Cpu gflops chart: Processors ranking: GFLOPS performance — GadgetVersus

When the settings window 9 opens0017 graphics, select the option you prefer for graphics performance. ^(Graphics) Three options are displayed, namely

• Let Windows decide (energy saving).
• Energy saving.
• High performance.

When finished, press the button » Save ^{(Save )} » and exit the settings.

If your PC has multiple high-end GPUs ^(GPUs) and you want to specify which of these GPUs ^(GPUs) should be used for high performance use cases, now you can do so.

They also added the ability to specify which ^(GPU) graphics processor (GPU) you want to run the application on using the new option «Specific graphics processor». ^(GPU)

while Windows manages everything on its own, but it’s great that this option is available for user management. If you have a heavy application that uses the graphics processor ^(GPU) , you can force it to use the power saving graphics processor ^(GPU) to save battery power. You can bookmark this tip under Battery saving tips ^{for your laptop.}

This will also help reduce the load on the main integrated graphics processor ^(GPU) and make it easier to perform two tasks, one medium and one heavy.

To remove, select the application and click the button » Remove ^(Remove) » . It won’t ask for confirmation, so be careful.

How normal is GPU usage?

No such threshold, so even 100% usage GPU ^(GPU) is considered normal. However, this means that you are using your hardware to the fullest and leaving no performance on the table. Such cases usually occur during graphics-intensive games.

How to improve computer graphics?

The most preferred way is to increase FPS or frames ^{(Frame Per Second)} per second on your PC. You can do this by updating your graphics and video drivers, by optimizing your game settings, or by investing in upgrade software FPS .

Hope this helps!

Related posts

• How to backup and restore GPU Preferences for Apps in Windows 10

• Fix PowerShell causing High CPU usage in Windows 11/10

• How to use Performance Monitor in Windows 10

• How to install or uninstall Microsoft Store Apps on Windows 10

• Using the iTunes app from Microsoft Store on Windows 10

• Optimize and improve Windows 10 computer performance Windows 10

• Can’t Pin Apps or Programs to the Taskbar in Windows 10

• 3647

• How to fix 100% Disk, High CPU, High Memory usage in Windows 10

• How to view HEIC and HEVC files on Windows 10 Photos app

• How to run Computer Performance Benchmark Test on Windows 10

• Top 10 free Epub readers for Windows 11/10

• Windows Store Cache can be corrupted in Windows 11/10

• Tune-Up Windows 10 using these tips and free software

• performance of Intel processors of different generations / Habr

  Almost every year a new generation of Intel Xeon E5 central processors enters the market. In each generation, the socket and technological process alternately change. There are more and more cores, and the heat dissipation is gradually decreasing. But a natural question arises: “What does the new architecture give to the end user?”

  To do this, I decided to test the performance of similar processors of different generations. I decided to compare models of the mass segment: 8-core processors 2660, 2670, 2640V2, 2650V2, 2630V3 and 2620V4. Testing with such a generational spread is not entirely fair, because between V2 and V3 there is a different chipset, a new-generation memory with a higher frequency, and most importantly, there are no direct peers in frequency among models of all 4 generations. But, in any case, this study will help to understand to what extent the performance of new processors has increased in real applications and synthetic tests.

  The selected line of processors has many similar parameters: the same number of cores and threads, 20 MB SmartCache, 8 GT/s QPI (except 2640V2) and the number of PCI-E lanes equal to 40.

  To assess the feasibility of testing all processors, I turned to the test results PassMark.

  Below is a summary graph of the results:

  Since the frequency is significantly different, it is not entirely correct to compare the results. But despite this, conclusions immediately suggest themselves:

  1. 2660 equivalent in performance to 2620V4
  2. 2670 outperforms 2620V4 (obviously due to frequency)
  3. 2640V2 sags, and 2650V2 beats everyone (also because of the frequency)

  I divided the result by the frequency and got a certain performance value at 1 GHz:

  Here the results are more interesting and visual:

  1. 2660 and 2670 — an unexpected run for me within one generation, 2670 justifies only that its overall performance is very high
  2. 2640V2 and 2650V2 — a very strange low result, which is worse than 2660
  3. 2630V3 and 2620V4 are the only logical growth (apparently just due to the new architecture. ..)

  , and not very successful, in my opinion — I remove
  from the candidates
  2. 2630V3 is an excellent result, but it is unreasonably more expensive than 2620V4, given the similar performance and, moreover, this is already an outgoing generation of 9 processors3731
  3. 2620V4 is an adequate price (compared to 2630V3), high performance and, most importantly, this is the only model of the latest generation 8-core processor with Hyper-threading in our list, so we definitely leave
  for further tests
  4. 2660 and 2670 are excellent results compared to 2620V4. In my opinion, it is the comparison of the first and last (at the moment) generation in the Intel Xeon E5 line that is of particular interest. In addition, we still have sufficient stocks of first-generation processors in stock, so this comparison is very relevant for us.

  The cost of servers based on 2660 and 2620V4 processors can differ by almost 2 times against the latter, so by comparing their performance and choosing a server based on V1 processors, you can significantly reduce the budget for buying a new server. But I will talk about this proposal after the results of testing.

  3 stands were assembled for testing:

  1. 2 x Xeon E5-2660, 8 x 8Gb DDR3 ECC REG 1333, SSD Intel Enterprise 150Gb
  2. 2 x Xeon E5-2670, 8 x 8Gb DDR3 ECC REG 1333, SSD Intel Enterprise 150Gb
  3. 2 x Xeon E5-2620V4, 8 x 8Gb DDR4 ECC REG 2133, SSD Intel Enterprise 150Gb

  PassMark PerformanceTest 9.0

  When selecting processors for tests, I already used the results of synthetic tests, but now it is interesting to compare these models in more detail. The comparison was made in groups: the 1st generation against the 4th.

  A more detailed test report allows us to draw some conclusions:

  1. Mathematics, incl. and floating point, mainly depends on the frequency. The difference of 100 MHz allowed the 2660 to outperform the 2620V4 in computational operations, in encryption and compression (and this despite a significant difference in memory frequency)
  2. Physics and calculations using extended instructions on the new architecture are performed better, despite the low frequency
  3. And, of course, the test with the use of memory passed in favor of the V4 processors, since in this case different generations of memory competed — DDR4 and DDR3.

  It was synthetic. Let’s see what specialized benchmarks and real applications will show.

  7ZIP archiver

  Here the results are similar to the previous test — direct binding to the processor frequency. It does not matter that a slower memory is installed — V1 processors confidently take the lead in frequency.

  CINEBENCH R15

  CINEBENCH is a computer performance benchmark for MAXON Cinema 4D professional animation software.

  Xeon E5-2670 pulled out in frequency and beat 2620V4. But the E5-2660, which has a less visible advantage in frequency, lost to the 4th generation processor. Hence the conclusion — this software uses useful additions to the new architecture (although it may be all about memory . ..), but not so much that it was a decisive factor.

  3DS MAX + V-Ray

  To evaluate the performance of processors when rendering in a real application, I took a bunch: 3ds Max 2016 + V-ray 3.4 + a real scene with several light sources, specular and transparent materials, and an environment map.

  The results were similar to CINEBENCH: the Xeon E5-2670 had the lowest render time, while the 2660 couldn’t beat the 2620V4.

  1: SQL/File

  In conclusion of testing, I enclose the results of gilev tests for 1C.

  When testing a database with file access, the E5-2620V4 processor confidently leads. The table shows the average values ​​of 20 runs of the same test. The difference between the results of each stand in the case of the file base was no more than 2%.

  A single-threaded SQL database test showed very strange results. The difference turned out to be negligible, given the different frequencies for 2660 and 2670, and the different frequencies for DDR3 and DDR4. There was an attempt to optimize the SQL settings, but the results turned out to be worse than it was, so I decided to test all the stands on the basic settings.

  The results of the multi-threaded SQL test turned out to be even more strange and inconsistent. The maximum speed of 1 thread in MB/s was equivalent to the performance index in the previous single-threaded test.

  The next parameter was the maximum speed (of all streams) — the result was almost identical for all stands. Since the results of different runs fluctuated greatly (+ -5%) — sometimes they were at different stands with a significant margin both in one direction and in the other. The same average multi-threaded SQL test results lead me to 3 thoughts:

  1. This situation is caused by an unoptimized SQL configuration
  2. SSD became the bottleneck of the system and did not allow processors to overclock
  3. There is almost no difference between the frequency of memory and processors for these tasks (which is extremely unlikely)

  If you have reliable explanations for such results, please share them in the comments.

  Also, the result for the “Recommended number of users” parameter turned out to be inexplicable. The average result of the 2660 turned out to be the highest of all — and this is despite the low results of all tests.
  I will also be glad to see your comments on this issue.

  Terminals

  The results of several comprehensive computational tests showed that the frequency of the processor in most cases turned out to be more important than the generation, architecture, and even memory frequency. Of course, there is modern software that uses all the improvements of the new architecture. For example, video transcoding is sometimes performed incl. using AVX2.0 instructions, but this is specialized software — and most server applications are still tied to the number and frequency of cores.

  Of course, I’m not saying that there is absolutely no difference between processors, I just want to note that for certain applications there is no point in a «planned» transition to a new generation.

  If you do not agree with me or you have suggestions for testing, the stands have not yet been sorted out, and I will be happy to test your tasks.

