1060 ti vs 1650 super: MSI GeForce GTX 1650 Super Gaming vs Nvidia GeForce GTX 1060: What is the difference?

3 Votes

9000

Reliability

7.0 /10

3 Votes

10.0 /10

2 VOTES

1. TECH FRIENTS GP

1506MHZ

004 1530MHz

The graphics processing unit (GPU) has a higher clock speed.

2.turbo GPU

1708MHz

1725MHz

When the GPU is running below its limits, it can jump to a higher clock speed to increase performance.

3.pixel rate

72.3 GPixel/s

55.2 GPixel/s

The number of pixels that can be displayed on the screen every second.

4.flops

3.85 TFLOPS

4.42 TFLOPS

FLOPS is a measurement of GPU processing power.

5.texture size

120.5 GTexels/s

138 GTexels/s

Number of textured pixels that can be displayed on the screen every second.

6.GPU memory speed

2002MHz

1500MHz

Memory speed is one aspect that determines memory bandwidth.

7.shading patterns

Shading units (or stream processors) are small processors in a graphics card that are responsible for processing various aspects of an image.

8.textured units (TMUs)

TMUs accept textured units and bind them to the geometric layout of the 3D scene. More TMUs generally means texture information is processed faster.

9 ROPs imaging units

ROPs are responsible for some of the final steps of the rendering process, such as writing the final pixel data to memory and performing other tasks such as anti-aliasing to improve the appearance of graphics.

Memory

1.memory effective speed

8008MHz

12000MHz

The effective memory clock frequency is calculated from the memory size and data transfer rate. A higher clock speed can give better performance in games and other applications.

2.max memory bandwidth

192.2GB/s

192GB/s

This is the maximum rate at which data can be read from or stored in memory.

3.VRAM

VRAM (video RAM) is the dedicated memory of the graphics card. More VRAM usually allows you to run games at higher settings, especially for things like texture resolution.

4.memory bus width

192bit

128bit

Wider memory bus means it can carry more data per cycle. This is an important factor in memory performance, and therefore the overall performance of the graphics card.

5. GDDR memory versions

Later versions of GDDR memory offer improvements such as higher data transfer rates, which improve performance.

6. Supports memory debug code

✖Nvidia GeForce GTX 1060

✖Nvidia GeForce GTX 1650 Super

Memory debug code can detect and fix data corruption. It is used when necessary to avoid distortion, such as in scientific computing or when starting a server.

Functions

1.DirectX version

DirectX is used in games with a new version that supports better graphics.

2nd version of OpenGL

The newer version of OpenGL, the better graphics quality in games.

OpenCL version 3.

Some applications use OpenCL to use the power of the graphics processing unit (GPU) for non-graphical computing. Newer versions are more functional and better quality.

4. Supports multi-monitor technology

✔Nvidia GeForce GTX 1060

✔Nvidia GeForce GTX 1650 Super

The video card has the ability to connect multiple screens. This allows you to set up multiple monitors at the same time to create a more immersive gaming experience, such as a wider field of view.

5.GPU Temperature at Boot

Lower boot temperature means that the card generates less heat and the cooling system works better.

6.supports ray tracing

✔Nvidia GeForce GTX 1060

✖Nvidia GeForce GTX 1650 Super

Ray tracing is an advanced light rendering technique that provides more realistic lighting, shadows and reflections in games.

7.Supports 3D

✔Nvidia GeForce GTX 1060

✔Nvidia GeForce GTX 1650 Super

Allows you to view in 3D (if you have a 3D screen and glasses).

8. supports DLSS

✖Nvidia GeForce GTX 1060

✖Nvidia GeForce GTX 1650 Super

DLSS (Deep Learning Super Sampling) is an AI based scaling technology. This allows the graphics card to render games at lower resolutions and upscale them to higher resolutions with near-native visual quality and improved performance. DLSS is only available in some games.

9. PassMark result (G3D)

Unknown. Help us offer a price. (Nvidia GeForce GTX 1650 Super)

This test measures the graphics performance of a graphics card. Source: Pass Mark.

Ports

1.has HDMI output

✔Nvidia GeForce GTX 1060

✔Nvidia GeForce GTX 1650 Super

Devices with HDMI or mini HDMI ports can stream HD video and audio to the connected display.

