Post Your TIMESPY, PCMARK10 & FIRESTRIKE SCORES! (2019)

[Mystic1978]

Ah, I think I have 6 first place scores and several to many #2 scores. 4070 Super. Not sure what 3DMark URL says about it. I don't worry about unofficial scores.


You may beat my CloudGate with your 5070 ti.

Here is the screeny.

Edit, yes that card is factory stock air cooler.

View attachment 398224

Wait… you picked Cloud Gate? Is that supposed to test GPUs or CPUs? That makes zero sense!
You could sell that 7900 XT and get a 5070 Ti instead. Take the first 15 synthetic benchmarks from my video and run some tests on air cooling with DDR4 and PCIe 4.0, like I did, then we can compare results, if you’re up for it

 

[ShrimpBrime]

Wait… you picked Cloud Gate? Is that supposed to test GPUs or CPUs? That makes zero sense!
You could sell that 7900 XT and get a 5070 Ti instead. Take the first 15 synthetic benchmarks from my video and run some tests on air cooling with DDR4 and PCIe 4.0, like I did, then we can compare results, if you’re up for it

I dunno, I just picked the first one the list.
Looks like CloudGate has GPU and CPU scores, so it tests both? Yes, it does!!

I've done like 200 video cards. Just about worn out benching and benching for many years.

If I had a 5070 ti and wanted to beat your scores, no doubt it wouldn't be an issue. Just as long as my sample clocks similar to yours!!

It is not doubt legacy is my sweet spot. 3DMark03 and I get along well.

I no longer have ddr4 systems. Edit

 

[Hardi]

i'm just gonna say that i got below 90k in cloud gate with 5080 and 7800X3D , so a 4070 getting over 100k shows that.. well, not real perfomance.
those legacy benches are a easy way to get top scores, but for what.

 

[ShrimpBrime]

but for what.

Specifically for...

Overclocking, overclocking, and much more! Like overclocking.

HWBOT is a site dedicated to overclocking. We promote overclocking achievements and competitions for professionals as well as enthusiasts with rankings and a huge hardware database.

hwbot.org

My profile

ShrimpBrime @ HWBOT

HWBOT is a site dedicated to overclocking. We promote overclocking achievements and competitions for professionals as well as enthusiasts with rankings and a huge hardware database.

hwbot.org

 

[Mystic1978]

I dunno, I just picked the first one the list.
Looks like CloudGate has GPU and CPU scores, so it tests both? Yes, it does!!

I've done like 200 video cards. Just about worn out benching and benching for many years.

If I had a 5070 ti and wanted to beat your scores, no doubt it wouldn't be an issue. Just as long as my sample clocks similar to yours!!

It is not doubt legacy is my sweet spot. 3DMark03 and I get along well.

I no longer have ddr4 systems. Edit

But Cloud Gate runs at such a low resolution and is so undemanding on the GPU that it obviously becomes a benchmark more dependent on the CPU than the GPU. It's like benchmarking Shadow of the Tomb Raider at 720p on low settings. I actually get more FPS with the 5070 Ti than with the 5090, but that makes no sense to me. Nobody cares about Cloud Gate or playing SOTTR at 720p low

 

[ShrimpBrime]

But Cloud Gate runs at such a low resolution and is so undemanding on the GPU that it obviously becomes a benchmark more dependent on the CPU than the GPU. It's like benchmarking Shadow of the Tomb Raider at 720p on low settings. I actually get more FPS with the 5070 Ti than with the 5090, but that makes no sense to me. Nobody cares about Cloud Gate or playing SOTTR at 720p low

What.

You're acting like Firestrike and TimeSpy where released yesterday when in fact both are like a decade old now.

Also, you can change resolution in ALL the benchmarks even down to and including 3DMark01.

So the excuse you give is irrelevant. You want a higher res, run a custom run then. But this is not a requirement to make a submission.

 

[Mystic1978]

What.

You're acting like Firestrike and TimeSpy where released yesterday when in fact both are like a decade old now.

Also, you can change resolution in ALL the benchmarks even down to and including 3DMark01.