  Economic benefit

  As I wrote at the beginning of the article, we offer a line of servers based on the first generation Xeon E5 processors, which are significantly more expensive than servers based on the E5-2620V4.
  These are the same new servers (not to be confused with used ones) with a 3-year warranty.

  Below is an approximate calculation:

  STSS Flagman server RX227.4-008LH in the configuration 2 x Intel Xeon E5-2620V4 + 8 x 8Gb DDR4 ECC REG at retail costs today

  r 265066

  Similar configuration STSS Flagman EX227.3-008LH based on 2 x Intel Xeon E5-2660 + 8 x 8Gb DDR3 ECC REG is available for 175275r.

  Habr readers can get an additional discount when ordering 5% . To do this, you need to select the desired case form factor from the list of models on our website . Model EX217.3-004LH is 1U, EX227.3-008LH is 2U, and EX240.3-008LH is Tower/4U.
  In the model configurator, you can select the necessary parameters of memory, disk subsystem and additional devices. When sending a request for calculation, you must specify the promo code HABRAHABR .

  Thank you for your attention! I will wait for your comments and wishes on the tests.

  Article writing and testing: Usikoff
  Testing 1C: sarge74

  CPU performance table

  1	256	CPU AMD Athlon X4 730 (AD730XO) 2.8 GHz/4core/ 4 Mb/65W/5 GT/s Socket FM2	9.4	1819	193.51
  2	210	CPU AMD Athlon X4 845 (AD845XA) 3.5 GHz/4core/ 2 Mb/65W/5 GT/s Socket FM2+	13.2	2658	201.36
  3	224	CPU AMD Athlon X4 840 (AD840XY) 3. 1 GHz/4core/ 4 Mb/65W/5 GT/s Socket FM2+	12.5	2565	205.2
  4	252	CPU Intel Celeron G3930 2.9 GHz/2core/SVGA HD Graphics 610/0.5+2Mb/51W/8GT/s LGA1151	9.9	2297	232.02
  5	254	CPU Intel Celeron G3900 2.8 GHz/2core/SVGA HD Graphics 510/0.5+2Mb/51W/8GT/s LGA1151	9.7	2320	239.18
  6	275	CPU AMD Athlon X2 370K (AD370KO) 4.0 GHz/2core/ 1 Mb/65W/5 GT/s Socket FM2	6	1487	247.83
  7	228	CPU AMD Athlon X4 860K (AD860KX) 3.7 GHz/4core/ 4 Mb/95W/5 GT/s Socket FM2+	12.4	3089	249.11
  8	172	CPU AMD FX-6300 (FD6300W) 3. 5 GHz/6core/ 6+8Mb/95W/5200 MHz Socket AM3+	17.1	4605	269.3
  9	197	CPU AMD Athlon X4 870K (AD870KX) 3.9GHz/4core/ 4Mb/95W/5 GT/s Socket FM2+	13.8	3730	270.29
  10	278	CPU AMD A4-6300 (AD6300O) 3.7 GHz/2core/SVGA Radeon HD 8370D/ 1 Mb/65W/5 GT/s Socket FM2	5.9	1609	272.71
  11	223	CPU AMD FX-4300 (FD4300W) 3.8 GHz/4core/ 4+4Mb/95W/5200 MHz Socket AM3+	12.6	3439	272.94
  12	132	CPU AMD FX-8320 (FD8320F) 3.5 GHz/8core/ 8+8Mb/125W/5200 MHz Socket AM3+	22.4	6120	273.21
  13	206	CPU AMD FX-4330 (FD4330W) 4. 0 GHz/4core/ 4+8Mb/95W/5200 MHz Socket AM3+	13.3	3730	280.45
  14	196	CPU AMD FX-4350 (FD4350F) 4.2 GHz/4core/ 4+8Mb/125W/5200 MHz Socket AM3+	13.9	3905	280.94
  15	148	CPU AMD FX-8320E (FD832EW) 3.2 GHz/8core/ 8+8Mb/95W/5200 MHz Socket AM3+	20.7	5887	284.4
  16	216	CPU AMD Athlon X4 950 (AD950XA) 3.5 GHz/4core/2 Mb/65W/5 GT/s Socket AM4	13.1	3730	284.73
  17	285	CPU AMD A4-4000 (AD4000O) 3.0 GHz/2core/SVGA RADEON HD 7480D/ 1 Mb/65W/5 GT/s Socket FM2	5.2	1492	286.92
  18	282	CPU AMD A4-5300 (AD5300O) 3. 4 GHz/2core/SVGA RADEON HD 7480D/ 1 Mb/65W/5 GT/s Socket FM2	5.6	1627	290.54
  19	232	CPU AMD A8-7500 (AD7500Y) 3.0 GHz/4core/SVGA RADEON R7/ 4 Mb/65W/5 GT/s Socket FM2+	12.3	3614	293.82
  20	283	CPU AMD Athlon X2 340 (AD340XO) 3.2 GHz/2core/ 1 Mb/65W/5 GT/s Socket FM2	5.3	1562	294.72
  21	187	CPU AMD FX-6100 (FD6100W) 3.3 GHz/6core/ 6+8Mb/95W/5200 MHz Socket AM3+	15.1	4488	297.22
  22	240	CPU Intel Pentium G4400 3.3 GHz/2core/SVGA HD Graphics 510/0.5+3Mb/54W/8 GT/s LGA1151	11.3	3381	299.2
  23	144	CPU AMD FX-8300 (FD8300W) 3. 3 GHz/8core/ 8+8Mb/95W/5200 MHz Socket AM3+	21.2	6353	299.67
  24	284	CPU AMD A4-4020 (AD4020O) 3.2 GHz/2core/SVGA RADEON HD 7480D/ 1 Mb/65W/5 GT/s Socket FM2	5.2	1574	302.69
  25	110	CPU AMD FX-8350 (FD8350F) 4.0 GHz/8core/ 8+8Mb/125W/5200 MHz Socket AM3+	25.2	7635	302.98
  26	183	CPU Intel Pentium G4600 3.6 GHz/2core/SVGA HD Graphics 630/0.5+3Mb/51W/8GT/s LGA1151	15.4	4721	306.56
  27	274	CPU AMD A4-6320 (AD6320O) 3.8 GHz/2core/SVGA Radeon HD 8370D/ 1 Mb/65W/5 GT/s Socket FM2	6.1	1895	310.66
  28	157	CPU AMD FX-6350 (FD6350F) 3. 9 GHz/6core/ 6+8Mb/125W/5200 MHz Socket AM3+	18.9	5945	314.55
  29	250	CPU Intel Celeron G3950 3.0 GHz/2core/SVGA HD Graphics 610/0.5+2Mb/51W/8GT/s LGA1151	10	3148	314.8
  30	291	CPU AMD Athlon II X2 240 (ADX240O) 2.8 GHz/2core/ 2Mb/65W/ 4000MHz Socket AM3	4.5	1475	327.78
  31	179	CPU Intel Pentium G4620 3.7 GHz/2core/SVGA HD Graphics 630/0.5+3Mb/51W/8GT/s LGA1151	15.8	5187	328.29
  32	190	CPU Intel Pentium G4560 3.5 GHz/2core/SVGA HD Graphics 610/0.5+3Mb/54W/8GT/s LGA1151	15	5071	338.07
  33	226	CPU AMD A8-7600 (AD7600Y) 3. 1 GHz/4core/SVGA RADEON R7/ 4 Mb/65W/5 GT/s Socket FM2+	12.4	4197	338.47
  34	288	CPU AMD Athlon II X2 245 (ADX245O) 2.9 GHz/2core/ 2Mb/65W/ 4000MHz Socket AM3	4.6	1562	339.57
  35	258	CPU Intel Celeron G3900T 2.6 GHz/2core/SVGA HD Graphics 510/0.5+2Mb/35W/8GT/s LGA1151	9	3089	343.22
  36	122	CPU Intel Core i3-8100 3.6 GHz/4core/SVGA UHD Graphics 630/ 6Mb/65W/8 GT/s LGA1151	24	8393	349.71
  37	231	CPU AMD A8 9600 (AD9600AG) 3.1 GHz/4core/SVGA RADEON R7/ 2 Mb/65W Socket AM4	12.3	4313	350.65
  38	233	CPU AMD A8-7650K (AD765KX) 3. 3 GHz/4core/SVGA RADEON R7/ 4 Mb/95W/ Socket FM2+	12.3	4313	350.65
  39	235	CPU Intel Pentium G4500 3.5 GHz/2core/SVGA HD Graphics 530/0.5+3Mb/51W/8 GT/s LGA1151	12.1	4255	351.65