2.HDMI connectors

More HDMI connectors allow you to connect multiple devices at the same time, such as game consoles and TVs.

HDMI 3.Version

HDMI 2.0

HDMI 2.0

New HDMI versions support higher bandwidth for higher resolutions and frame rates.

4. DisplayPort outputs

Allows connection to a display using DisplayPort.

5.DVI outputs

Allows connection to a display using DVI.

Mini DisplayPort 6.outs

Allows connection to a display using Mini DisplayPort.

Price match

Cancel

Which graphic cards are better?

GTX 1650 video card test and comparison with GTX 1650 Super and GTX 1060

This article was written by a website visitor and has been rewarded.

Those who read my posts from time to time probably still remember that not so long ago I already had a test of the GTX 1650 Super video card by Palit StormX. Today, I got my hands on a version of a video card without Super registration, but performed by the same Palit StormX.

These video cards actually differ significantly. In fact, the difference in the number of execution units between the standard version of the GTX 1650 and the GTX 1650 Super is even greater than the difference between the GTX 1650 Super and the GTX 1660 already located in the line above. It is worth noting, however, that the GTX 1660 already has 6GB of video memory on board, while both versions of the GTX 1650 are content with only 4BG of video memory, but of different standards, GDDR5 in the case of the standard GTX 1650 and GDDR6 in the case of the GTX 1650 Super version, which, as a result, due to the higher clock frequency, gives the latter a noticeably higher bandwidth value with the same the same bus width, bringing it closer in this value, again, to the older one in the GTX 1660 line.

recommendations

But the easiest way, oddly enough, is to draw parallels with the previous generation of video cards. If the GTX 1650 Super is a direct analogue of the GTX 1060 6GB in terms of the number of execution units, then the standard version of the GTX 1650 is closer to the GTX 1050 Ti, although it has more processors at its disposal. For greater clarity, you should look at the comparative table of the basic characteristics of these video cards.

I got the video card in the OEM version, that is, without a box, in a dense black pimply bag with brief documentation inside. To be honest, the last time I bought a video card in this design was 15 years ago. Since then, as a rule, even the most budget video cards have been in boxes.

From images on the internet, it might seem that the GTX 1650 Palit StormX and the GTX 1650 Super Palit StormX are pretty much the same in terms of looks. In reality, these video cards are quite different in their dimensions, offhand one and a half times. The GTX 1650 Palit StormX is very compact in real life, slightly larger than a 2.5” hard drive. The plastic shroud also looks a bit different. The video card occupies one slot, or rather, one slot in the case is occupied by a bar, inside the case the video card itself occupies about one and a half slots due to bulging cooling.

GTX 1650 Palit StormX

GTX 1650 Super Palit StormX

The radiator contacts only with the video card chip itself, I did not find any additional cooling of the power subsystem or memory chips. Although the memory is hidden from view, it is located just under the radiator and I could not understand if there was any contact there. But I did not find any hints of this or the presence of thermal pads. Which is logical, in the older version of the GTX 1650 Super Palit StormX, the memory is also not cooled by anything else. In both cases, all elements on the board, including the memory subsystem and memory chips, are simply blown through by the air flow created by the fan.

Despite the modest cooling system, the video card does not get very hot, on average we have a temperature in the range of 60-63 degrees, while the fan speed increases from a minimum of 30% to about 40-42%. In the older GTX 1650 Super Palit StormX, despite the larger system, the temperatures and fan speeds were higher, as was the noise emitted. Unfortunately, as in the older version, here the fans do not turn off when there is no load and constantly rotate at a minimum value of 30%, which, unfortunately, is audible against the general ambient noise background. By itself, the noise level of the fan creates a low, but characteristic, standing out against the background. The number of revolutions, unfortunately, is not displayed for this video card, as for the older version. Therefore, one has to be content with only percentage values.

I will test the video card in the same games, on the same settings and under the same conditions as last time. This will give us the opportunity to compare the results obtained with the results already available and understand what place this video card will take in comparison with the GTX 1060 3GB and GTX 1650 Super. Just in case, I’ll remind you and point out separately on the graphs that the GTX 1060 3GB was overclocked (the operating frequencies of the video card in the tests were 2139 MHz for the core and 9300 MHz for the memory). GTX 1650 Super worked normally. GTX 1650 will also work normally.

