Russia Defence Forum

Would you like to react to this message? Create an account in a few clicks or log in to continue.

Military Forum for Russian and Global Defence Issues


+80
Isos
LMFS
kvs
KomissarBojanchev
eridan
Pierre Sprey
d_taddei2
RTN
[ F l a n k e d ]
AlfaT8
zg18
JohninMK
Swede55
onwiththewar
Hole
marcellogo
havok
Mindstorm
magnumcromagnon
dino00
archangelski
Manov
tomazy
rambo54
JackRed
The-thing-next-door
Tsavo Lion
Peŕrier
YG_AJ
GRIM 44
BKP
SeigSoloyvov
Dr.Snufflebug
TheArmenian
Neutrality
medo
Azi
MC-21
wilhelm
KiloGolf
Stealthflanker
Luq man
Cyberspec
Tingsay
thegopnik
Nasr Hosein
flamming_python
AMCXXL
ZoA
iwanz
par far
T-47
GarryB
Cheetah
miketheterrible
OminousSpudd
Singular_Transform
chicken
ATLASCUB
berhoum
Vann7
Big_Gazza
hoom
Viktor
HM1199
Cyrus the great
tanino
coolieno99
franco
jaguar_br
Svyatoslavich
mack8
yavar
Benya
George1
Austin
higurashihougi
Rmf
Kimppis
Project Canada
84 posters

    PAK-FA, T-50: News #4

    avatar
    Vann7


    Posts : 5385
    Points : 5485
    Join date : 2012-05-16

    PAK-FA, T-50: News #4 - Page 5 Empty Re: PAK-FA, T-50: News #4

    Post  Vann7 Fri Apr 28, 2017 8:24 am

    tanino wrote:Thanks for the photo of the cockpit of the Super Hornet.

    At Suckoy Team a single screen is not safety in case of malfunction, much better to have two singles pushed together in the event of default of any of them, fighting performance would not be affected. In Russia (and earlier in the Soviet Union there have been numerous studies (some still stumbled) on ergonomics combat workstation (advanced tech apart)

    Regarding the discussion I agree that the data represented on the display are not a video game created by corporate Perception (example), beautiful, modern and effect, but do not correspond to actual use in flight and in combat.
    But though: is there a point where for obvious reasons, russia must catch up: the Visual information of its pilots.

    I think the basic point is unique: no one likes to be represented on the display old-analog dials instead of three-dimensional maps.

    No problem..

    I read a year ago ,one Indian pilot comments about his view of more than ~5 combat jets
    that he tested or allowed to be passenger , and he was able to see Russian planes and Europeans ,ones and American ones . he was testing plane performance ,and as i said ,
    it was quite interesting how for him ,was important to have modern digital displays and
    also the helmet too. He told about american planes (i think was F-15) that was hard to read
    the panels with direct sun light and a bit difficult to have instant situational awareness. and he praised F/18 superhornets panels. So also a surprise . generally his favorite plane was the
    rafale very modern avionics ,electronics and computers. he did not tested however Pak-fa or
    F-22.
    avatar
    HM1199


    Posts : 49
    Points : 51
    Join date : 2016-07-03

    PAK-FA, T-50: News #4 - Page 5 Empty Re: PAK-FA, T-50: News #4

    Post  HM1199 Fri Apr 28, 2017 1:21 pm

    hmm , i think there are people here with knowlege about computer stuff ,i don't know much in that but i want an explannation about something , in the ''integrated avionics" article they said that the transfer rate of the PAK FA's systems is 8Gb/s. Does that mean this is the speed of the databus?
    i also found a similar thing here : http://www.rpkb.ru/lines-of-business/electronic_direction/on-board-computers/family-btsvm-baget-53/
    they said that the communication mechanism is done through the bus VME32, 2 full duplex channel Fibre Channel interface for information exchange. Linear data rate - 1062.5 Mbit / s , would that mean that the databus in the baguette 53-31M computer ?
    As a reference , in a file , they said that the HSDB or high speed databus of f22 has a speed of 50Mb/sec.So i'm not really understanding all of this , maybe i'm doing a wrong analogy.
    miketheterrible
    miketheterrible