  40	279	CPU AMD A6-5400K (AD540KO) 3.6 GHz/2core/SVGA RADEON HD 7540D/ 1 Mb/65W/5 GT/s Socket FM2	5.8	2046	352.76
  41	272	CPU AMD A4-7300 (AD7300O) 3.8 GHz/2core/SVGA Radeon HD 8470D/ 1 Mb/65W/5 GT/s Socket FM2	6.2	2203	355.32
  42	289	CPU AMD Athlon II X2 250 (ADX250O) 3.0 GHz/2core/ 2Mb/65W/ 4000MHz Socket AM3	4.6	1644	357.39
  43	53	CPU AMD Ryzen 5 1600 (YD1600B) 3. 2 GHz/6core/3+16Mb/65W Socket AM4	39.1	13988	357.75
  44	112	CPU AMD FX-8370 (FD8370F) 4.0 GHz/8core/ 8+8Mb/125W/5200 MHz Socket AM3+	25.2	9150	363.1
  45	199		13.7	5071	370.15
  46	260	CPU Intel Pentium G2030 3.0 GHz/2core/SVGA HD Graphics/0.5+3Mb/55W/5 GT/s LGA1155	8.8	3264	370.91
  47	217	CPU AMD A8-7670K (AD767KX) 3.9 GHz/4core/SVGARADEON R7/4 Mb/95W/5 GT/s Socket FM2+	13	4838	372.15
  48	99	CPU AMD FX-9370 (FD9370F) 4.4 GHz/8core/ 8+8Mb/220W/5200 MHz Socket AM3+	26.6	9966	374. 66
  49	158	CPU AMD Ryzen 3 1200 (YD1200B) 3.1 GHz/4core/2+8Mb/65W Socket AM4	18.8	7052	375.11
  50	263	CPU Intel Pentium G2020 2.9 GHz/2core/SVGA HD Graphics/0.5+3Mb/55W/5 GT/s LGA1155	8.5	3206	377.18
  51	41	CPU AMD Ryzen 5 1600X (YD160XB) 3.6 GHz/6core/3+16Mb/95W socket AM4	42.6	16435	385.8
  52	264	CPU Intel Pentium G2010 2.8 GHz/2core/SVGA HD Graphics/0.5+3Mb/55W/5 GT/s LGA1155	8.4	3322	395.48
  53	103	CPU AMD Ryzen 5 1400 (YD1400B) 3.2 GHz/4core/2+8Mb/65W Socket AM4	25.9	10316	398.3
  54	220	CPU AMD A10-7800 (AD7800Y) 3. 5 GHz/4core/SVGA RADEON R7/ 4Mb/65W/5 GT/s Socket FM2+	12.7	5071	399.29
  55	267	CPU AMD A6-7400K (AD740KY) 3.5 GHz/2core/SVGA RADEON R5/ 1Mb/65W/5 GT/s Socket FM2+	6.5	2629	404.46
  56	244	CPU Intel Core 2 Quad Q8300 2.5 GHz/4core/ 4Mb/95W/ 1333MHz LGA775	10.8	4371	404.72
  57	237	CPU Intel Pentium G4520 3.6 GHz/2core/SVGA HD Graphics 530/0.5+3Mb/51W/8 GT/s LGA1151	12	4896	408
  58	84	CPU AMD Ryzen 5 1500X (YD150XB) 3.5 GHz/4core/2+16Mb/65W Socket AM4	29.7	12298	414.07
  59	143	CPU AMD Ryzen 3 1300X (YD130XB) 3. 5 GHz/4core/2+8Mb/65W Socket AM4	21.5	8917	414.74
  60	259	CPU Intel Core 2 Quad Q8200 2.33 GHz/4core/ 4Mb/95W/ 1333MHz LGA775	8.9	3730	419.1
  61	270	CPU AMD A6-6400K (AD640KO) 3.9 GHz/2core/SVGA RADEON HD 8470D/ 1 Mb/65W/5 GT/s Socket FM2	6.3	2652	420.95
  62	92	CPU AMD FX-9590 (FD9590F) 4.7 GHz/8core/ 8+8Mb/220W/5200 MHz Socket AM3+	28.2	11890	421.63
  63	202	CPU AMD A10-7870K (AD787KX) 3.9GHz/4core/SVGA RADEON R7/ 4 Mb/95W/5 GT/s Socket FM2+	13.6	5945	437.13
  64	215	CPU Intel Pentium G4600T 3. 0 GHz/2core/SVGA HD Graphics 630/0.5+3Mb/35W/8GT/s LGA1151	13.1	5770	440.46
  65	31	CPU AMD Ryzen 7 1700 (YD1700B) 3.0 GHz/8core/4+16Mb/65W Socket AM4	46.9	20865	444.88
  66	171	CPU Intel Core i3-7100 3.9 GHz/2core/SVGA HD Graphics 630/0.5+ 3Mb/51W/8 GT/s LGA1151	17.1	7693	449.88
  67	211	CPU AMD A10-9700 (AD9700AG) 3.5 GHz/4core/SVGA RADEON R7/2 Mb/65W/Socket AM4	13.2	6003	454.77
  68	219	CPU AMD A10-7860K (AD786KY) 3.6 GHz/4core/SVGA RADEON R7/ 4 Mb/65W/5 GT/s Socket FM2+	12.8	5887	459.92
  69	271	CPU AMD A6-7470K (AD747KY) 3. 7 GHz/2core/SVGA RADEON R5/ 1Mb/65W/5 GT/s Socket FM2+	6.3	2903	460.79
  70	177	CPU Intel Core i3-6100 3.7 GHz/2core/SVGA HD Graphics 530/0.5+ 3Mb/51W/8 GT/s LGA1151	16.4	7635	465.55
  71	249	CPU Intel Core 2 Quad Q8400 2.66 GHz/4core/ 4Mb/95W/ 1333MHz LGA775	10	4663	466.3
  72	280	CPU Intel Core 2 Duo E8300 2.83 GHz/2core/ 6Mb/65W/ 1333MHz LGA775	5.8	2751	474.31
  73	290	CPU Intel Core 2 Duo E7200 2.53 GHz/2core/ 3Mb/65W/ 1066MHz LGA775	4.6	2186	475.22
  74	213	CPU AMD A10-7850K (AD785KX) 3. 7 GHz/4core/SVGA RADEON R7/ 4 Mb/95W/5 GT/s Socket FM2+	13.1	6236	476.03
  75	97	CPU Intel Core i3-8350K 4.0 GHz/4core/SVGA UHD Graphics 630/ 8Mb/91W/8 GT/s LGA1151	26.7	12822	480.22
  76	286	CPU Intel Core 2 Duo E8200 2.66 GHz/2core/ 6Mb/65W/ 1333MHz LGA775	5.1	2460	482.35
  77	247	CPU Intel Pentium G4400T 2.9 GHz/2core/SVGA HD Graphics 510/0.5+3Mb/35W/8 GT/s LGA1151	10.1	4896	484.75
  78	238	CPU AMD A10 9700E (AD9700AH) 3.0 GHz/4core/SVGA RADEON R7/2 Mb/35W Socket AM4	12	5828	485.67
  79	245	CPU Intel Pentium G3260 3. 3 GHz/2core/SVGA HD Graphics/0.5+3Mb/53W/5 GT/s LGA1150	10.2	4954	485.69
  80	115	CPU Intel Core i5-6400 2.7 GHz/4core/SVGA HD Graphics 530/1+6Mb/65W/ LGA1151	24.6	11948	485.69
  81	261	CPU Intel Celeron G1840 2.8 GHz/2core/SVGA HD Graphics/0.5+2Mb/53W/5GT/s LGA1150	8.8	4371	496.7
  82	194	CPU AMD A10-7890K (AD789KX) 4.1 GHz/4core/SVGA RADEON R7/4 Mb/95W/5 GT/s Socket FM2+	13.9	7052	507.34
  83	166	CPU Intel Core i3-7300 4 GHz/2core/SVGA HD Graphics 630/ 4Mb/ LGA1151	17.7	9092	513.67
  84	277	CPU Intel Core 2 Duo E8400 3. 0 GHz/2core/ 6Mb/65W/ 1333MHz LGA775	5.9	3089	523.56
  85	164	CPU Intel Core i3-7320 4.1 GHz/2core/SVGA HD Graphics 630/ 4Mb/ LGA1151	18.1	9500	524.86
  86	51	CPU Intel Core i5-8600K 3.6 GHz/6core/SVGA UHD Graphics 630/1.5+9Mb/95W/8 GT/s LGA1151	39.3	20632	524.99
  87	169	CPU Intel Core i3-6320 3.9 GHz/2core/SVGA HD Graphics 530/0.5+ 4Mb/51W/8 GT/s LGA1151	17.4	9209	529.25
  88	135	CPU Intel Core i5-7400 3 GHz/4core/SVGA HD Graphics 630/1+6Mb/65W/8 GT/s LGA1151	22.3	11831	530.54
  89	161	CPU Intel Core i3-7350K 4. 2 GHz/2core/SVGA HD Graphics 630/ 4Mb/60W/8 GT/s LGA1151	18.5	9908	535.57
  90	269	CPU AMD A6-6420K (AD642KO) 4.0 GHz/2core/SVGA RADEON HD 8470D/ 1 Mb/65W/5 GT/s Socket FM2	6.4	3439	537.34
  91	265	CPU Intel Celeron G1840T 2.5 GHz/2core/SVGA HD Graphics/0.5+2Mb/35W/5 GT/s LGA1150	8	4313	539.13
