Qx9650 vs i7: Core 2 Extreme QX9650 vs Core i7-12800HE

Core2 Quad Q9650 vs i7-3770 — Fortnite Battle Royale with GT 1030 Benchmarks 1080p, 1440p, Ultrawide, 4K Comparison

GT 1030 with

Intel Core2 Quad Q9650 @ 3. 00GHz

Fortnite Battle Royale

GT 1030 with

Intel Core i7-3770 @ 3.40GHz

Core2 Quad Q9650
i7-3770

Multi-Thread Performance

4177 Pts

9293 Pts

Single-Thread Performance

1262 Pts

2068 Pts

Fortnite Battle Royale

Core2 Quad Q9650 vs i7-3770 in Fortnite Battle Royale using GT 1030 — CPU Performance comparison at Ultra, High, Medium, and Low Quality Settings with 1080p, 1440p, Ultrawide, 4K resolutions

Core2 Quad Q9650
i7-3770

Ultra Quality

High Quality

Medium Quality

Low Quality

Core2 Quad Q9650

i7-3770

Compare Core2 Quad Q9650 vs i7-3770 specifications

Core2 Quad Q9650 vs i7-3770 Architecture

	Core2 Quad Q9650	i7-3770

Core2 Quad Q9650 vs i7-3770 Cache

	Core2 Quad Q9650	i7-3770

Core2 Quad Q9650 vs i7-3770 Cores

	Core2 Quad Q9650	i7-3770

Core2 Quad Q9650 vs i7-3770 Features

	Core2 Quad Q9650	i7-3770

Core2 Quad Q9650 vs i7-3770 Performance

	Core2 Quad Q9650	i7-3770

Core2 Quad Q9650 vs i7-3770 Physical

	Core2 Quad Q9650	i7-3770

Share Your Comments 36

Compare Core2 Quad Q9650 vs i7-3770 in more games

Elden Ring

2022

God of War

2022

Overwatch 2

2022

Forza Horizon 5

2021

Halo Infinite

2021

Battlefield 2042

2021

Assassin’s Creed Valhalla

2020

Microsoft Flight Simulator

2020

Valorant

2020

Call of Duty: Black Ops Cold War

2020

Death Stranding

2020

Marvel’s Avengers

2020

Godfall

2020

Cyberpunk 2077

2020

Apex Legends

2019

Anthem

2019

Far Cry New Dawn

2019

Resident Evil 2

2019

Metro Exodus

2019

World War Z

2019

Gears of War 5

2019

F1 2019

2019

GreedFall

2019

Borderlands 3

2019

Call of Duty Modern Warfare

2019

Red Dead Redemption 2

2019

Need For Speed: Heat

2019

Assassin’s Creed Odyssey

2018

Battlefield V

2018

Call of Duty: Black Ops 4

2018

Final Fantasy XV

2018

Shadow of the Tomb Raider

2018

Forza Horizon 4

2018

Fallout 76

2018

Hitman 2

2018

Just Cause 4

2018

Monster Hunter: World

2018

Strange Brigade

2018

Assassin’s Creed Origins

2017

Dawn of War III

2017

Ghost Recon Wildlands

2017

Destiny 2

2017

PlayerUnknown’s Battlegrounds

2017

Fortnite Battle Royale

2017

Need For Speed: Payback

2017

For Honor

2017

Project CARS 2

2017

Forza Motorsport 7

2017

Ashes of the Singularity

2016

Ashes of the Singularity: Escalation

2016

Battlefield 1

2016

Deus Ex: Mankind Divided

2016

Doom

2016

F1 2016

2016

Hitman

2016

Rise of the Tomb Raider

2016

The Division

2016

Total War: Warhammer

2016

Overwatch

2016

Dishonored 2

2016

DiRT Rally

2015

Grand Theft Auto V

2015

The Witcher 3

2015

Rocket League

2015

Need For Speed

2015

Project CARS

2015

Rainbow Six Siege

2015

Battlefield 4

2013

Crysis 3

2013

Counter-Strike: Global Offensive

2012

League of Legends

2009

Minecraft

2009

Test Intel Core i7-950 — CompLine

The update of the Core i7 line was not limited to the release of the 975 Extreme Edition. Also on sale is a more affordable processor model, the performance of which is not damn inferior to the flagship model — the Core i7-950.