So the excuse you give is irrelevant. You want a higher res, run a custom run then. But this is not a requirement to make a submission.

I'm not making excuses,I'm just saying that the Cloud Gate benchmark isn't a good reference to compare a 5070Ti with a 4070 Super. If you want to compare your 4070 Super's results, there are 15 other benchmarks on the list you can use. Cloud Gate, especially at the low resolution you ran it, is more of a test of how fast is your CPU and RAM rather than how well the 4070 Super actually performs. If you really want to see what your 4070 Super is capable of, run Port Royal, Steel Nomad, or Speed Way, and then we can compare results.

Or you could just buy a 5070 Ti and settle the debate. I've already had two 5090s, one 5080, and two 5070 Tis, there's plenty of stock available if you're looking to buy one.

This is a comparison in Port Royal between the top 1 result with the 4070 Super and my best Port Royal score with the 5070 Ti

A 50.7% difference in favor of the 5070 Ti

 

[ShrimpBrime]

I'm not making excuses,I'm just saying that the Cloud Gate benchmark isn't a good reference to compare a 5070Ti with a 4070 Super. If you want to compare your 4070 Super's results, there are 15 other benchmarks on the list you can use. Cloud Gate, especially at the low resolution you ran it, is more of a test of how fast is your CPU and RAM rather than how well the 4070 Super actually performs. If you really want to see what your 4070 Super is capable of, run Port Royal, Steel Nomad, or Speed Way, and then we can compare results.

Or you could just buy a 5070 Ti and settle the debate. I've already had two 5090s, one 5080, and two 5070 Tis, there's plenty of stock available if you're looking to buy one.

This is a comparison in Port Royal between the top 1 result with the 4070 Super and my best Port Royal score with the 5070 Ti

A 50.7% difference in favor of the 5070 Ti

View attachment 398302

No I get that. The 4070 cant be measured against 5070. Total agreement.
But you know a 5900X is a slow dog compared to almost all the Raptor parts right?

I also agree that a lot of benchmarks are CPU intensive like legacy benchmarks, which I did note I prefer benching. Because that's where I started.... back in the day ya know??

My original comment was kind of joking like. Cause people will see, notice, admire, and then beat your scores. All of them. And I don't need to buy a card to do it, someone will cover your scores. Just how it is young brother. So like my 7900 XT for example scores decent with lower clocks today because today's drivers are much better than when it was released. So in time, you may have to "re-run" all these scores on a different driver perhaps to get back that top placement.

There's a lot of depth to competitive overclocking. I have the experience to say I have the confidence to beat your scores if you sent your entire rig over. Might even have some tricks up the sleeve, which honestly I don't do much tweaking like that anymore. In some cases, it's a gross waste of time. But that really depends.

realistically, some benchmarks do better on AMD processors if we are getting into modern stuff strictly speaking. A 9950X/3D pretty much spanks Intel at the same clock frequency (I've seen plenty of 6ghz from both), in which some benchmarks just do better on Intel, and some do better on AMD.

Lastly, you should take notice to the CPU clock speeds I'm running some benchmarks at. The reason my CloudGate score is so high, isn't the GPU. It's all front end low latency and high cpu clocks. But a lot of low end GPUs I've done with basically a stock 13700K and just ran it. So that depends on the day really, where the CPU is at. Today, we are testing the board with an air cooler after last night's DryIce runs. I was benching well above your daily air cooled clocks.

So to example, 8c 16t 14900K at 6600mhz 3DMark05. Yes, this configuration with your 5070 ti at the same clocks would likely kill many of your submissions, modern benchmark or not. No I don't want to run air against your air. We are in it to win it. So you need DryIce or LN2. Then, we go fast

 

[Mystic1978]

No I get that. The 4070 cant be measured against 5070. Total agreement.
But you know a 5900X is a slow dog compared to almost all the Raptor parts right?

I also agree that a lot of benchmarks are CPU intensive like legacy benchmarks, which I did note I prefer benching. Because that's where I started.... back in the day ya know??