  92	175	CPU Intel Core i3-6300 3.8 GHz/2core/SVGA HD Graphics 530/0.5+ 4Mb/51W/8 GT/s LGA1151	17	9267	545.12
  93	201	CPU AMD A12 9800 (AD9800AU) 3.8 GHz/4core/SVGA RADEON R7/2 Mb/65W/Socket AM4	13.6	7460	548.53
  94	121	CPU Intel Core i5-7500 3. 4 GHz/4core/SVGA HD Graphics 630/1+6Mb/65W/8 GT/s LGA1151	24.1	13521	561.04
  95	234	CPU Intel Core i3-3240 3.4 GHz/2core/SVGA HD Graphics 2500/0.5+3Mb/55W/5 GT/s LGA1155	12.1	6819	563.55
  96	268	CPU AMD A6 9500E (AD9500AH) 3.0 GHz/2core/SVGA RADEON R5/ 1 Mb/35W Socket AM4	6.4	3614	564.69
  97	192	CPU Intel Core i3-4170 3.7 GHz/2core/SVGA HD Graphics 4400/0.5+3Mb/54W/5 GT/s LGA1150	14.4	8160	566.67
  98	25	CPU Intel Core i7-8700K 3.7 GHz/6core/SVGA UHD Graphics 630/1.5+12Mb/95W/8 GT/s LGA1151	50.6	29257	578.2
  99	189	CPU Intel Core i3-7100T 3. 4 GHz/2core/SVGA HD Graphics 630/0.5+ 3Mb/35W/8 GT/s LGA1151	15	8684	578.93
  100	76	CPU Intel Xeon E3-1230 V6 3.5 GHz/4core/1+8Mb/72W/8 GT/s LGA1151	31.7	18417	580.98
  101	146	CPU Intel Core i5-4570 3.2 GHz/4core/SVGA HD Graphics 4600/1+6Mb/84W/5 GT/s LGA1150	21.1	12414	588.34
  102	102	CPU Intel Core i5-7600 3.5 GHz/4core/SVGA HD Graphics 630/1+6Mb/65W/8 GT/s LGA1151	26	15386	591.77
  103	80	CPU Intel Xeon E3-1230 V5 3.4 GHz/4core/1+8Mb/80W/8 GT/s LGA1151	30.8	18300	594.16
  104	71	CPU Intel Xeon E3-1240 V6 3. 7 GHz/4core/1+8Mb/72W/8 GT/s LGA1151	33.2	20107	605.63
  105	151	CPU Intel Core i5-4460 3.2 GHz/4core/SVGA HD Graphics 4600/1+6Mb/84W/5 GT/s LGA1150	19.9	12064	606.23
  106	155	CPU Intel Core i5-4440 3.1 GHz/4core/SVGA HD Graphics 4600/1+6Mb/84W/5 GT/s LGA1150	19.3	11715	606.99
  107	257	CPU Intel Core 2 Quad Q9400 2.66 GHz/4core/ 6Mb/95W/ 1333MHz LGA775	9.1	5537	608.46
  108	141	CPU Intel Core i5-6402P 2.8 GHz/4core/SVGA HD Graphics 510/1+6Mb/65W/8 GT/s LGA1151	21.6	13172	609.81
  109	193	CPU Intel Core i3-6100T 3. 2 GHz/2core/SVGA HD Graphics 530/0.5+3Mb/35W/ LGA1151	14.3	8801	615.45
  110	139	CPU Intel Core i5-6500 3.2 GHz/4core/SVGA HD Graphics 530/1+6Mb/65W/ LGA1151	22.1	13638	617.1
  111	294	CPU Intel Pentium E6300 2.8 GHz/2core/ 2Mb/65W/ 1066MHz LGA775	3.9	2407	617.18
  112	105	CPU Intel Core i5-6600K 3.5 GHz/4core/SVGA HD Graphics 530/1+6Mb/91W/ LGA1151	25.8	15969	618.95
  113	118	CPU Intel Core i5-6600 3.3 GHz/4core/SVGA HD Graphics 530/1+6Mb/65W/ LGA1151	24.2	15095	623.76
  114	73	CPU Intel Xeon E3-1240 V5 3. 5 GHz/4core/1+8Mb/80W/8 GT/s LGA1151	31.8	19991	628.65
  115	96	CPU Intel Xeon E3-1230 V3 3.3 GHz/4core/1+8Mb/80W/5 GT/s LGA1150	26.9	16960	630.48
  116	55	CPU Intel Core i7-5820K 3.3 GHz/6core/1.5+15Mb/140W/5 GT/s LGA2011-3	37.8	23895	632.14
  117	242	CPU Intel Core 2 Quad Q9450 2.66 GHz/4core/ 12Mb/95W/ 1333MHz LGA775	10.9	6936	636.33
  118	70	CPU Intel Xeon E3-1245 V6 3.7 GHz/4core/SVGA HD Graphics P630/1+8Mb/73W/8 GT/s LGA1151	33.2	21214	638.98
  119	136	CPU Intel Xeon E3-1220 V6 3.0 GHz/4core/1+8Mb/72W/8 GT/s LGA1151	22. 3	14279	640.31
  120	72	CPU Intel Xeon E5-2660 2.2 GHz/8core/2+20Mb/95W/8 GT/s LGA2011	33.2	21273	640.75
  121	65	CPU Intel Core i7-7700 3.6 GHz/4core/SVGA HD Graphics 630/1+8Mb/65W/8 GT/s LGA1151	34	21797	641.09
  122	137	CPU Intel Xeon E3-1220 V5 3.0 GHz/4core/1+8Mb/80W/8 GT/s LGA1151	22.3	14396	645.56
  123	81	CPU Intel Xeon E5-2650 2.0 GHz/8core/2+20Mb/95W/8 GT/s LGA2011	29.7	19233	647.58
  124	44	CPU Intel Core i7-6800K 3.4 GHz/6core/1.5+15Mb/140W LGA2011-3	41.6	27159	652. 86
  125	101	CPU Intel Xeon E3-1231 V3 3.4 GHz/4core/1+8Mb/80W/5 GT/s LGA1150	26.4	17251	653.45
  126	124	CPU Intel Xeon E3-1225 V6 3.3 GHz/4core/SVGA HD Graphics P630/1+8Mb/73W/8 GT/s LGA1151	23.8	15561	653.82
  127	113	CPU Intel Xeon E5-2630 2.3 GHz/6core/1.5+15Mb/95W/7.2 GT/s LGA2011	24.8	16261	655.69
  128	58	CPU Intel Xeon E5-2670 2.6 GHz/8core/2+20Mb/115W/8 GT/s LGA2011	36.9	24245	657.05
  129	185	CPU Intel Core i3-4370 3.8 GHz/2core/SVGA HD Graphics4600/0.5+4Mb/54W/5 GT/s LGA1150	15.1	9966	660
  130	59	CPU Intel Core i7-7700K 4. 2 GHz/4core/SVGA HD Graphics 630/1+8Mb/91W/8 GT/s LGA1151	36.7	24420	665.4
  131	79	CPU Intel Xeon E3-1245 V5 3.5 GHz/4core/SVGA HD Graphics P530/1+8Mb/80W/8 GT/s LGA1151	30.9	20632	667.7
  132	89	CPU Intel Core i7-4770 3.4 GHz/4core/SVGA HD Graphics 4600/1+8Mb/84W/5 GT/s LGA1150	28.3	18941	669.29
  133	149	CPU Intel Xeon E3-1220 V3 3.1 GHz/4core/1+8Mb/80W/5 GT/s LGA1150	20.4	13696	671.37
  134	188	CPU Intel Core i3-6300T 3.3 GHz/2core/SVGA HD Graphics 530/0.5+4Mb/35W/ LGA1151	15	10083	672.2
  135	128	CPU Intel Core i5-4690 3. 5 GHz/4core/SVGA HD Graphics4600/1+6Mb/84W/5 GT/s LGA1150	22.8	15445	677.41
  136	293	CPU Intel Core 2 Duo E6550 2.33 GHz/2core/ 4Mb/65W/ 1333MHz LGA775	3.9	2687	688.97
  137	36	CPU Intel Core i7-8700 3.2 GHz/6core/SVGA UHD Graphics 630/1.5+12Mb/65W/8 GT/s LGA1151	44.7	30889	691.03
  138	125	CPU Intel Xeon E3-1225 V5 3.3 GHz/4core/SVGA HD Graphics P530/1+8Mb/80W/8 GT/s LGA1151	23	15911	691.78
  139	64	CPU Intel Core i7-6700K 4.0 GHz/4core/SVGA HD Graphics 530/1+8Mb/91W/8 GT/s LGA1151	34.2	24128	705.5
  140	152	CPU Intel Core i5-3570 3. 4 GHz/4core/SVGA HD Graphics 2500/1+6Mb/77W/5 GT/s LGA1155	19.8	14163	715.3
  141	156	CPU Intel Core i5-4460S 2.9 GHz/4core/ LGA1150	19	13638	717.79
  142	140	CPU Intel Xeon E3-1226 V3 3.3 GHz/4core/SVGA HD Graphics P4600/1+8Mb/84W/5 GT/s LGA1150	21.9	15736	718.54
  143	68	CPU Intel Xeon E3-1270 V6 3.8 GHz/4core/1+8Mb/72W/8 GT/s LGA1151	33.9	24420	720.35