Intel is now systematically updating its line of Core i7 processors. This happens, most likely, to widen the gap in performance between it and the upcoming i5, in which processors of the middle price category start. 9The 75 processor replaces the 965, and the 950 replaces the 940. The entry-level model, the i7-920, is still on sale. In addition, it was transferred to a fresh stepping (thus, all the processors of the line «moved» to it), and they plan to remove it from production only together with the announcement of the i5 line.

Earlier we published a review of the Core i7-975 processor, where we described the main features that the new stepping brought us. Just in case, we will point them out again: reduced power consumption and improved overclocking potential. Note that no work has been done on processor performance.

We’ll now look at a more affordable processor, the Core i7-950. It is also based on the D0 stepping and its main difference from the i7-975 is the locked multiplier. We do not think that for many this will be a huge problem.

As in the case of the i7-975, the formal difference with the previous model is minimal — the multiplier was increased by 1, which provided a frequency increase of 133 MHz. The cost of the upgrade remained the same — $562 in bulk. This is complete, of course, but a lot has already been said about the price of the Core i7 platform.

S-Spec of the new processor — SLBEN, CPUID — 0x000106A5.

The CPU-Z screenshot clearly shows all the changes that have taken place.

Checking the overclocking potential showed that the processor «took» the bar at 4 GHz without any problems. This time, we had an MSI Eclipse Plus motherboard at our disposal, which provided us with more stability in overclocking than the model from Intel. As a result, we raised the frequency to 4117 MHz (23×179), for which we also had to increase the effort (under load, a value of 1.384 V was noted). It seems that this is not the limit for this combination of processor and board, and yet we still lack the cooling capacity.

We did not come to this conclusion immediately. Temperature monitoring was carried out by us using the Speedfan 4.37 program. As you know, Core i7 processors have a temperature sensor (digital thermal sensor) inside each core, the one that captures the difference between the current temperature and the maximum Tjmax. Next, most programs convert this data into more user-friendly values. At the same time, the program itself can assign different values ​​to Tjmax, which accordingly affects the results.

The Speedfan experienced 84°C during overclocking, suggesting some provision. Therefore, we thought we had reached the overclocking limit for the processor itself. For all that, due to the fact that the recorded temperature on the graph directly “rested” on this significance, we had some suspicions.

In this regard, the RealTemp 3.00 program was used, which, in addition to the usual temperature data (normally based on Tjmax=100 ?C), directly displays the data read by the sensor. It was she who showed that the temperature of the processor was very close to Tjmax.

For this reason, starting from this article, I will use the program RealTemp 3.00 to monitor the temperature. The comparisons made before remain valid, but it is not worth comparing the Speedfan and RealTemp readings.

In addition, I would like to note that the copy of the processor that came to us had a somewhat uneven surface of the heat spreader cover. We used a high-quality thermal interface (Arctic Silver Cermamique), and yet the cooler’s heatsink warmed up relatively weakly, which indicates poor heat transfer.

In any case, we had no opportunities for further overclocking of the processor, and we achieved decent results. Of course, it is unrealistic to say unequivocally about the growth of overclocking potential for one copy, but this result is above average for the C0 stepping. In addition, world statistics show that processors based on D0 really overclock better.

We conducted a full-scale test, in which we compared the performance of the 950 model (in normal mode and with overclocking) with one of the fastest processors for Socket 775 — Core 2 Extreme QX9650, and in addition with the Core i7-975, the one that we have already considered before. Comparison with processors of the previous stepping was not carried out, since even in our review of the Core i7-975 it was found that there was no difference in performance. The results of comparison with the Core i7-975 are also quite predictable, but the correlation with the QX9650 may be interesting.

The following test bench configuration was used:

Cooler: Thermaltake ISGC-300 + Fan-12

Synthetics started: Sandra 2009, 3D Mark’06, PC Mark’05, Super Pi.

The stock frequencies of the QX9650 and i7-950 are quite close. We didn’t specifically align them; there is already plenty written about the performance of the Nehalem architecture in comparison with the Core. For all that, in all tests, in addition to PC Mark, it is clear that the i7-950 is faster by more than 20%. In the PC Mark test, the gap is only around 5%.

Now let’s look at tests close to reality. Let’s start with games. Testing was carried out using Smart FPS. We tested performance in Crysis, Crysis: Warhead, Unreal Tournament 3 and Prey. A resolution of 1280*1024 was used, thanks to which the difference between the processors should have become more noticeable.

Nevertheless, Crysis ran into the performance of the video card, moreover, in such a sparing mode. In addition, a surprise awaits us here — even with overclocking, the i7 cannot overtake the QX9650. We did not experiment with disabling Hyper-threading, although in some games this leads to an increase in performance.