All test segments were taken from this recording. The settings can be found here. All measurements were made using the FRAPS program.

Test configuration looks like this:

• Ryzen 3 3100 @3800 MHz+ be quiet! Pure Rock Slim
• ASUS TUF B350M-Plus Gaming (Bios ver. 5602, AGESA 1.0.0.6)
• 32GB (4×8) @2933CL16 Kingston HyperX (HX424C15FB2K2/16 + HX430C15FB3K2/16)
• GTX 1650 (WHQL 460.89)
• SSD WDS500G2B0A 500GB
• NZXT S340
• Cooler Master V1000

Ultra preset, resolution 1920×1080. The GTX 1650 is in third place, while the overclocked GTX 1060 3GB is in second. All three video cards lined up in a fairly even line. The GTX 1650 Super outperforms the GTX 1060 3GB by 16%, which in turn outperforms the GTX 1650 by 19%. All three video cards provide a comfortable level of frame rate for the game.

Manual graphics settings, 2x multisampling, resolution 1920×1080. The GTX 1650 is again in last place, but here the GTX 1060 3GB already had some problems with a lack of video memory, which affects the minimum frame rate, the game on this video card periodically stutters. The GTX 1650, despite the lower average, has no problems with the minimum value. The GTX 1650 Super outperforms the GTX 1060 3GB by 17% in the average and 57% in the minimum frame rate. That, in turn, is ahead of the GTX 1650 by 12% in terms of the average frame rate, but inferior by 21% in the minimum.

Max graphics preset, 4x multisampling, 1920×1080 resolution. One of the few examples in which the GTX 1060 3GB, due to overclocking, still managed to outperform the GTX 1650 Super at least a little. Although the advance is extremely conditional. The GTX 1650 is about 20% behind them. Playable framerate is provided by all three video cards.

Maximum graphics settings preset, resolution 1920×1080. In Far Cry 5, the balance of power is again the same, but in this game, the gap between the GTX 1650 and the GTX 1060 3GB is no longer so noticeable. All video cards provide an acceptable frame rate, although on the GTX 1650 this value is already on the verge and can sink below 60 frames.

Maximum graphics settings, no anti-aliasing, advanced settings disabled, resolution 1920×1080. GTX 1650 Super and GTX 1060 3GB (in overclocking, let me remind you) go head to head, although the GTX 1650 Super still breaks out a little, especially in terms of the maximum frame rate. The younger GTX 1650 is not so inferior to them. All three video cards provide a comfortable level of frame rate.

Medium graphics preset, tessellation off, HairWorks off, resolution 1920×1080. Here we again see a balance of power similar to Battlefield 1 with a relatively even gap between all three video cards. But the GTX 1650 Super breaks away from the other two cards more noticeably. It will be possible to play comfortably on all three video cards.

Manual settings from last post, resolution 1920×1080. In Resident Evil 2 Remake, the GTX 1650 managed to nearly match the overclocked GTX 1060 3GB, with the GTX 1650 Super beating it by 19%. It will be possible to play comfortably on all three video cards.

Maximum graphics settings, FXAA, resolution 1920×1080. The second game in the test, in which the GTX 1060 3GB managed to take the lead, albeit formally and only due to overclocking. In fact, the GTX 1060 3GB and GTX 1650 Super caught up, which can not be said about the younger GTX 1650, which is very significantly inferior to other video cards. You can play on all three video cards, but on the GTX 1650, the frame rate can drop even below 50 frames.

Maximum graphics settings, TAA anti-aliasing, resolution 1920×1080. Another game in which the GTX 1060 3GB and GTX 1650 Super are extremely close in performance, while the younger GTX 1650 is largely inferior to them. A comfortable game is possible on all three video cards.