  144	170	CPU Intel Xeon E5-2603 V4 1.7 GHz/6core/1.5+15Mb/85W/6.4 GT/s LGA2011-3	17.3	12647	731.04
  145	126	CPU Intel Core i5-7600K 3.8 GHz/4core/SVGA HD Graphics 630/1+6Mb/91W/8 GT/s LGA1151	22. 8	16669	731.1
  146	38	CPU Intel Core i7-6850K 3.6 GHz/6core/1.5+15Mb/140W LGA2011-3	43.3	31880	736.26
  147	109	CPU Intel Xeon E3-1241 V3 3.5 GHz/4core/1+8Mb/80W/5 GT/s LGA1150	25.5	18941	742.78
  148	69	CPU Intel Xeon E3-1275 V6 3.8 GHz/4core/SVGA HD Graphics P630/1+8Mb/73W/8 GT/s LGA1151	33.9	25236	744.42
  149	87	CPU Intel Core i7-6700 3.4 GHz/4core/SVGA HD Graphics 530/1+8Mb/65W/8 GT/s LGA1151	28.8	21739	754.83
  150	83	CPU Intel Xeon E5-2620 V3 2.4 GHz/6core/1.5+15Mb/85W/8 GT/s LGA2011-3	29. 7	22497	757.47
  151	160	CPU Intel Xeon E3-1220 V2 3.1 GHz/4core/69W LGA1155	18.6	14104	758.28
  152	77	CPU Intel Xeon E3-1270 V5 3.6 GHz/4core/1+8Mb/80W/8 GT/s LGA1151	31.7	24070	759.31
  153	62	CPU Intel Xeon E5-2620 V4 2.1 GHz/8core/+20Mb/85W/8 GT/s LGA2011-3	35.8	27683	773.27
  154	90	CPU Intel Xeon E5-1620 V4 3.5 GHz/4core/1+10Mb/140W LGA2011-3	28.3	21914	774.35
  155	78	CPU Intel Core i7-4790K 4.0 GHz/4core/SVGA HD Graphics 4600/1+8Mb/88W/5 GT/s LGA1150	31.4	24362	775. 86
  156	243	CPU Intel Core 2 Quad Q9550 2.83 GHz/4core/ 12Mb/95W/ 1333MHz LGA775	10.8	8451	782.5
  157	292	CPU Intel Core 2 Duo E8500 3.16 GHz/2core/ 6Mb/65W/ 1333MHz LGA775	4.3	3381	786.28
  158	74	CPU Intel Xeon E3-1275 V5 3.6 GHz/4core/SVGA HD Graphics P530/1+8Mb/80W/8 GT/s LGA1151	31.7	25177	794.23
  159	222	CPU Intel Core i3-4350T 3.1 GHz/2core/SVGA HD Graphics 4600/0.5+4Mb/35W/5 GT/s LGA1150	12.6	10141	804.84
  160	108	CPU Intel Core i7-3770 3.4 GHz/4core/SVGA HD Graphics 4000/1+8Mb/77W/5 GT/s LGA1155	25.6	20748	810. 47
  161	49	CPU Intel Core i7-5930K 3.5 GHz/6core/1.5+15Mb/140W/5 GT/s LGA2011-3	41.2	33745	819.05
  162	295	CPU Intel Pentium E5500 2.8 GHz/2core/ 2Mb/65W/ 800MHz LGA775	2.8	2297	820.36
  163	182	CPU Intel Xeon E5-2603 V3 1.6 GHz/6core/1.5+15Mb/85W/6.4 GT/s LGA2011-3	15.4	12997	843.96
  164	241	CPU Intel Core 2 Quad Q9650 3.0 GHz/4core/ 12Mb/95W/ 1333MHz LGA775	10.9	9209	844.86
  165	297	CPU Intel Pentium Dual-Core E5400 2.7 GHz/2core/ 2Mb/65W/ 800MHz LGA775	2.7	2308	854.81
  166	116	CPU Intel Core i7-4790 3. 6 GHz/4core/SVGA HD Graphics 4600/1+8Mb/84W/5 GT/s LGA1150	24.3	21040	865.84
  167	287	CPU Intel Core 2 Duo E8600 3.33 GHz/2core/ 6Mb/65W/ 1333MHz LGA775	4.9	4313	880.2
  168	43	CPU Intel Xeon E5-2660 V2 2.2 GHz/10core/2.5+25Mb/95W/8 GT/s LGA2011	41.8	37300	892.34
  169	248	CPU Intel Xeon E5-2603 V2 1.8 GHz/4core/1+10Mb/80W/6.4 GT/s LGA2011	10.1	9209	911.78
  170	119	CPU Intel Xeon E5-2620 V2 2.1 GHz/6core/1.5+15Mb/80W/7.2 GT/s LGA2011	24.1	23312	967.3
  171	131	CPU Intel Xeon E5-2609 V4 1.7 GHz/8core/2+20Mb/85W/6. 4 GT/s LGA2011-3	22.6	21914	969.65
  172	35	CPU Intel Xeon E5-2630 V4 2.2 GHz/10core/+25Mb/85W/8 GT/s LGA2011-3	45.3	45575	1006.07
  173	207	CPU Intel Core i5-4590 3.3 GHz/4core/SVGA HD Graphics 4600/1+6Mb/84W/5 GT/s LGA1150	13.3	13638	1025.41
  174	85	CPU Intel Xeon E5-1630 V4 3.7 GHz/4core/1+10Mb/140W LGA2011-3	29	30364	1047.03
  175	54	CPU Intel Xeon E5-2630 V3 2.4 GHz/8core/2+20Mb/85W/8 GT/s LGA2011-3	38.8	40971	1055.95
  176	42	CPU Intel Xeon E5-1650 V4 3.6 GHz/6core/1.5+15Mb/140W/5 GT/s LGA2011-3	41. 9	44876	1071.03
  177	18	CPU Intel Core i7-6900K 3.2 GHz/8core/2+20Mb/140W LGA2011-3	57.9	63467	1096.15
  178	27	CPU Intel Core i7-5960X 3.0 GHz/8core/2+20Mb/140W/5 GT/s LGA2011-3	50.3	57406	1141.27
  179	165	CPU Intel Xeon E5-2609 V3 1.9 GHz/6core/1.5+15Mb/85W/6.4 GT/s LGA2011-3	17.8	20690	1162.36
  180	296	CPU Intel Pentium D 945 3.4GHz/2core/ 4Mb/95W/ 800MHz LGA775	2.7	3206	1187.41
  181	88	CPU Intel Xeon E3-1281 V3 3.7 GHz/4core/1+8Mb/82W/5 GT/s LGA1150	28.4	35610	1253. 87
  182	67	CPU Intel Xeon E3-1280 V6 3.9 GHz/4core/1+8Mb/72W/8 GT/s LGA1151	33.9	43477	1282.51
  183	28	CPU Intel Xeon E5-2640 V4 2.4 GHz/10core/+25Mb/90W/8 GT/s LGA2011-3	48.5	62651	1291.77
  184	75	CPU Intel Xeon E3-1280 V5 3.7 GHz/4core/1+8Mb/80W/8 GT/s LGA1151	31.7	43128	1360.5
  185	29	CPU Intel Xeon E5-2650 V3 2.3 GHz/10core/2.5+25Mb/105W/9.6GT/s LGA2011-3	46.9	66032	1407.93
  186	47	CPU Intel Xeon E5-2640 V3 2.6 GHz/8core/2+20Mb/90W/8 GT/s LGA2011-3	41.5	58922	1419.81
  187	130	CPU Intel Xeon E5-2623 V4 2. 6 GHz/4core/1+10Mb/85W/8 GT/s LGA2011-3	22.8	32696	1434.04
  188	20	CPU Intel Xeon E5-2650 V4 2.2 GHz/12core/3+30Mb/105W/9.6 GT/s LGA2011-3	53.2	80427	1511.79
  189	14	CPU Intel Xeon E5-2680 V3 2.5 GHz/12core/3+30Mb/120W/9.6 GT/s LGA2011-3	58.4	88994	1523.87
  190	16	CPU Intel Xeon E5-2660 V4 2.0 GHz/14core/3+35Mb/105W/9.6 GT/s LGA2011-3	58.2		1603.21
  191	19	CPU Intel Xeon E5-2670 V3 2.3 GHz/12core/3+30Mb/120W/9.6 GT/s LGA2011-3	53.7	86605	1612.76
  192	10	CPU Intel Xeon E5-2680 V4 2.4 GHz/14core/3+35Mb/120W/9. 6 GT/s LGA2011-3	67.9	116852	1720.94
  193	11	CPU Intel Core i7-6950X 3.0 GHz/10core/2+25Mb/140W LGA2011-3	63.1	111607	1768.73
  194	12	CPU Intel Xeon E5-2690 V3 2.6 GHz/12core/3+30Mb/135W/9.6 GT/s LGA2011-3	61.8	118950	1924.76
  195	6	CPU Intel Xeon E5-2690 V4 2.6 GHz/14core/3+35Mb/135W/9.6 GT/s LGA2011-3	72.6	146108	2012.51
  196	46	CPU Intel Xeon E5-2643 V4 3.4 GHz/6core/1.5+20Mb/135W/9.6 GT/s LGA2011-3	41.5	104904	2527.81
  197	23	CPU Intel Xeon E5-2667 V4 3.2 GHz/8core/2+25Mb/135W/9.6 GT/s LGA2011-3	52. 2	143602	2751