Unreal Tournament 3 and Prey showed a much greater bondage to the processor frequency. In Prey i7-950 in normal mode, it also outperformed Core by more than 10%. However, in general, you can notice a slight increase in performance in games. It is unlikely that many people, having such a computer configuration, will play at 1280 * 1024 and medium graphics settings, and at other resolutions the superiority will be even less significant.

Next is a group of tests comparing processor performance in other real use cases:

These tasks show a much greater (close to linear) servitude to the processor frequency. Exceptional archiving accelerates rather weakly, since the memory subsystem and the hard disk are of great importance for it. The Nehalem architecture performs very well in these tasks. The advantage in Paint. NET is particularly impressive. Thus, for these purposes, the purchase of a line 9 processor0004 Core i7 can be considered justified.

As for the i7-950 specifically, it can be considered a very appropriate announcement. The only real competitor to Nehalem at the moment is they themselves, so it doesn’t make much sense to discuss Intel’s pricing policy. Such an update of the model range can only be welcomed, because for the same money we are now offered greater performance and less heat generation.

Materials used Ferra.ru

Intel Core i5, Core i7 and others: a practical study

A comprehensive report on testing new processors is somewhat reminiscent of the realities of student life: having examined in detail the architectural and technological features of the Core i5 / i7 8xx CPUs in a «theoretical lesson» in the form of yesterday’s article , we are moving from a «lecture» to a «seminar», during which we will find out in practice how Lynnfield performs in comparison with its Bloomfield counterpart and even earlier predecessor — Core 2 Quad.

We help

Performance evaluation of new processors was carried out at their nominal frequencies both with and without Turbo Boost enabled. In order to compare the performance of the updated Nehalem architecture with its original version, we used an Intel Core i7 965 set to 2.93 GHz, the same as the Core i7 870. The Core 2 Quad QX9650 was tested in a similar way as the latest representative of the Core architecture. 2: The frequency was fixed at the standard 3 GHz mark for this CPU. In all cases, the RAM worked in DDR3-1333 mode with the same timings, and for Core i7 965 — in three-channel mode, since it is obvious that those who strive for maximum performance will use all the possibilities of the platform to achieve it.

Test bench configuration

Processors	Intel Core i7 965 Extreme Edition
	Intel Core i7 870
	Intel Core i5 750
	Intel Core 2 Quad QX9650
Coolers	Thermalright Ultra-120 eXtreme
	Zalman CNPS-9700 LED
Motherboards	ASUS P6T7 WS Supercomputer
	ASUS P7P55D Deluxe
	MSI X48C Platinum
RAM	Kingston KHX16000D3ULT1K3/3GX 3x1GB DDR3-2000
Hard disk	WD VelociRaptor WD3000GLFS 300 GB 10000 rpm 16 MB SATA-II AHCI
Monitor	Samsung T260HD 26″ 1920×1200

As test applications, both more or less traditional home PC programs (games, archiver, video encoders) and professional software for 3D design and image processing were used. Some of them (for example, Photoshop, WinRAR and the x264 encoder) allow you to evaluate the efficiency of the memory subsystem and caches, others — pure computing performance and gain from Hyper-Threading in real conditions.

Let’s start with the most low-level tests — throughput and latency of RAM. As you might expect, the triple-channel controller in the Core i7 965 maxes out at almost 16 GB/s of bandwidth — but that’s only 7% more than the Core i7 870. dual-channel mode, these figures confirm our thesis about the redundancy of this solution. Especially since latency in three-channel mode turns out to be worse by about 10%, and many applications, not using the increased bus bandwidth, will suffer from higher RAM latencies. In addition, if we talk about obtaining maximum frequencies or minimum DDR3 timings, then you can achieve better results with a traditional configuration, because the fewer modules and channels, the more efficient overclocking. This is also an important aspect, because such an optimization of the memory subsystem can compensate for the difference in throughput and at the same time further reduce latency. There are no tests with different DDR3 parameters in the diagrams, but indirectly this conclusion is confirmed by the results of the Core i7 870 with Turbo Boost in testing the memory subsystem.