First, looking at the results, keep in mind that the GTX 1060 3GB was overclocked, while the GTX 1650 was running normally, so the actual distribution between these two cards under equal conditions will be different. Secondly, you need to understand that the games used in the tests are mostly old, released even before the advent of the Turing architecture, in the era of the total use of the DirectX 11 API. All these games have been optimized to take advantage of legacy architectures and do not give us a real picture of the performance of newer graphics cards. In the last entry of the GTX 1650 Super test, I touched on this topic in more detail, but I will briefly remind you that new architectures (Turing, and now Ampere) have brought with them certain architectural changes and improvements, which, if used, can give a noticeable advantage over to now obsolete architectures, including Pascal. These improvements include the possibility of parallel execution of INT and FP calculations, while previously these operations were performed in turn, plus Turing architecture video cards were taught to execute FP16 calculations at twice the speed, and INT calculations became much faster. Do not forget about the support of such technologies as, for example, VRS and Mesh Shading.

Last time I used the Neon Noir Ray Tracing benchmark for clarity, which uses its own method of implementing ray tracing on video cards without using hardware blocks, and this example very well demonstrated the advantage of the fresh architecture. Moreover, if last time we could not yet observe the real application of this solution, now we already have a living example in the form of the game Crysis Remastered, which uses this method already directly in the game. So the use of this benchmark is quite viable and realistic. Therefore, I propose to turn to him again.

As expected, the GTX 1650 performs better in this benchmark than in older games. In stock, it performs on par with the overclocked GTX 1060 3GB, and in a small overclock it outperforms it by 9%. Speaking of overclocking, in this case it would be strange to expect something special. On the core, the video card reaches an average of ~2000 MHz, like the older GTX 1650 Super from the previous test. As for the memory, artifacts appear already at a value of about +600 MHz. So I set 500 MHz as a stable operating value, with a small margin of stability. As a result, the effective memory frequency increased from 8000 MHz to 9000 MHz. For comparison, on the older GTX 1650 Super, which has GGDR6 memory installed on board (plus there is an additional 6-pin power supply), it was possible to raise the effective operating frequency from the standard 12000 MHz to 14600 MHz.

In general, the video card leaves a very pleasant impression, as for a budget compact version that occupies minimal space in the case and does not require additional power. By the way, there is a version of the video card with GDDR6 memory, operating at an effective frequency of 12000 MHz, like the GTX 1650 Super. Another question is whether there is any sense in such memory on a video card of this level. I don’t think so. Whether it is worth buying such a video card is an individual question. If you are not a particularly demanding gamer and you have a FHD monitor with a refresh rate of 60-75 Hz, then it is fine. Moreover, chasing some tops in this case does not make much sense, a more or less unpretentious non-reference option will provide quite effective and quiet cooling.

This material was written by a site visitor and has been rewarded.

GTX 1650 Super vs. GTX 1060

This content has been written by a website visitor and has been rewarded.

Super

GTX 1650 Super is an extremely interesting video card that significantly outperforms the simple GTX 1650 in its characteristics and is only slightly inferior to the older GTX 1660. Perhaps, in terms of price to performance ratio, this is one of the best options on the market at the moment . However, the performance and features will be lower.

We’re talking about Palit’s StormX graphics card, which is essentially the most affordable and affordable GTX 1650 Super graphics card solution on the market at the moment. The video card is delivered in a box of modest size, which is customary in design for Palit. There is nothing to talk about some kind of equipment, except for the video card and booklets, there is nothing and is hardly necessary.

recommendations

The video card has extremely modest dimensions and a simple cooling system, which is an aluminum radiator bar with a single 100 mm fan. The heat sink contacts only the chip, the memory chips and the VRM zone are cooled according to the residual principle only by the air flow of the main fan. The radiator fins are arranged longitudinally, so that part of the heated air leaves the body.

Not all drivers are equally useful

Unfortunately, the first thing I encountered after installing the driver was a problem, and quite a complex one at that. After installing the driver, the familiar Nvidia control panel was missing, and every time Windows 10 started reporting that this Nvidia control panel was missing, offering to click on the link to install it, but this did not lead to any result, no installation took place, nor when clicking on the link, or when manually trying to install this control panel through the Microsoft Store. Moreover, all my data in the browser began to be reset after each reboot, I had to manually log into all accounts again, and my email was full of messages that I was logged into my account from a new device.