  Graphics: fast, slow and integrated

  Part 24: Intel HD Graphics of the third and fourth generations

  an article that studied Celeron, Pentium and Core i3 was published more than a year ago, so it was limited to Sandy Bridge and Ivy Bridge. From the point of view of a potential buyer, of course, this situation is wrong. After all, the integrated graphics core in a top desktop processor is usually used by those for whom its characteristics are not important, so, by and large, the HDG 2500 is enough. If not enough, then a discrete video card is usually purchased, especially since owners of computers on Core i7 or Core i5 can easily afford not to save on the latter. Yes, and in older models of laptops, manufacturers often put a discreet on the principle of «so that it is.» Let it often turn out to be a GPU, comparable in performance to the built-in one, but it is not always possible to fight off such a “concern”.

  But in the budget segment, everything is completely different. Of course, Pentium (not to mention Core i3) can be used to assemble a good gaming computer. And if we limit ourselves to a single-user mode, then it’s not even “not bad”, but good (as we have already seen). However, with serious performance requirements, you usually have to purchase expensive video cards without saving on other systems, so here you can not save too much on the processor (especially since, as we have already written more than once, at the moment all processors in the consumer segment are very inexpensive ). Who needs the cheapest models? Mostly for those who have to save every dollar (and even more often — a ruble or hryvnia), so buying a decent discrete video card is not even considered (or is being considered, but somewhere in the future). «Indecent» now, as has been shown more than once, there is no point in acquiring it at all — thrown out money, which still will not allow you to get a qualitative advantage over the use of integrated graphics. But in this case, the characteristics of the latter may begin to be of decisive importance — simply because in interactive applications (which include games), quantitative characteristics translate into quite qualitative differences. In other words, there is not much difference — in how many minutes it will eventually take to import into the database or process a large number of images: of course, 15 minutes is better than 30, but in the end the work will be done (even if you have to drink an extra cup of coffee or look for another some occupation). At the same time, 15 (and even 20-25) and 30 frames per second in the game are already qualitative differences: in the second case, the game can be played with the selected settings, but not yet in the first case. In general, the question is fundamental. So the answer to it is interesting to many. Today we will look for him.

  Testing: goals and objectives, configurations, methodology

  This relatively large section will be common and the same for all articles: unfortunately, it is not enough for all people to explain something once 🙂 Moreover, not all readers will carefully study all the articles in the cycle — the probability of “starting from the middle” or simply limiting ourselves to one or two materials is extremely high, of which we are fully aware. Therefore, we immediately apologize to those who are against the constant repetition of the same truths. Which, however, as you know, is the mother of learning 🙂

  So, first and foremost, it should be taken into account that within the framework of this testing, we are not dealing exclusively with components — we are testing systems that consist of them. Separately, the processors are tested within the framework of the «main line» articles. Always in a fixed configuration — with a powerful video card, a large amount of RAM, etc. We have on our website and directly test video cards in gaming applications, updated monthly. As part of i3D-Speed, all video cards (from a simple budget card to a multi-GPU) are tested on a powerful configuration, chosen based on the sufficiency for the graphics component of any power. That is, we believe that from the point of view of traditional «component» testing, these two lines of articles are quite enough.

  But for the practical use of the results obtained within their framework, a certain link is needed. The fact is that applications whose performance does not depend on the central processor do not exist in nature. There are, of course, cases when it is limited to other components, but this very often happens at different levels for different processors. Gaming and similar applications significantly depend on the performance of the GPU, but they also put a lot of load on the CPU. If the task is too “light” for graphics, only the processor begins to determine everything. If it is “heavy”, then the influence of the processor, on the contrary, becomes minimal, and sometimes it can even be ignored. Between these extreme cases, both components are important, and the degree of their importance can be reversed. A priori in an unknown way. That is, from the fact that one processor is faster than another using a powerful video card, it does not follow that the ratio will be maintained if it is replaced with a budget one. More precisely, in some modes it will remain, in some it will change, in some everything will simply be the same. A similar problem is inherent in video cards — the level of «sufficiency» of the CPU varies depending on the GPU and its mode of operation.

  It would seem that it would be enough to simply test all the «processor + video» bundles. The solution is obvious and correct in theory, but practically impracticable in practice, since the amount of work is growing exponentially. In other words, 40 video cards on one system — 40 test configurations. 40 processors with one video card — also 40 configurations. And if you combine this, you get 1600 test configurations. Although, of course, if all this work can be done, truly invaluable results will be obtained. But by the time they are received, they will no longer be needed by anyone, since they will become obsolete (looking ahead — even the “simplified” method we have chosen allows us to test no more than a dozen configurations in a working week, so 1600 is a task for three years when using one stand).

  But you can also approach it from the other side: do not try to find exact answers to all questions, but confine yourself to qualitative assessments. At least for some of the processors, you can try to «feel» the lower level of performance. Which is integrated graphics, since recently it has become an integral part of most modern processors. And there are younger models of discrete adapters that are at least as good. But it is many times simpler and slower than top-end solutions — on the graphics market, the spread of characteristics is still greater than on the processor one. With this choice of equipment, we can significantly reduce the list of test configurations and modes. Indeed, the most relevant results will be for buyers of budget computers, since with the price of a system unit of 1000 dollars, you can give 10% of this amount for a slightly more powerful video card than the lower level, and not mess with the same integrated video. Simply — to be. So processors of the middle class and higher often do not need to be tested with weak video. Sometimes, of course, we will also do this — in order to have the necessary guidelines, but only sometimes. In addition, systems of this class do not require tests in some outstanding modes, such as 2560 x 1600 with older variations on the theme of full-screen anti-aliasing 🙂 In a word, the work can be significantly simplified.

  The fact that 90% of applications of the standard processor technique do not depend on video performance at all reduces the amount of work. In the previous series, we used all the programs, so four parts of it are quite sufficient proof of this fact. For whom it’s still not enough — there’s nothing we can do about it 🙂 Be that as it may, GPGPU is still nothing more than a curious experiment, and all work in this direction shows that it is generally special for systems with weak GPUs does not differ in relevance: powerful video cards on “good” tasks are really capable of speeding up something, but when trying to squeeze something worthwhile out of an entry-level discrete, very often all the steam goes into the whistle — the complication of algorithms and unnecessary data transfers “eat up” all the potential gain. From which, however, one should not conclude that we will pass by some curious and popular application that can actively use GPU resources. Of course, we will not go through and add it to this experimental technique. Only here while the main problem that anything similar does not come across. More precisely, there are already “curious” programs, but are popular they still do not become for one reason or another. The same video transcoding, around which a lot of copies were broken, in fact, few people need it regularly, and the quality of work developed by enthusiasts of programs leaves much to be desired (this is still very mildly speaking). And (here it is a grimace of fate) is most quickly performed using specialized hardware blocks available in integrated Intel GPUs, and not at all on general-purpose pipelines.

  Thus, we have not so many programs left that it makes sense to «drive» on systems with poor graphics. In fact, the «standard» method is literally simplified to five groups, three of which are experimental in it. These are: Interactive work in three-dimensional packages No changes Mathematical and engineering calculations MAPLE and MATLAB are thrown out, since nothing is displayed on the screen, but the remaining three applications are interesting to readers, judging by the reviews (it is clear that it is hardly advisable to save so much on the workplace, but suddenly you have to work on a weak computer). In fact, it turns out that the composition of these two groups is the same as a result, but in the previous case, the “graphic” score of the corresponding test is taken into account, and in this case, the “processor” one: as testing practice has shown, in fact, both of them depend on both the processor and video cards, which is what we needGames No changesGames with low resolution and quality settings Within the framework of the «main» method, this group is practically not used in any way and does not affect the overall score, but it was made just for systems with weak graphics. Primarily mobile, but not that different from what we’re testing in this series High-definition video playback Needs no special comment

  Since we do not have many groups, and all of them are quite specific, we will not give a general assessment. We are primarily interested in results. Which, as usual, will be fully compatible with those obtained on the configurations of the main testing line, since we already know for sure that video cards do not affect other applications in any way. So, if you wish, you can simply replace the corresponding piece in the “large” table, since we do not hide them in any way. However, it should be borne in mind that the scores of this test are in no way compatible with the main line: here we take a system with Celeron G540 and Radeon HD 6450 512 MB GDDR3 as a scale unit, so for independent fraud, you should download a table in Microsoft Excel format, in which all the results are given both in the converted into points and in the «natural» form. 953 55 54 55 54

  Desktop Celerons based on the Haswell microarchitecture have been announced recently and have not yet reached our hands, and the Bay Trail is a completely different story: only BGA and W make these models maximally competitors of CULV processors, but by no means «standard desktop» platforms. But Pentium and Core i3 of various modifications are massively available for both LGA1155 and the new LGA1150. Accordingly, three pairs of processors will take part in our testing — two Pentiums and four Core i3s. With Pentium, everything is simple — we took two processors with the same clock speed of computing cores: the old G2140 and the new G3430. Please note that the graphics core of lower-end models is still called HD Graphics, although this is already the fourth GPU with this name, and it differs from the previous two not only architecturally, but also the number of pipelines has increased from 6 to 10. That is, the difference with Ivy Bridge will be sure, but there is nothing to compare with the Pentium and Celeron on Sandy Bridge that are still on sale — the functionality is very different, which we already noted a little over a year ago.