    Posts : 7383
    Points : 7341
    Join date : 2016-11-06

    PAK-FA, T-50: News #4 - Page 5 Empty Re: PAK-FA, T-50: News #4

    Post  miketheterrible Fri Apr 28, 2017 5:52 pm

    You are more or less close to correct but it wont be the baguette computer
    GarryB
    GarryB


    Posts : 40376
    Points : 40876
    Join date : 2010-03-30
    Location : New Zealand

    PAK-FA, T-50: News #4 - Page 5 Empty Re: PAK-FA, T-50: News #4

    Post  GarryB Sun Apr 30, 2017 12:20 pm

    You have to keep in mind a couple of things...

    first of all the F-22 is not new and 50Mb/s is actually very fast because it has a radar sensor but no thermal optics so it will be transmitting digital data but not video streams or anything data heavy.

    In comparison the PAK FA will be transmitting data to the various networks it is operating within to share and receive target data and information.

    The F-22 will keep its emissions to a minimum so it wont be spotted.

    The PAK FA will be using information to hunt down threats including stealth fighters and stealth bombers and cruise missiles.
    avatar
    hoom


    Posts : 2352
    Points : 2340
    Join date : 2016-05-06

    PAK-FA, T-50: News #4 - Page 5 Empty Re: PAK-FA, T-50: News #4

    Post  hoom Sun Apr 30, 2017 2:27 pm

    F-22 has really old processors.
    Its like 20 GFLOPS total after upgrade (same as original Xbox)

    F-35 has newer processors, seen a number more like 200 GFLOPS which is much better but still not that impressive, a bit less than an Xbox 360.
    Saw another number of 1 TFLOPS which would be much better but below XboxOne (1.3 TFLOPS)

    There is a rumor the new T-50 processor setup is this:
    The architecture of processor cores: "Elbrus" with an extended set of commands
    The total number of processor cores OBC: 24 kernel
    Core clock frequency (not less than): 600 MHz
    Total capacity: up to 20,000 DMIPS (Dhrystone test), up to 150 GFLOPS for double-precision numbers, up to 300 GFLOPS for single-precision (for comparison, Intel Core2 Quad Q8300 2.5 GHz, 4 cores – 40 GFLOPS)
    So that'd be somewhat better than the low F-35 number but well behind the high.
    avatar
    HM1199


    Posts : 49
    Points : 51
    Join date : 2016-07-03

    PAK-FA, T-50: News #4 - Page 5 Empty Re: PAK-FA, T-50: News #4

    Post  HM1199 Sun Apr 30, 2017 2:57 pm

    i heared that the 150 and 300 GFlops are with the elbrus 4s , maybe it'll get better with 8s.
    Viktor
    Viktor


    Posts : 5796
    Points : 6429
    Join date : 2009-08-25
    Age : 44
    Location : Croatia

    PAK-FA, T-50: News #4 - Page 5 Empty Re: PAK-FA, T-50: News #4

    Post  Viktor Sun Apr 30, 2017 3:13 pm

    GarryB wrote:You have to keep in mind a couple of things...

    first of all the F-22 is not new and 50Mb/s is actually very fast because it has a radar sensor but no thermal optics so it will be transmitting digital data but not video streams or anything data heavy.

    In comparison the PAK FA will be transmitting data to the various networks it is operating within to share and receive target data and information.

    The F-22 will keep its emissions to a minimum so it wont be spotted.

    The PAK FA will be using information to hunt down threats including stealth fighters and stealth bombers and cruise missiles.

    Thats why they united into Russian aerospace forces Very Happy
    avatar
    hoom


    Posts : 2352
    Points : 2340
    Join date : 2016-05-06

    PAK-FA, T-50: News #4 - Page 5 Empty Re: PAK-FA, T-50: News #4

    Post  hoom Mon May 01, 2017 6:30 am

    i heared that the 150 and 300 GFlops are with the elbrus 4s
    Yes it fits 6* Elbrus-4S (25 GFLOPS per 4 core @800Mhz)
    http://mcst.ru/mikroprocessor-elbrus4s
    https://en.wikipedia.org/wiki/Elbrus-2S%2B

    Elbrus-4S doesn't sound all that great relative to modern desktop PCs & being a VLIW architecture it may be quite hard to actually obtain a significant % of theoretical performance.
    But its a big improvement over the 4.8 GFLOPS @300Mhz Elbrus-2SM.