Actually, the increase from Turbo Boost is visible in almost all applications, and in some cases the difference in speed is very noticeable. Of course, such a result was absolutely predictable — after all, in fact, we see the performance indicators of a corny overclocked CPU. But if earlier the overclocking process was somehow shifted to the user (and the vast majority of PC owners simply do not practice overclocking for various reasons, including the understandable desire to “operate the components as the manufacturer intended”), now Intel has taken care of increasing the clock frequency by planning safely possible overclocking under various conditions for the stability of the system. Thus, if it is important to compare products in the most identical configurations to evaluate the effectiveness of the architecture, then for a real consumer, the diagrams of the CPU operation with Turbo Boost enabled are of the greatest interest — the way it will be used in reality.

The performance in gaming applications is interesting: Hyper-Threading technology does not always help in them, and often even reduces performance. The reason, in particular, is that game engines do not constantly load the processor, threads are generated depending on the situation (processing physical effects, calculating artificial intelligence, etc.). This can cause slowdowns when redirecting threads to virtual cores. We can see an example in the results of the Core i5 750: even considering the lower frequency, this CPU is at least as good as older models, and in some cases (in particular, Unreal Tournament 3) even works a little faster.

Noteworthy are the results in Adobe Photoshop. Our test package is a set of filters and photo transformations, some of which are purely computational tasks, and some require a considerable speed of the memory subsystem. In additional materials on the site, you can find a detailed report of this test for each of the filters, and it shows that the Core 2 Quad, which has 12 MB of fast L2 cache versus 1 MB of L2 and 8 MB of slow L3 in Core i7, noticeably in single-threaded processing ahead of innovation. At the same time, Photoshop became a benefit of i7 965 — the gap from the i7 870 is quite large.

At the same time, sometimes the built-in memory controller and Hyper-Threading are expected to bring excellent dividends, as can be seen in video encoding, rendering and archiving, i.e. tasks that parallelize computational threads well. In these tests, the gap between the older processors of the Nehalem architecture (with 8 virtual cores) and the Core 2 is very significant, but the Core i7 750 without Hyper-Threading support shows itself much more modestly, although it improves the results with Turbo Boost enabled.

Summing up the results of testing, we can note the following key points. Firstly, when compared at the same frequency, processors based on the Bloomfield core, despite the three-channel access to DDR3, in most tasks differ minimally in speed from the new products based on LGA1156. Perhaps this is good news for consumers, since the new platform is cheaper. Secondly, the main competitive advantage of Nehalem in relation to its predecessor is support for Turbo Boost and Hyper-Threading, which is clearly visible when comparing the results of QX9650, Core i7 750 and Core i7 870. Unlike traditional overclocking, the user does not need to make any effort to speed up the new CPUs, so the real performance of the Core i7 is with Turbo Boost enabled. And finally, thirdly, it is worth recognizing that, compared with the revolutionary Core 2, the performance gain of the new generation of processors is, let’s say, «evolutionary». At one time, Pentium 4 also showed similar results compared to Pentium III, only after a couple of years becoming much faster. This time around, a similar breakthrough might be expected from Westmere’s upcoming 32nm generation, which is months away from the announcement of its first products. So once again, let’s fix the main conclusion that suggested itself in the theoretical material: Lynnfield are a good replacement for Core 2 in all respects, and in many respects they will even be competitors for older models of the i7 series 9xx. But you should not expect a long life of this generation of processors — even more interesting new items are already being prepared to replace them.

Intel is now systematically updating the line of Core i7 processors. This is done, most likely, to widen the performance gap between it and the upcoming i5, which will launch mid-range processors. 9The 75 processor replaces the 965, and the 950 replaces the 940. The junior model, the i7-920, is still on sale. It was also transferred to a new stepping (thus, all the processors of the line «moved» to it), and they plan to remove it from production only together with the announcement of the i5 line.

Earlier we published a review of the Core i7-975 processor, where we described the main features brought to us by the new stepping. Just in case, we will point them out again: reduced power consumption and improved overclocking potential. Note that no work has been done on processor performance.

Today we’re looking at a more affordable processor, the Core i7-950. It is also based on the D0 stepping and its main difference from the i7-975 is the locked multiplier. We do not think that for many this will be a big problem.

As in the case of the i7-975, the formal difference with the previous model is minimal — the multiplier was increased by 1, which gave a frequency increase of 133 MHz. The upgrade price remained the same — $562 in bulk lots. This is a lot, of course, but a lot has already been said about the price of the Core i7 platform.

S-Spec of the new processor — SLBEN, CPUID — 0x000106A5.

The CPU-Z screenshot clearly shows all the changes that have taken place.