  There is no naming confusion in the Core i3 family. Moreover, there is more order in general — earlier the company offered both processors with the HDG 2500 core (the most popular in desktop Ivy Bridge), and several modifications with the HDG 4000. This ensured the equality of selling prices, but the frequency of computing cores was always higher (with this condition) for models with a lower graphics core. The new generation was divided into two families. The heirs of the old Core i3 are models of the 41×0 line, similar to them in terms of frequencies and cache memory capacity and equipped with HDG 4440. The more expensive processors of the 43×0 line have become a relatively new product, where not only the oldest among the “socket” GPU HDG 4600 processors, but and all 4 MiB of L3 cache is used: as in the first generation Core i3 or mobile dual-core Core i7. In general, the positioning of new processors has become simpler and more logical: we pay more — we get more. In all respects. There are also intersections in clock frequency with the previous generation, which gave us two equal pairs 3245-4130 and 3250-4330.

  Another guest from another world is a video card based on the GeForce GT 630. We already tested something with this name a year ago, but exactly what with the name: the old products were based on the GF108, and the new ones use the GK208 chip. NVIDIA itself claims that this is a new development, but in fact the GPU is very similar to the cropped GK107 (previously used in the GT 640 and higher). Moreover, cropped in a software way — both have the same area and partially coinciding wiring. Why partially? Because the GK208 lacks one memory channel, and the bus interface is only PCIe x8, not x16. Thus, it is obvious that at comparable frequencies, the GT 630 is not a competitor to the old GT 640, despite the same number of GPUs. But compared to the old GT 630 DDR3, everything should not be so bad: the «narrow» memory bus is partially compensated by its higher clock frequency (1800 MHz against the official 1600 MHz, which in real products often dried up to 1400 MHz), and the arithmetic capabilities of the chip are much higher — at the level of GT 640. Another question is whether such a level is needed in a modern computer or is it better to use integrated video? 🙂 On the other hand, what is important, the cards based on the GK208 are compact and completely equipped with passive cooling (because the GPU does not heat up very well), and at the price they can compete with the GT 610 / 620, which differ by absolutely no performance. In general, these solutions have a certain niche — at least an upgrade of old compact systems. Well, we will determine the exact performance level using a card from ASUS with 2 GB DDR3 (we did not test the modification with 1 GB because there is nothing to do — different volumes in video cards of this level will not affect in any way), working together with Core i3-4330 (so that certainly the processor did not interfere).

  Interactive work in 3D packages

  As we already wrote, in driver version 9.18.10.3257, Intel programmers fixed another batch of errors, which led to a curious effect: even Pentium on Ivy Bridge (adding 20% ​​to last year’s results) is already reaching the level any AMD APUs (with the possible exception of Kaveri, but these models are just starting to hit the retail chains). Moreover, this is the level of junior NVIDIA discrete gaming chips, even paired with a faster processor. In general, you should no longer be afraid of integrated Intel graphics. Especially after the release of Haswell, this is an even higher level of performance. Moreover, as you can see, installing a lower game discrete (which was almost mandatory for such programs in the days of Sandy Bridge) significantly reduces performance, i.e. it’s better not to do this anymore.

  Math and Engineering

  HD Graphics didn’t get in the way here before as the results were mostly dependent on single-threaded CPU performance, which put Intel devices in a winning position, and now the situation has only worsened. But, by the way, pay attention — a discrete graphics card can improve the results. Simply because it does not claim either the processor’s cache memory or the thermal package. However, the gain is extremely small, which, together with a decrease in the «graphics» score, does not allow changing the conclusion — if you buy a discrete video card for professional programs, then certainly not a junior gaming one.

  Aliens vs. Predator

  As expected, the 3rd Gen HDG and the HDG 2500 are identical, something we’ll see more than once, so we won’t dwell on this result in the future. The 4400 is only slightly faster than the 4000, which is forgivable — one of the younger decisions against the once older one. The HDG 4600 almost reaches the A6 performance — a noticeable step forward because, as we have already said, the HDG 4000 was only enough to fight the A4. And the difference between the two HDGs is even greater. Although in practice in this mode everything breaks down that even the A8-6600K (faster than the GT 630, by the way) is still not enough to get a comfortable frame rate. Therefore, the settings will have to be lowered.

  At a minimum, of course, everything flies. In addition to the younger graphics configuration Ivy Bridge — even in this mode, it was barely enough to cross the border at 30 FPS. So I’m glad that the new graphics at least don’t have such problems. And even only Pentium is already behind the discrete GT 630 level, and even then a little, and installing such cards in a computer based on any new Core i3 is definitely a bad idea. Well, APU is ahead by a wide margin from the others. The result was not unexpected, although there are hopes for at least an approximate parity of the older Core i3 with at least much cheaper A6. We have seen the once lower results even in very old A8, of course, however, AMD engineers and programmers have not been idle for the last year either 🙂

  Batman: Arkham Asylum GOTY Edition

  The high-quality (in our testing) mode of this game «surrendered» to integrated graphics from Intel after the introduction of the HDG 4000, and newer GPUs from the company, of course, are even faster. And even the Pentium was not enough just a little to make it to 30 FPS. An achievement that, however, fades against the background of the fact that even the old A4-5300 or the very ancient A6-3500 is still faster — AMD set the bar high, you can’t say anything. Actually, there is nothing surprising in the fact that the company’s APUs are already pushing out even the lower discreet from the market. And Intel, despite rapid progress, the pipe is lower and the smoke is thinner 🙂 However, it is also already clear that it makes no sense to install solutions of the GT 630 class (especially lower) into systems based on its new processors — there will be no fundamental increase in performance.

  With a low picture quality and an old graphics engine, you get a comparison of processors for the most part. With slight variations: after all, the HDG 2500 (and its relatives of low-end families) are too weak a solution, and the use of a discrete one prevents the processor component from working at full capacity. But in general, it was possible to play in this mode even on the Celeron G555, and progress since its appearance allows you not to limit yourself so much.

  Crysis: Warhead x64

  An example of the opposite situation — so far no integrated graphics solution can handle this game with the selected settings. Moreover, as we can see, despite the steady increase in productivity, next year is unlikely to change much. Which is not surprising, since even the discrete Radeon HD 7750 DDR5 is enough for this with almost no speed margin. But if we evaluate not only the absolute results themselves, but the dynamics of their growth, the assessment of the situation changes somewhat. As you can see, modern Pentiums have already reached a level that only a year ago was available only to some modifications of the Core i3. And the older representatives of the latest in terms of performance of the graphics core are now at the level of APUs of the A6 family or … Discrete video cards of not so long ago, such as the Radeon HD 6670 DDR3. Or quite modern GeForce GT 630. That is, the boundary between older (and even not the oldest) models of integrated GPUs and younger discrete ones is increasingly blurred.

  It is worth lowering the picture quality to the level of ten-year-old games, as it immediately turns out that anything is enough, which is quite correlated with “worldly wisdom”. But it also makes such modes not too revealing, of course, however, as we have already said more than once, they were chosen at one time in an attempt to force graphics of lower classes to provide acceptable performance — for example, integrated into three-year-old low-power Celerons. However, some interesting information can be “squeezed” out of them even now. In particular, the progress of Intel drivers is quite visible — just over a year ago, the Pentium G2120 here produced less than 50 frames per second, and with the new drivers, the G2140 became one and a half times faster. However, this is not enough to keep up with at least the cheap AMD A6, but the new Pentiums in games with simple graphics (either initially simple or simplified settings) can already “wrestle” with the A8. And, again, the only plus of a weak discrete is that it does not prevent the processor part from giving all the best. Although the effect of this when using inexpensive video cards, as we see, cannot be called significant.

  F1 2010

  Although the game will soon be four years old, it is still a tough nut to crack for integrated graphics. But in a slightly different way than Crysis — if the whole load fell on the GPU, then the processor performance is already important here, and it is desirable that the latter support more than two computation threads. As a result, most of the lower-end solutions keep at 12.5 FPS thanks to the engine itself — it tries to “not fall” as much as possible, further simplifying the picture. Here HDG 4000 and above, as well as integrated Radeon HD work «honestly», but still too slowly. And no wonder — as we already know, more or less only top-end A10s can cope with this mode. And even better, as before, use discrete. Preferably at least Radeon HD 7730 DDR5 or higher.

  In light mode, even with weak graphics, the shortcomings of dual-threaded processors are already visible. However, once again this is most noticeable when using AMD processors, but for Intel, the difference between Pentium and Core i3 is small (and the new Pentium may overtake the old i3). Therefore, the minimum should be considered something of class A8. Or buy a discrete graphics card — the specificity of the EGO engines (used in the entire Formula One series) is such that even a decrease in the graphics quality does not make it useless.

  Far Cry 2

  Far Cry 2 is even older, so only Intel and AMD A4/A6 processors can’t cope with the task even in high-quality mode. In general, the qualitative difference between Intel HD Graphics and an APU or a lower-end discrete — as you can see, it still persists, despite a very noticeable increase in performance in the new generation of GPUs.

  But for the light mode, only Sandy Bridge was missing, and in the case of more modern devices, we get almost a test of the performance of the processors themselves. With quite predictable results.

  Metro 2033

  In fact, another stress test for integrated graphics — getting something more or less acceptable from it will not work for a long time. But for evaluating the performance itself, the GPU fits well. However, there is almost nothing new for us here, with the exception of perhaps the most noticeable difference between the two generations of Intel IGPs — Haswell really became a big step forward, allowing the company to almost catch up with the integrated Radeon. More precisely, the HDG 4000 could already compete with the A4, which, however, did not pull on the achievement — too low a level for relatively expensive solutions. But the approximate parity with the A8 is all right. In theory, of course — in practice, as we already know, even $100 discs are too few.