    And then the next gen Elbrus-8S is kinda a monster with 125 GFLOPS @1.3Ghz on much more modern 28nm manufacturing process
    PAK-FA, T-50: News #4 - Page 5 Elbrus-8c
    http://www.mcst.ru/novyj-8yadernyj-mikroprocessor-elbrus-8c
    https://en.wikipedia.org/wiki/Elbrus-8S

    If they make a similar step up for the following generation...
    miketheterrible
    miketheterrible


    Posts : 7383
    Points : 7341
    Join date : 2016-11-06

    PAK-FA, T-50: News #4 - Page 5 Empty Re: PAK-FA, T-50: News #4

    Post  miketheterrible Mon May 01, 2017 7:18 am

    25 gflops under 64bit operation 50 under 32bit.

    You guys seem to also forget that this isn't specifically a basic processor. These require DSP cores which Elbrus 2C+ has and it is eluded to the fact that Elbrus 4C+ has it too for specific lines, meaning 2 or more DSP cores. DSP, digital signal processors, are ideal for radar systems and other systems that need to translate radio signals and alike. They run at roughly 10mhz but are significantly more powerful than what they do against most modern CPU's.

    As well, VLIW has little to do with what you are even talking about, since the architecture itself runs at its intended initiative, only matters when it is working under an environment and if it is supported by the OS. In this case, it may not work all that well in Windows OS compared to it working in BSD. Since planes don't use Windows OS, then you are just talking out of your ass. I work with VLIW processor Intel Itanium over where I work, so I know what I am talking about.
    avatar
    hoom


    Posts : 2352
    Points : 2340
    Join date : 2016-05-06

    PAK-FA, T-50: News #4 - Page 5 Empty Re: PAK-FA, T-50: News #4

    Post  hoom Mon May 01, 2017 4:39 pm

    As well, VLIW has little to do with what you are even talking about,
    I was going to write a bit about it depending on how avionics are programmed whether VLIW would be easy to get good performance from & the assumption that chosing Elbrus for T-50 implies that it is considered a good match but couldn't find a short way to say it  dunno

    25 gflops under 64bit operation 50 under 32bit.
    Used the DP numbers to match the quoted 150 GFLOPS DP in the quote, works out same for the SP numbers but yes I should have made that clear pale

    Overall point is contrary to Vann the T-50 will have an order of magnitude better processing power than F-22, ~on par with F-35 (depending what end of the range it actually has)
    Singular_Transform
    Singular_Transform


    Posts : 1032
    Points : 1014
    Join date : 2016-11-13

    PAK-FA, T-50: News #4 - Page 5 Empty Re: PAK-FA, T-50: News #4

    Post  Singular_Transform Mon May 01, 2017 4:50 pm

    hoom wrote:
    i heared that the 150 and 300 GFlops are with the elbrus 4s
    Yes it fits 6* Elbrus-4S (25 GFLOPS per 4 core @800Mhz)
    http://mcst.ru/mikroprocessor-elbrus4s
    https://en.wikipedia.org/wiki/Elbrus-2S%2B

    Elbrus-4S doesn't sound all that great relative to modern desktop PCs & being a VLIW architecture it may be quite hard to actually obtain a significant % of theoretical performance.
    But its a big improvement over the 4.8 GFLOPS @300Mhz Elbrus-2SM.

    And then the next gen Elbrus-8S is kinda a monster with 125 GFLOPS @1.3Ghz on much more modern 28nm manufacturing process
    PAK-FA, T-50: News #4 - Page 5 Elbrus-8c
    http://www.mcst.ru/novyj-8yadernyj-mikroprocessor-elbrus-8c
    https://en.wikipedia.org/wiki/Elbrus-8S

    If they make a similar step up for the following generation...

    Actually the VLIW architecture means that it can work close to its theoretical performance.