My original comment was kind of joking like. Cause people will see, notice, admire, and then beat your scores. All of them. And I don't need to buy a card to do it, someone will cover your scores. Just how it is young brother. So like my 7900 XT for example scores decent with lower clocks today because today's drivers are much better than when it was released. So in time, you may have to "re-run" all these scores on a different driver perhaps to get back that top placement.

There's a lot of depth to competitive overclocking. I have the experience to say I have the confidence to beat your scores if you sent your entire rig over. Might even have some tricks up the sleeve, which honestly I don't do much tweaking like that anymore. In some cases, it's a gross waste of time. But that really depends.

realistically, some benchmarks do better on AMD processors if we are getting into modern stuff strictly speaking. A 9950X/3D pretty much spanks Intel at the same clock frequency (I've seen plenty of 6ghz from both), in which some benchmarks just do better on Intel, and some do better on AMD.

Lastly, you should take notice to the CPU clock speeds I'm running some benchmarks at. The reason my CloudGate score is so high, isn't the GPU. It's all front end low latency and high cpu clocks. But a lot of low end GPUs I've done with basically a stock 13700K and just ran it. So that depends on the day really, where the CPU is at. Today, we are testing the board with an air cooler after last night's DryIce runs. I was benching well above your daily air cooled clocks.

So to example, 8c 16t 14900K at 6600mhz 3DMark05. Yes, this configuration with your 5070 ti at the same clocks would likely kill many of your submissions, modern benchmark or not. No I don't want to run air against your air. We are in it to win it. So you need DryIce or LN2. Then, we go fast

View attachment 398307

I hear you, and I totally get the legacy overclocking passio, it's clear you’ve been around the scene for a while. But just to clarify, my point was never to dismiss older benchmarks entirely, just to say that Cloud Gate in particular isn’t the best reference for comparing raw GPU performance between a 5070 Ti and a 4070 Super. Especially at low resolution, it leans more on CPU and RAM latency than actual GPU throughput.


About the 5900X with the 4070 Super in port royal ( that guy was top 1 in the overal lederboards with any cpu in any country ) and yeah, it's not top-tier by today’s standards, but again, in benchmarks like Port Royal, Steel Nomad, or Speed Way, CPU and RAM have very minimal impact unless you're bottlenecking hard. I’ve tested that myself by loosening and tightening my DDR4 RAM and pushing the CPU harder, and the difference was barely noticeable in those runs.


And hey, I respect your confidence and experience, no doubt you’ve got the skills and tricks from years of benching. But in this case, I’m just comparing results in the same test environment, using currently available hardware, without exotic cooling. Of course, with LN2 or Dry Ice, the game changes entirely.


I’m not trying to claim some eternal record, just presenting data from comparable runs, under similar conditions. That's all. If anything, this is good fun, and if you or someone else beats my scores later on, all the better. That’s the spirit of the scene.

I'm 47 years old and I've been tinkering with hardware since I was 6. You might have underestimated me, my friend.

 

[ShrimpBrime]

I hear you, and I totally get the legacy overclocking passio, it's clear you’ve been around the scene for a while. But just to clarify, my point was never to dismiss older benchmarks entirely, just to say that Cloud Gate in particular isn’t the best reference for comparing raw GPU performance between a 5070 Ti and a 4070 Super. Especially at low resolution, it leans more on CPU and RAM latency than actual GPU throughput.


About the 5900X with the 4070 Super in port royal ( that guy was top 1 in the overal lederboards with any cpu in any country ) and yeah, it's not top-tier by today’s standards, but again, in benchmarks like Port Royal, Steel Nomad, or Speed Way, CPU and RAM have very minimal impact unless you're bottlenecking hard. I’ve tested that myself by loosening and tightening my DDR4 RAM and pushing the CPU harder, and the difference was barely noticeable in those runs.


And hey, I respect your confidence and experience, no doubt you’ve got the skills and tricks from years of benching. But in this case, I’m just comparing results in the same test environment, using currently available hardware, without exotic cooling. Of course, with LN2 or Dry Ice, the game changes entirely.