Checking the overclocking potential showed that the processor «took» the bar at 4 GHz without any problems. This time we had an MSI Eclipse Plus motherboard at our disposal, which gave us more stability in overclocking than the model from Intel. As a result, we raised the frequency to 4117 MHz (23×179), for which we also had to increase the voltage (1.384 V was fixed under load). It seems that this is not the limit for this combination of processor and board, however, we do not have enough cooling capacity.

We did not come to this conclusion immediately. Temperature monitoring was carried out by us using the Speedfan 4.37 program. As you know, Core i7 processors contain a temperature sensor (digital thermal sensor) inside each core, which records the difference between the current temperature and the maximum Tjmax. Most programs then convert this data into more user-friendly values. In this case, the program itself can assign different values ​​to Tjmax, which accordingly affects the results.

The Speedfan experienced 84 ºC during overclocking, suggesting some headroom. Therefore, we thought we had reached the overclocking limit for the processor itself. However, due to the fact that the recorded temperature on the chart directly «rested» against this value, we had some suspicions.

In this regard, the RealTemp 3.00 program was used, which, in addition to the usual temperature data (standardly based on Tjmax=100 ºC), directly displays the data read by the sensor. It was she who showed that the temperature of the processor was very close to Tjmax.

For this reason, starting with this article, I will use RealTemp 3.00 to monitor temperature. The comparisons made earlier remain valid, but it is not worth comparing the Speedfan and RealTemp readings.

I would also like to note that the copy of the processor that came to us had a somewhat uneven surface of the heat spreader cover. We used a high-quality thermal interface (Arctic Silver Cermamique), but the heatsink of the cooler warmed up relatively weakly, which indicates poor heat transfer.

In any case, we had no opportunities for further overclocking of the processor, and we achieved decent results. Of course, we can’t unambiguously say about the growth of the overclocking potential based on one copy, but this result is higher than the average for the C0 stepping. In addition, world statistics show that processors based on D0 overclock really better.

We conducted a full-scale test, in which we compared the performance of the 950 model (in normal mode and overclocked) with one of the fastest processors for Socket 775 — Core 2 Extreme QX9650, as well as with the Core i7-975, which we already reviewed earlier. Comparison with processors of the previous stepping was not carried out, since even in our review of the Core i7-975 it was found that there was no difference in performance. The results of the comparison with the Core i7-975 are also quite predictable, but the comparison with the QX9650 can be interesting.

The following test bench configuration was used:

Cooler: Thermaltake ISGC-300 + Fan-12
Motherboard:

Synthetics to start with: Sandra 2009, 3D Mark’06, PC Mark’05, Super Pi.

The stock frequencies of the QX9650 and i7-950 are quite close. We didn’t specifically align them; there is already plenty written about the performance of the Nehalem architecture in comparison with the Core. However, in all tests except PC Mark, the i7-950 is over 20% faster. In the PC Mark test, the gap is only about 5%.

Now let’s look at tests close to reality. Let’s start with games. Testing was carried out using Smart FPS. We tested performance in Crysis, Crysis: Warhead, Unreal Tournament 3 and Prey. The resolution used was 1280*1024, so the difference between the processors should have become more noticeable.

However, Crysis ran into graphics card performance even in such a gentle mode. Also here we are in for a surprise — i7 even with overclocking cannot overtake the QX9650. We didn’t experiment with disabling Hyper-threading, although it does improve performance in some games.

Unreal Tournament 3 and Prey showed a much greater dependence on the processor frequency. In the Prey i7-950 in normal mode, it also outperformed the Core by more than 10%. However, in general, a slight increase in performance in games can be noted. It is unlikely that many people, having such a computer configuration, will play at 1280 * 1024 and medium graphics settings, and at other resolutions the advantage will be even less significant.

The following is a group of benchmarks comparing processor performance in other real use cases:

These tasks show a much greater (close to linear) dependence on processor frequency. Only archiving accelerates rather weakly, since the memory subsystem and the hard disk are of great importance for it. The Nehalem architecture performs very well in these tasks. The advantage in Paint.NET is particularly impressive. Thus, for these purposes, the purchase of a Core i7 processor can be considered justified.

Products provided by

Intel	Representative office of Intel in Ukraine	www.intel.ua
Thermal right
ASUS	MTI	www.mti.ua
MSI	MSI Representative Office in Ukraine	www.msi.com
Kingston	Representative office of Kingston in Ukraine	www.kingston.com
WD	Western Digital	www.wdc.com
Samsung	Samsung Representative Office in Ukraine	www.samsung.ua