    The whole Elbrus CPU family has been designed for military signal processing, so every function in it is tailored to get out the most for fast fourier & co.

    avatar
    hoom


    Posts : 2352
    Points : 2340
    Join date : 2016-05-06

    PAK-FA, T-50: News #4 - Page 5 Empty Re: PAK-FA, T-50: News #4

    Post  hoom Tue May 02, 2017 2:36 am

    The whole Elbrus CPU family has been designed for military signal processing, so every function in it is tailored to get out the most for fast fourier & co.
    Is that so?!
    I had the impression they were intended for govt servers/workstations & other sensitive stuff with future potential for domestic desktops (hence ability to run X86 code) but not actually for going into military equipment -> possibility that using it on T-50 would have been a case of a Politician/Manager going for the big numbers ignoring advice...

    If military equipment was the original intent then US better get working on better military processors ASAP if they don't want to fall behind...
    Singular_Transform
    Singular_Transform


    Posts : 1032
    Points : 1014
    Join date : 2016-11-13

    PAK-FA, T-50: News #4 - Page 5 Empty Re: PAK-FA, T-50: News #4

    Post  Singular_Transform Tue May 02, 2017 8:31 pm

    hoom wrote:
    The whole Elbrus CPU family has been designed for military signal processing, so every function in it is tailored to get out the most for fast fourier & co.
    Is that so?!
    I had the impression they were intended for govt servers/workstations & other sensitive stuff with future potential for domestic desktops (hence ability to run X86 code) but not actually for going into military equipment -> possibility that using it on T-50 would have been a case of a Politician/Manager going for the big numbers ignoring advice...

    If military equipment was the original intent then US better get working on better military processors ASAP if they don't want to fall behind...


    So, it is a bit more complicated than that.

    The Elbrus design started around 1980, the main target was to develop efficient and fast cpu for research centers and high demand military puproses, like phased array radars ,sonars and so on.

    To accomlish it they created a CPU deisgn that provided functions in 85 that was implemented by the intel in 93.

    So, this is the VLIW design.


    The current out of order execution require comlicated hardware, actualy 75% of the CPU die is for the predictor circuits to discover the paralel execution opportunities.

    In the case of the Elbrus the compiler (or the sofware designer in assembly) needs to make the software parelell executable.


    This actualy means that an elbrus CPU can be as fast as a four -eight times bigger / faster (by MHz) x86 CPU .

    So the Russians actualy can make as fast CPUs on 65 nm as the intel on 32nm , maybe on 22 nm.
    BUT the drawback is that the software deisng OR the compiler deisng become very complicated.


    There is no known compiler for VLIW instructions - maybe the russian military industry has one.


    But even if only part of the code is hand coded the CPU can works close to its theretical performance - like the algorithms for radar signal processing.


    By the wiki actualy the Elbrus like architecture quite popular in the DSP processors, and GPUs.


    The whole "goverment, industrial" line born due to the implementation of backdors in the intel/amd cpus.

    The AMD started to roll out this around 2013, so the Russian goverment had to make a domestic CPU for high risk / secure goverment communication. The Elbrus was the first and best choice, considering its availability.


    https://libreboot.org/faq.html#hardware-compatibility
    https://en.wikipedia.org/wiki/Very_long_instruction_word
    George1
    George1


    Posts : 18488
    Points : 18991
    Join date : 2011-12-22
    Location : Greece

    PAK-FA, T-50: News #4 - Page 5 Empty Re: PAK-FA, T-50: News #4

    Post  George1 Wed May 03, 2017 10:38 am

    PAK-FA will carry Kh-35UE anti-ship missile

    Big_Gazza
    Big_Gazza


    Posts : 4839
    Points : 4829
    Join date : 2014-08-25
    Location : Melbourne, Australia

    PAK-FA, T-50: News #4 - Page 5 Empty Re: PAK-FA, T-50: News #4

    Post  Big_Gazza Tue May 09, 2017 4:30 pm

    PAK-FA, T-50: News #4 - Page 5 209286
    avatar
    Vann7


    Posts : 5385
    Points : 5485
    Join date : 2012-05-16

    PAK-FA, T-50: News #4 - Page 5 Empty Re: PAK-FA, T-50: News #4

    Post  Vann7 Wed May 10, 2017 7:03 pm

    hoom wrote:F-22 has really old processors.
    Its like 20 GFLOPS total after upgrade (same as original Xbox)