I’m not trying to claim some eternal record, just presenting data from comparable runs, under similar conditions. That's all. If anything, this is good fun, and if you or someone else beats my scores later on, all the better. That’s the spirit of the scene.

I'm 47 years old and I've been tinkering with hardware since I was 6. You might have underestimated me, my friend.

A little bit, my bad. Some of the way it's worded sounds like modern green type stuff. Had another guy knocking wprime, but fails to realize prime digits is used by a lot of people in a non competitive environment. So if I sound abrasive, that's not the intent.

Your subs are good. They rock all of mine, I haven't gotten a card with that much muscle yet. I try to cap the hobby spending at a certain dollar amount. Otherwise, I'd love to bench and compare with you! 100%.

I don't even like TimeSpy. Honestly. Seems to take forever lol. Is it done yet?

So Speed Way, Steel Nomad ect, GPU mostly. Get the core clocks up. Some benches like Vmem OC a lot too. That guy with 4070 S in Port Royal probably had 300mhx GPU clocks over mine. Think I ran many at only 2750mhz because I use that card daily. The 7900 XT isn't even as fast as that card and almost double the money

The only logical thing for me to do is NOT give NV more money directly for a 5070 TI. I don't game. It would be useless. When they are 250 bucks, I'll grab one in a few years maybe.

 

[Mystic1978]

@ShrimpBrime​

On the other hand, I should point out that some people are already using more extreme cooling. Take Burti_TecLab, for example, he had his 5070 Ti running at just 5°C, likely with a hardmod or shuntmod given his rock-solid average frequencies. He paired it with a 9800X3D running close to 5.7GHz and DDR5 memory on PCIe 5.0. Meanwhile, my result was on air, averaging 54°C with a 14900KF running just at 5,9Ghz and DDR4 on PCIe 4.0.


In the end, it’s clear who optimized their setup better. It's not about going to extremes, competitiveness doesn't always require going that far.

In my opinion, the comparison is almost embarrassing, and not for me, but for the #2 spot. Even with more extreme methods, they still couldn’t reach the top position. Being competitive isn't that serious, man... Anyway, I’ll leave the debate here. A word to the wise is enough.

 

[ShrimpBrime]

@ShrimpBrime​

On the other hand, I should point out that some people are already using more extreme cooling. Take Burti_TecLab, for example, he had his 5070 Ti running at just 5°C, likely with a hardmod or shuntmod given his rock-solid average frequencies. He paired it with a 9800X3D running close to 5.7GHz and DDR5 memory on PCIe 5.0. Meanwhile, my result was on air, averaging 54°C with a 14900KF running just at 5,9Ghz and DDR4 on PCIe 4.0.


In the end, it’s clear who optimized their setup better. It's not about going to extremes, competitiveness doesn't always require going that far.

In my opinion, the comparison is almost embarrassing, and not for me, but for the #2 spot. Even with more extreme methods, they still couldn’t reach the top position. Being competitive isn't that serious, man... Anyway, I’ll leave the debate here. A word to the wise is enough.

View attachment 398318

Shunt mod is very popular in the gaming community actually. I see it around a lot.

Correct. HwBot and OC competitive is not what it used to be.

cheers

 

[Mystic1978]

Shunt mod is very popular in the gaming community actually. I see it around a lot.

Correct. HwBot and OC competitive is not what it used to be.

cheers

I just found the video of Burti_TecLabof and his modded 5070Ti:

View this content on Instagram

View this content on Instagram

So in the end, being number #2 using LN2 with a modded 5070Ti and me beating him with DDR4 on air and under PCIe 4.0... As i said before is not that serious my man...

View this content on Instagram

View this content on Instagram

And that's just in Speed Way, in the other 15 benchmarks I haven't even seen him show up.
Of course, he probably used air cooling, and he's nowhere near my results.
He needs to mod the PCB and use LN2 to even be competitive.
It's clear he has no damn clue how to properly tune his test setup.

 

[ShrimpBrime]

I just found the video of Burti_TecLabof and his modded 5070Ti:

View this content on Instagram

View this content on Instagram

So in the end, being number #2 using LN2 with a modded 5070Ti and me beating him with DDR4 on air and under PCIe 4.0?... As i said before is not that serious my man...