The problem turned out to be quite simple — the official site www.nvidia.ru downloads the wrong version of the driver, and there is no way to select the desired version on the Russian-language site. There are two versions: standard and DCH. We need a standard driver, but the official site downloads the DCH version which is causing all these problems. There are at least two solutions: use the foreign version of the site www.nvidia.com, where you can select the version of the driver to download, or change the link address manually. When trying to access a foreign site, I encountered the fact that it automatically redirects me to a Russian-language one. Therefore, I used the second method. When we select the driver version and get to the page where we are offered to click on the button and download the driver, we copy the link address, paste it into the browser line, remove the extra letters like “-dch” and only then press Enter. That’s it, we download the standard version of the driver. Moreover, apparently, this problem has been at least six months old, but it has not yet been solved. Below I give an example of coverage and solution of this problem with reference to a source dated January 2020, that is, six months have passed.

Image of taken from here

Game tests

The clock speed of the video card, even despite its budget and an extremely modest cooling system under load, is kept within 1900 MHz, which is quite good.

We will compare this video card with my already former GTX 1060 3GB MSI Gaming X in overclocking to 2139 MHz for the core and 9300 MHz for memory. Yes, we will be comparing the stock budget GTX 1650 Super with the flagship GTX 1060 3GB overclocked. Why is that? We will assume that this is a compensation for a small difference in computing units and older video memory. Still, this is a budget video card four years ago, let’s give it a small head start as a tribute. A visual comparison of the characteristics of both video cards can be seen below.

All test segments were taken from this recording. The settings can be found here. All measurements were made using the FRAPS program.

Test configuration:

• Ryzen 5 1500X @3700 MHz + be quiet! pure rock slim
• ASUS TUF B350M-Plus Gaming
• 2x8GB Kingston HyperX @2800 MHz CL14-16-14-30-48 1T (HX424C15FB2K2/16)
• GTX 1650 Super 4GB Palit StormX @1900/12000
• GTX 1060 3GB MSI Gaming X @2139/9300
• SSD WDS500G2B0A 500GB
• HDD WD10EZEX 1TB

Ultra preset, resolution 1920×1080. The GTX 1650 Super 4GB outperforms the GTX 1060 3GB by 16% despite the overclocking of the latter, both video cards provide a comfortable frame rate.

Manual graphics settings, 2x multisampling, 1920×1080 resolution. Here, the GTX 1060 3GB had some low VRAM issues. The GTX 1650 Super 4GB leads it by 17% on average and 57% on the minimum frame rate.

Maximum graphics preset, 4x multisampling, 1920×1080 resolution. One of the few examples in which the GTX 1060 3GB, due to overclocking, still managed to outperform the GTX 1650 Super 4GB at least a little. Although the advance is extremely conditional.

Maximum graphics settings preset, resolution 1920×1080. The GTX 1650 Super 4GB again takes over the GTX 1060 3GB by 15%.

Maximum graphics settings, no anti-aliasing, advanced settings disabled, resolution 1920×1080. Both graphics cards go head to head, although the GTX 1650 Super 4GB still breaks out a bit, especially in terms of maximum frame rate. It is worth noting that in this game the load of the video card was not always kept at the maximum value, so the real gap and the potential of the video card may be higher.

Medium graphics preset, tessellation off, HairWorks off, resolution 1920×1080. Here we again observe the already quite familiar GTX 1650 Super 4GB lead by 17% in terms of the average frame rate and by 25% in the minimum one.

Manual settings from last post, resolution 1920×1080. The situation is repeating itself. The GTX 1650 Super 4GB outperforms the GTX 1060 3GB by 19% in terms of average frame rate.

Maximum graphics settings, FXAA, resolution 1920×1080. The second example for today, in which the GTX 1060 3GB, thanks to overclocking, still managed to pull ahead a little. Although, as in the first case, breaking out is very formal.

Maximum graphics settings, TAA, resolution 1920×1080. In this case, video cards are extremely close in their performance, although with some variation in the minimum and maximum values. And in both directions, which is interesting. The novelty has an advantage in terms of the maximum value, but for some reason it is inferior to the older video card in terms of the minimum value. Randomness is excluded, there were three runs in both cases, the pattern is preserved.