    F-35 has newer processors, seen a number more like 200 GFLOPS which is much better but still not that impressive, a bit less than an Xbox 360.
    Saw another number of 1 TFLOPS which would be much better but below XboxOne (1.3 TFLOPS)

    There is a rumor the new T-50 processor setup is this:
    The architecture of processor cores: "Elbrus" with an extended set of commands
    The total number of processor cores OBC: 24 kernel
    Core clock frequency (not less than): 600 MHz
    Total capacity: up to 20,000 DMIPS (Dhrystone test), up to 150 GFLOPS for double-precision numbers, up to 300 GFLOPS for single-precision (for comparison, Intel Core2 Quad Q8300 2.5 GHz, 4 cores – 40 GFLOPS)
    So that'd be somewhat better than the low F-35 number but well behind the high.

    You need to keep in mind ,that the needs for a combat plane processing power are totally different than a gaming console. The combat jets will not operate heavy graphics at all. but
    very simplistic ones , of inferior quality of what you have in your iphone. The advantage of designing a very simple processor is that it will not need an advanced cooling solution ,that will need to be replaced its fan every x number of months or that if the fan fails ,the entire
    computer of the combat plane shuts down. So even though im not aware of the exact details
    of what Pak-fa is using , what ,they will like to see is a plane that is very powerful but at the same time very easy and cheap to maintain. and if possible a processor that use passive cooling ,that is without fan ,and just a heatsink. So simplicity will be given a really good
    preference. Because in times of emergency , when your country is at war, even 1-2 minutes
    late is too late.. and could be the difference of your base being destroyed or Pak-fa to intercept the intruder. So all this processor speeds ,are meaningless. Americans could equip
    and upgrade their F-35 if they wanted it ,with the latest kabylake 4.2Ghz processors. or the 10 core ones running overclocked at 3.5ghz. that will be pure power . but those kind of processors will over heat a lot , and will not last more than 10 years. and when a military build
    a combat plane ,they try to build it ,with hardware that will last for a very long time and with very little service if any. So a Pentium 2 processor running at 2ghz that does not produce much heat and needs no monitoring and dont use fans to cool it , that is things that needs to be cleaned every week ,.could be more ideal at times than a latest generation moder Processor.
    avatar
    hoom


    Posts : 2352
    Points : 2340
    Join date : 2016-05-06

    PAK-FA, T-50: News #4 - Page 5 Empty Re: PAK-FA, T-50: News #4

    Post  hoom Thu May 11, 2017 6:28 am

    Lolwat?! Suspect
    You were the one complaining that T-50 was going to be lacking in processing power in the first place.

    And CPU cooling is most certainly not the reason for one plane to have weaker processing than another Laughing

    Can't just stick a Kabylake in unless you don't mind being incredibly vulnerable to EMP/EW, you need specifically hardened processors.
    Singular_Transform
    Singular_Transform


    Posts : 1032
    Points : 1014
    Join date : 2016-11-13

    PAK-FA, T-50: News #4 - Page 5 Empty Re: PAK-FA, T-50: News #4

    Post  Singular_Transform Thu May 11, 2017 11:07 pm

    Tha aircrafts cooling the electronics with fuel means higher power consumption decrease the effective range of the aircraft.

    And the radar using up a lot of processing power.
    GarryB
    GarryB


    Posts : 40376
    Points : 40876
    Join date : 2010-03-30
    Location : New Zealand

    PAK-FA, T-50: News #4 - Page 5 Empty Re: PAK-FA, T-50: News #4

    Post  GarryB Fri May 12, 2017 12:33 pm

    pumping fuel through a radiator system to cool the electronics wont use "power".


    The AESA elements of the 5 radar systems will require cooling too... not really that big a deal... fluids are quite efficient at removing excess heat and heating fuel before putting it through the engine makes the engine more efficient... the whole concept of fuel in a jet engine is to produce heat... the more heat the better.
    PapaDragon
    PapaDragon


    Posts : 13454
    Points : 13494
    Join date : 2015-04-26
    Location : