[SOLVED] Should i feel scammed? Lenovo prebuilt timespy results

[MOgens1337]

I bought a prebuild LENOVO LEGION T7 34IMZ5 I7, on a black friday sale (you can see the full specs in my timespy links). I wanted to build my own, but with gfx prices these days and with the deal i got, it really made sense. I also saw alot of reviews saying that the lenovo prebuilds have good thermals, and i must say i agree. When i use a custom fan curve (Nothing too aggressive or noisy) in msi afterburner my temps are very good. However, when i run timespy, i get discouraging results (or at least i think so) Everything is run at stock speeds in my following links, however i think the link where im running vantage (Inbuilt oc/monitoring program from lenovo) i think there is a slight overclock. However you can't control the fans from there, and my thermals are way better when running afterburner (But results are worse since no oc yet), so i would like that to be my main tool and not vantage.

http://www.3dmark.com/spy/24647592 (With vantage running)
http://www.3dmark.com/spy/24639073 (With afterburner running instead)

So my gfx score is about 1k below average, is my card just a really bad one, or is everyone overclocking just a bit before they run timespy?
oh and btw. my nvidia settings are all set to perfomance, no AA and 1 prerendered frame. I tried taking off low-latency mode (More than one pre rendered frame), which actually gave a worse score, that surprised me, as i thought that should give more fps.

Anyways, thanks in advance! You are a great community and i have learned alot from reading your posts through the many years.

 

[Eximo]

Probably not the best cooled GPU out there. Certainly not one of the high end ones, and it probably doesn't have a factory overclock. So it will boost less than other cards even if the temperatures are good.

 

[logainofhades]

People are probably overclocking, for timespy, as they want to try and get on the leaderboards.

 

[MOgens1337]

I get that you wont get the best silicon in a prebuilt, but almost 1k below average just seems like alot. How much does these score matter in a game scenario tho? Also since my temps are comfy in the 50 c range, that should leave me some headroom for a light oc right?

 

[MOgens1337]

People are probably overclocking, for timespy, as they want to try and get on the leaderboards.

But their core clock and hz numbers dont seem very high, and is it really reasonable to think that the average, is the product of overclocking?

 

[logainofhades]

Do you have an actual model number, for the gpu? It might be a model with not so great cooling.

 

[Eximo]

Do you have an actual model number, for the gpu? It might be a model with not so great cooling.

Or a fixed power limit to keep it quiet.

 

[logainofhades]

Or a fixed power limit to keep it quiet.

That is also possible. Kinda like Evga locking out the 2060 KO's power limit.

 

[MOgens1337]

Do you have an actual model number, for the gpu? It might be a model with not so great cooling.

Or a fixed power limit to keep it quiet.

Im trying to find the actual gpu model, but having trouble doing so. The power slider in afterburner is at 100% from default - and cant go higher. Don't know if that means that there is a fixed/ or low power limit

 

[faalin]

You have to remember that they lump all 3080 together in one listing whether they are overclocked or not.

So your card Graphic score 16397
Clock 1965MHz
Memory 1188Mhz

Vs

Someone overclocking Graphic score 21560
Clock 2415Mhz
Memory 1314Mhz

Its not fair to compare scores if the cards are not on equal grounds. You would have to search through all the scores till you find someone that is running there card at stock settings to compare scores. That is why your score is only 98% better then most 3080 cards

 

[MOgens1337]

If i run timespy, and look at the statitics from afterburner, i should be able to see the actual power draw right?

 

[Eximo]

A locked slider is a good indicator of a locked down vBIOS. That might be something you can deal with if you want to take that risk. Other MSI vBIOS are likely compatible, but then you would have to be careful with temperatures.

And if the GPUs were binned, this might be a poor performer and thus relegated to an OEM card. Binning costs money though, so cheap cards are sometimes just assembled from a tray of GPUs from Nvidia without checking for good ones (though at this point I think Nvidia does their own binning to keep founder's edition silicon at the top)

 

[geofelt]

Does this pc play the games you want with acceptable performance?

Best, I think to ignore such comparisons.
Those with good results will post.
Those with dogs remain silent.

Considering how hard it is to find any 3080, you probably did well.
The cpu is no slouch either.

 

[MOgens1337]

You have to remember that they lump all 3080 together in one listing whether they are overclocked or not.

So your card Graphic score 16397
Clock 1965MHz
Memory 1188Mhz

Vs

Someone overclocking Graphic score 21560
Clock 2415Mhz
Memory 1314Mhz

Its not fair to compare scores if the cards are not on equal grounds. You would have to search through all the scores till you find someone that is running there card at stock settings to compare scores. That is why your score is only 98% better then most 3080 cards

When i look through peoples scores, i do find rigs that have like 1k better gfx score than me, but around the same clocks and Mhz, maybe slightly higher but not really.

Also i think my score is better than 98% of all scores, not all 3080 scores. Since im close to bottom 10% of 3080 scores.

 

[MOgens1337]

A locked slider is a good indicator of a locked down vBIOS. That might be something you can deal with if you want to take that risk. Other MSI vBIOS are likely compatible, but then you would have to be careful with temperatures.

And if the GPUs were binned, this might be a poor performer and thus relegated to an OEM card. Binning costs money though, so cheap cards are sometimes just assembled from a tray of GPUs from Nvidia without checking for good ones (though at this point I think Nvidia does their own binning to keep founder's edition silicon at the top)

Well its not locked, i just cant move it above 100%, writing that i realize that might mean locked

Also yeah i realize that there is a good chance, that the silicon isnt the best in an OEM, but i came from an overperforming 960, and an underperforming 3080 is just sort of bittersweet tbh - nerdtears.

Does this pc play the games you want with acceptable performance?

Best, I think to ignore such comparisons.
Those with good results will post.
Those with dogs remain silent.

Considering how hard it is to find any 3080, you probably did well.
The cpu is no slouch either.

It does, and yeah i try to not care, but bottom 10% just feels bad man :/
Oh i thought it automatically published the results! That makes me feel better.

And yeah, the cpu seems to be performing well, which is good, because i want to stream.

 

[sizzling]

This type of benchmarking tool is used by those tuning their gpu’s. Could be overclocking, could be increasing power or thermal limits or even undervolting can help performance. Those doing this will run the tests many times with likely higher than stock results for the majority of runs. You are comparing to skewed numbers.

 

[Eximo]

Lenovo system memory is not likely to be amazing either, could just have high latency. CPU at stock settings, and being unable to change boost times and power limits would certainly put you behind on the CPU side of things. The GPU is just not a top performer, likely down to cooler and the vBIOS.

Nothing to worry about really. Only if you want to start tweaking and overclocking and getting the most performance possible.

Messing with the vBIOS you risk bricking the card, not a wise move in this market. But that would probably get you a power slider that goes over 100%.

 

[MOgens1337]

Lenovo system memory is not likely to be amazing either, could just have high latency. CPU at stock settings, and being unable to change boost times and power limits would certainly put you behind on the CPU side of things. The GPU is just not a top performer, likely down to cooler and the vBIOS.

Nothing to worry about really. Only if you want to start tweaking and overclocking and getting the most performance possible.

Messing with the vBIOS you risk bricking the card, not a wise move in this market. But that would probably get you a power slider that goes over 100%.

When you say system memory, do you mean ram that came with the pc?

 

[Eximo]

Yes, generally OEMs don't buy the best memory.

I believe it was DDR4 3200 2x8GB. Typical 'good' latency would start at 16 and go down from there. OEMs will tend to buy things like CL22, CL20, whatever they could get in bulk.

You can download CPU-Z and see what your memory timings are like.

 

[helper800]

You could also check the MSI afterburner setting to see if there is an option that you have to turn on to allow above 100% power limits. I am not sure, its been a while since I used MSI afterburner. Also, humor me and do a Userbenchmark and like the results here for us to see. Most people don't OC their cards for that one and you can get a breakdown of how fast your particular model of 3080 is compared to others of the same model, not just all other 3080s.

 

[Biaz]

When you say system memory, do you mean ram that came with the pc?

3200 CL 22 on Jedec Profile for Lenovo's Ram. I replaced those with 2x16 Fury Beast DDR4 3200 with PnP. Frequency goes down to 2400 with CL16. Performances are boost for the lower latency (Intel CPUs prefer low timing insted higher frequency). With the stock ram, this Pc is very frustrating. With this brand new, i got a very good increase. I couldn't imagine it could make such a big difference.