  Actually, even with the low quality mode (it’s not that low in this game, it should be noted that the minimum supported resolution of 1024 x 768 was often used in practice only recently), the integrated graphics «learned» to cope not so long ago. And not every one — the A6 based on Llano was the first to break the border, and the transition to Trinity turned out to be even a step backwards in this family (because the game can fully use multi-core processors), but, in general, there are enough of them. And there are no slower solutions. However, we again observe that within the new Intel platform, even Pentium is “enough”, but most of the products for the previous one could not cope due to the weakness of the mass HDG 2500. That is, we actually have a transition from quantity to quality — something that “couldn’t” a year ago many Core i5, today «may» Pentium. Or any Core i3, not individual models of this family. Well, that’s good too.

  Summary results

  What do we have in the bottom line? If you remember that 100 points is a Radeon HD 6450 paired with a Celeron, then a lot. Indeed, mass graphics for LGA1155 (and this is the HDG 2500 and its analogue in Celeron / Pentium or generally weak, even functionally IGP Sandy Bridge) failed to even reach this level. The new Pentiums surpass it, i.e. the integrated GPU easily overtakes discrete products such as the aforementioned Radeon HD 6450 or GeForce GT 610/620. It is clear that all of them can only be called gaming solutions out of politeness, but they still exist and are still on sale (not to mention older video cards of a comparable level that continue to be used by many economical computer users). In addition, A4 for the FM1 platform was also left behind — also a base level, of course, and even for an outdated platform two years ago, but a couple of years ago, few believed that Intel would even be able to catch up with AMD in the foreseeable future: Sandy Bridge graphics in any variant could not be compared with desktop APUs of all modifications.

  Core i3 at first glance «grew up» weaker — HDG 4400 is faster than HDG 4000 only by 20%, not one and a half times. Which is easily explained — if in the budget segment the number of pipelines increased from 6 to 10, then only from 16 to 20 «on the floor above». and the 4400 is the lower end of the new desktop Cores: most already use the HDG 4600, which has slightly better performance. In fact, we can even talk about the transition of quantity into quality — just a year ago, only the HDG 4000 (the very rare variant) could provide frame rates in games at the level of AMD APUs of the A4 line, but now parity has already formed with faster A6. Naturally, this does not look like a victory in any way — after all, in terms of price, even the A8 keep at the Pentium level, and the Core i3 are faster, but also noticeably more expensive processors, but the fact of gradually equalizing positions takes place. However, the release of APUs based on Kaveri may well be able to restore the status quo, but the mass distribution of these devices (and their promotion to the lower segments of the AMD range) will have to wait. And the replacement of Trinity with Richland, as we already wrote, was only a cosmetic update. Not at all like switching from Ivy Bridge to Haswell.

  Of course, the «building of integrated muscles» in the products of both vendors more and more narrows the potential scope of junior discrete solutions. The new GT 630 is only marginally faster than the old one (memory is the bottleneck) and still lags behind the A8/A10. And the gap from the junior solutions from AMD and Intel has already decreased so much that the purchase of a discrete video adapter of this level has ceased to be a justified undertaking at all — the performance gain does not compensate for the extra costs and other shortcomings of the approach. In general, the only thing that video cards of this segment can claim is the modernization of old computers. And here, in most cases, a more attractive solution will either be buying a faster discrete, or simply replacing the platform.

  Well, you can gradually stop paying attention to the modes of minimum settings — all modern solutions already cope with them. In any case, desktop — surrogate systems still cannot boast of comfortable results even with the simplification of graphics to the level of a decade ago.

  OpenCL

  Despite the active talk about heterogeneous computing, their scope remains very limited. Especially if we talk about those areas that are applicable to integrated graphics — the use of discrete GPUs for some «heavy» calculations in the HPC field began several years ago, but this has little relevance to the mass market. And the main problem for the latter was, it seems to us, that OpenCL is not at all as “open” as it was declared. In fact, programmers are forced to take into account the peculiarities of the implementation of specifications by all three vendors, i.e. work at too low a level. WinZip turned out to be a typical example of the immaturity of the technology at one time — behind the victorious reports about the release of an application of at least some general purpose with OpenCL support, not everyone noticed that we were talking about support only for the AMD implementation, but not for Intel and NVIDIA.

  Curiously, these features are still showing up even in synthetic benchmarks, many of which simply execute different code branches on different solutions. In particular, such is the Basemark CL, which we started using some time ago as part of the tests of this line. What this leads to in practice is clearly seen in our study of the programs themselves: this utility is clearly not indifferent to AMD GPUs. And if you also remember that not so long ago, Intel processors executed OCL code only on the main cores, but without using the GPU, it becomes clear why this particular program became AMD’s favorite benchmark, the use of which was recommended to all testers. Recently, however, they stopped recommending it. Let’s try to understand why, considering, of course, that Basemark CL for cross-platform comparison must be used very carefully.

  In the diagram, we have collected the results of all the processors tested in this program, which painted an extremely interesting picture. First, as we can see, the HDG 2500 or the “numberless” relative of this GPU provide performance only at the level of junior mobile solutions. It is clear why — the code is well parallelized, so six pipelines are six pipelines, at least in the CULV Celeron, at least in the desktop Core i3. But the Pentium on Haswell is already much faster. However, it is still not possible to consider it as a serious OpenCL accelerator: up to A6, it still does not reach processors with HDG 4000 (again, it doesn’t matter: mobile or desktop). But certain preferences when using OpenCL can also be obtained with its help — at least more than the buyer of any processors based on the AMD Kabini core will receive. But the HDG 4400 is a much more attractive option: as you can see, only the new generation Core i3 turned out to be equal to the top-end Core i7 of the previous one! And compared to competing products, this is not so bad — the level of some A8. It is clear that they are cheaper, but the difference in price with the younger Core i3 is still much less than with the older Core i7 🙂 And the HDG 4600 is already the A10 level. Moreover, it is easy to see that all economical buyers, and not only those who choose AMD products, can get big benefits from the implementation of OpenCL: the difference between i3 and i7 is less than 10%. In general, the victorious reports only spoil the results of Kaveri — AMD managed to jump above its head once again. But there are not enough of these APUs yet, unlike the Core i3 lying on every corner. In addition, cheaper and more productive on the classic x86 code, which is extremely important in the current state of affairs with the implementation of OpenCL (a processor that is faster in a large number of programs and slower in a small number of programs looks more attractive than one that wins only in an exotic selected environment).

  You don’t have to comment on the results of GT 630 — as it has been noted more than once, this benchmark of NVIDIA’s solution does not like it (moreover, OpenCL 1.1 code is used in this case, not 1.2). On the other hand, no one is immune from the repetition of such a situation in real programs. Well, in this case, as we can see, the lower discrete can easily lag behind even inexpensive integrated graphics. What is an additional nail in its coffin 🙂

  Total

  If, when choosing a high-level processor (and even assuming the use of a discrete video card), no one managed to find Haswell’s special advantages over Ivy Bridge, then in the budget segment and when using an integrated graphs, the situation is the opposite: there is no point in buying “old” processors. Perhaps to upgrade the system to Sandy Bridge while maintaining the motherboard, but here it’s better to just buy a video card — cheaper and more efficient. And the new system is exclusively on the LGA1150. In that case, of course, if you choose from Intel solutions — as you can see, the lag behind AMD APUs has greatly decreased, but has not completely disappeared. Thus, if you want to save money and focus primarily on the performance of the graphics core, the FM2/FM2+ platform is still a good choice: the same A8-6600K is cheaper than any Core i3, and the A8-5600K can compete in price with Pentium. Naturally, in this case, you should not forget that this saving is not at all free — the processor part is very different, which is often very important (at least in this segment), and in the case of a subsequent purchase of a discrete video card, an additional payment for a «good» the integrated GPU will disappear entirely. In addition, the “appetite” of AMD APUs is somewhat higher than typical for dual-core Intel processors. In general, they are not direct competitors, but, we repeat, if the performance of integrated graphics is in the first place, then it is better to still pay attention to AMD developments — the new generation of devices from Intel has reduced the gap in this matter, but far from zero, even if we ignore the price difference.

  Well, in a global sense, progress certainly makes us happy. Especially if we talk about the basic level of performance. You can, of course, once again scold Intel for some confusion — after all, this is the fourth graphics core with the faceless name «HD Graphics», but more importantly, its performance has increased by the traditional one and a half times. This does not make HDG a gaming solution, but the very fact of “raising the bar” is already a good signal for programmers. Yes, even higher order has been added — after all, up to and including Ivy Bridge, the “main” level of Intel graphics in the desktop segment coincided with the “basic” one: the most massive GPU was the HDG 2500. Now, Core i3 is distinguished from Pentium not only by Hyper-Threading support, but also more powerful graphics: at least HDG 4400, and already this video core is better than any Ivy Bridge GPU. Even if not one and a half times already, but this (and higher) level of graphics capabilities is now available to every buyer — there is no need to chase after special processor models. Which, again, allows us to count on its more complete utilization by programmers.

  And, of course, such an increase in the graphics capabilities of junior processors is another nail in the coffin of budget discrete video cards. While there is still a performance advantage even in the «$60» segment, it’s already too small to buy a standalone device rather than using a «free» IGP.