrand last gen parts for the rest.

 

I do hope there will be a big perf/dollar jump though. I.e. I hope RX 5700 XT performance is available for USD 200. That will allow people to build fast 1440p gaming PCs for cheap.

 

Also AMD better have partner card designs available at launch. This is a very competitive launch because Nvidia is also launching at the same time, so AMD needs to give buyers options.

I bought a Sapphire R9 290 vapor-x in 2014.

It's still kicking ass in 2020 on the newest AAA games.

Updated: 28th March 2020: Low Quality Testing on old and budget GPUs

 

Please note that low quality in this game still looks great, you don't lose much visually. Shockingly bad results on Kepler. A former flagship GTX 780ti is losing to an even older low end Radeon HD 7790.

 

https://www.techspot.com/article/2001-doom-eternal-older-gpu-test/

Vega 56.

It's already faster today and will age better with the extra VRAM and AMD's good long term driver support.

 

 

 

14 minutes ago, BuckGup said:

People are going to buy a game regardless if it runs super well or not. So it makes sense from their perspective to spend more on a better story and content than better devs 

Debatable. Being well technically polished certainly is of value to the games reputation, especially on PC.

No doubt games like Assasin's Creed Unity and Batman Arkham Knight took a big sales hit because of bad performance, word gets around.

6 minutes ago, Froody129 said:

From what I've seen this game is optimized really well. Hopefully it becomes the new trend

I guess unlike id software most developers do not have the expert talent to really optimize graphics, including making the most of a low level API.

This is the follow up to the acclaimed 2016 game. Doom Eternal by id software uses the idtech 7 engine. Graphics pipeline is natively Vulkan based, no fallback option for older DirectX or OpenGL versions. id software has developed a good reputation for being extremely technically proficient, the John Carmack legacy is still alive in the company. Their games tend to perform smoothly while looking good and making the most of available resources.

 

Doom Eternals graphics are a step up from the previous entry. Please note the below benchmarks are absolutely maxed out, so you should be able to get quite a boost by optimizing the settings.

 

Even a 3GB GTX 1060 can do 60fps at 1080p maximum settings. Something like a  Vega56 easily clears 120fps.

 

An RX 5700 or RTX 2060 Super gets very close to 1440p 120fps maximum settings.

 

A Vega 64 or RTX 2060 super can do 4K 60fps maximum settings.

 

At the highest settings the game justifies AMD and Nvidia's move towards 8GB VRAM.

 

note- graphics on youtube always look a bit worse due to video compression.

Source

https://www.techpowerup.com/review/doom-eternal-benchmark-test-performance-analysis/

 

******************

Updated: 28th March 2020: Low Quality Testing on old and budget GPUs

 

Please note that low quality in this game still looks great, you don't lose much visually. Shockingly bad results on Kepler. A former flagship GTX 780ti is losing to an even older low end Radeon HD 7790.

 

https://www.techspot.com/article/2001-doom-eternal-older-gpu-test/

On 3/3/2020 at 2:19 AM, BlackManINC said:

I'll need to see some parts with more fast paced action like in previous iterations, some sprinting sections at least to see how well this system really works.

You are thinking in terms of regular 2d monitor games, and trying to map your wants from traditional fps games.

 

The level of immersion playing a game with this production quality on a mid or high end PC with a good VR headset is next level. You feel like you are in the world, not looking into a window. For example when an NPC comes up to you you are very aware that there is somebody life sized standing next to you because your brain has bought into your presence inside the 3d environment. Combine that with good story motivation, gameplay mechanics, music, some horror elements etc... Slow paced, with tension, exploration of the world will work just fine.

 

 

 

 

 

 

 

 

 

 

 

[

 

 

 

 

 

 

 

 

 

 

 

 

 

wow seems like lots of positive feedback so far on this release

 

reddit link 

1 hour ago, Guiltyx said:

Thoughts? Maybe AMD is finally going to overtake Nvidia.

 I have little doubt that AMD will indeed beat the RTX 2080ti without too much trouble. The RDNA architecture has been a huge step forward in terms of how much work it gets done per compute unit, and that is just RDNA 1.0 which we have seen thus far- in the form of small midrange GPUs with only 36-40 CUs. The bigger ones should be really fast with 60-80 CUs.

 

The thing is it is so late! They will hold the crown just for a couple of months and then Nvidia will launch their next gen stuff as well.

 

What AMD can do is bring sanity back to PC gaming in terms of pricing. By launching their full RX 6xxx series lineup consisting of high end RDNA 2.0 parts as well as rebranded RDNA 1.0 parts for the midrange. And they can market it as the saviour of PC gaming if they selll the top end SKUs for $700 while beating the RTX 2080ti. Have a huge marketing push around that.

 

But alas it seems that AMD too is following in Nvidia's footsteps in pushing PC gaming more and more out of reach. They will just undercut Nvidia by a small amount going by recent trends.

 

  

19 minutes ago, Medicate said:

Putting HBM2 on that card, and 24GB of that, would just be AMD shooting themself in the foot. There's no reason for such an expensive and big frame buffer on gaming cards.

That could even be the fully unlocked part designed to be sold as a workstation part for USD 2000, which will be cut down for gamers... we really don't know.

What we do know is that RDNA has been designed to support both types of memmory so AMD has that flexibility.

 

1 hour ago, TheUberMedic said:

It's strange. The last time I had an AMD GPU was the RX 480 and that thing ran really smoothly with very few crashes. Compared to my last 2 GPUs (laptops with GTX 1070 and RTX 2070), these crash ALOT more than the 480.

Ya in my experience too AMD has been a bit better than Nvidia for reliability, but judging by reports from others it seems like the last couple months have had an uptick in issues.