View this content on Instagram

View this content on Instagram

But as you said, nobody really cares. I find much of the opposite frankly.
You claim beating this dude, but say "not that serious man"... and continue praising yourself.

And I want to commend you on the win! Seriously I know it can be very difficult. It's like saying you've submitted 14900KF at HWBot and in the same boat as myself, not getting on the first page let alone top 10 without LN2. Cooling leagues with mixed point system. How on air are you ever going to really win? I have a FX 9590 submission 5.7ghz on water, fastest recorded. Years ago. Nobody cares. But, we have that to share

Keep up the good work. It's no big deal, and we do it for fun. That is exactly the fastest team's on the planet's 1st main rule. Just enjoy yourself

I enjoyed the chat with you. You're good peeps.

 

[Mystic1978]

But as you said, nobody really cares. I find much of the opposite frankly.
You claim beating this dude, but say "not that serious man"... and continue praising yourself.

And I want to commend you on the win! Seriously I know it can be very difficult. It's like saying you've submitted 14900KF at HWBot and in the same boat as myself, not getting on the first page let alone top 10 without LN2. Cooling leagues with mixed point system. How on air are you ever going to really win? I have a FX 9590 submission 5.7ghz on water, fastest recorded. Years ago. Nobody cares. But, we have that to share

Keep up the good work. It's no big deal, and we do it for fun. That is exactly the fastest team's on the planet's 1st main rule. Just enjoy yourself

I enjoyed the chat with you. You're good peeps.

Thanks, I really appreciate your words. You're right, at the end of the day, it's all about having fun and enjoying the process. But I also think it's worth highlighting that sometimes people overlook what’s actually being achieved with limited setups.


Beating someone on LN2 while just on air and DDR4 isn’t “praising myself”, it’s just putting things in perspective. Not everyone can or wants to go full LN2 and mod PCBs just to chase ranks. When someone gets close or even passes others without extreme cooling, that should be acknowledged too.


And yeah, you're absolutely right, most people won't care, especially outside the benching crowd. But for those of us who do this, we do notice.


So, thanks again, and same to you: keep pushing, keep enjoying it, and let the scores speak when they need to

 

[ShrimpBrime]

Thanks, I really appreciate your words. You're right, at the end of the day, it's all about having fun and enjoying the process. But I also think it's worth highlighting that sometimes people overlook what’s actually being achieved with limited setups.


Beating someone on LN2 while just on air and DDR4 isn’t “praising myself”, it’s just putting things in perspective. Not everyone can or wants to go full LN2 and mod PCBs just to chase ranks. When someone gets close or even passes others without extreme cooling, that should be acknowledged too.


And yeah, you're absolutely right, most people won't care, especially outside the benching crowd. But for those of us who do this, we do notice.


So, thanks again, and same to you: keep pushing, keep enjoying it, and let the scores speak when they need to

Praise may have been the wrong word, but you catch the drift. It wasn't really a jab, I'm pointing it out. I suppose I was numb to it because I've done a lot of what you are doing! I mean to beat one of Splave's submissions. (this guy top US bencher) I've done that! And it's like pure fucking luck too haha.

DDR4 can be fast when tuned well. I have just as much fun at 4600mhz as I did at 4000mhz fast and tight. But I'm not a memory guru, and I don't buy 350$ kits. I just like CPU frequency. It's always been my thing. Make cpu fast. Weeee. 7Ghz over the weekend. Cost, 20 bucks of DryIce.

 

[Ruru]

Never got 17k in Timespy before. I guess my CPU was power restricted previously.

I scored 17 058 in Time Spy

AMD Ryzen 7 5800X, NVIDIA GeForce RTX 3080 x 1, 32768 MB, 64-bit Windows 11}

www.3dmark.com

 

[hasegade]

Link 3D Mark
rx6800 + 14600kf

Your CPU is bottlenecking the graphics card

 

[NBrock]

9800X3D and MSI Suprim 5090