Ray Tracing

We all know and remember that the GTX series video cards are devoid of any hardware computing units, despite the fact that they are based on the Turing architecture. And many, according to my observations, mistakenly believe that only these blocks are responsible for the high speed of the Ray Tracing technology. But it’s not so simple. Even the official slides from Nvidia clearly show that the calculations associated with ray tracing fall not only, and maybe even not so much on the hardware units, but on the rest of the computing units of the video card. What this means in practice — I suggest you take a look at the Neon Noir Ray Tracing Benchmark technical demo from Crytek, comparing two of our video cards that are extremely similar in terms of their computing abilities, both of which lack any specialized hardware blocks, but are based on two different architectures: Pascal and Turing. As we have already seen from the results of gaming tests without ray tracing, these video cards are quite close to each other. What will happen in the case of using ray tracing?

The GTX 1650 Super based on the Turing architecture pulled ahead by as much as 42% even despite the lower clock speeds, that is, almost one and a half times.

As far as I understand, but I could be wrong, one of the features in the Turing architecture was the possibility of simultaneous parallel calculations of INT32 and FP32, while in the case of Pascal these operations were performed in turn, in addition, Turing was taught to execute FP16 at twice the speed, which in total and gave such an increase in ray tracing calculations even without any additional hardware blocks. The RT hardware cores themselves took on the task of calculating BVH, and the tensor cores clean the picture from noise. As a result, in the aggregate of all this work, we get a tangible increase compared to the old architectures.

I expect that in future games that use open standards for ray tracing and similar techniques, like the one we were shown in the Unreal Engine 5 tech demo, Turing GTX series graphics cards will show their architectural strengths and will continue to confidently stay afloat, while video cards of past generations will remain out of work, approximately the same as in the example above.

For this reason, I think that drawing parallels between video cards of past generations and video cards of the GTX series based on the Turing architecture based on the results in old games just because they do not have hardware cores is not entirely correct. Yes, in old games, these video cards can generally show close results, but they are old games because they do not use the architectural advantages of the new series of video cards. The example with Neon Noir Ray Tracing Benchmark clearly shows that these video cards, if more modern technologies are involved, can differ greatly from each other in terms of performance in favor of more modern ones. Turing GTX series graphics cards have potential even without hardware cores, and it will show itself. In addition to the differences and advantages mentioned above, do not forget about the support for new technologies, such as VRS and Mesh Shading, which will also bear fruit in future projects. In the same Unreal Engine 5 tech demo, they demonstrated the Nanite technique, which is clearly based on the same principle of operation as Mesh Shading, and, most likely, new video cards will show their advantage in this direction as well. And of the existing games, Wolfenstein II can be noted, taking advantage of VRS technology.

Power consumption, noise level, temperature, overclocking

For some reason, no program sees the speed of the video card fan, only percentage values ​​are available, so you will have to be content with the percentage level. This version of the video card does not support the fan stop mode at low temperatures, in idle mode without load, the speed is 35%, and the overall noise level does not stand out from other PC components. Under load, the speed increases to 57%, the temperature reaches 62 degrees, and the noise level increases and becomes distinct, it stands out from the rest of the PC components, however, it cannot be called especially loud, it is moderate, somewhat reminiscent of the level of reference Nvidia video cards Pascal generations.

The power consumption was measured with a household wattmeter, so in the graph this value reflects the power consumption of the entire system from the outlet. For the objectivity of the result, I took different measurements under different loads (processor load, video card load, processor load and video card load) and in idle time. The declared power consumption level for this video card is 100 watts. Monitoring programs report on average values ​​close to this figure when the video card is loaded. When undervolting, this value drops to about 70 watts. The power consumption of the processor is about 60 watts.

Despite the presence of an additional 6-pin power supply, the Power Limit level cannot be increased on this version of the video card. Therefore, in this case, as such, overclocking is possible only within the limits of the standard values ​​​​of voltage and power consumption.

After some manipulations, we managed to get about 2025-2040 MHz instead of the regular ~1900 MHz for the chip and 14600 MHz instead of the regular 12000 MHz for the memory. In Neon Noir Ray Tracing Benchmark, the performance gain was about 9%.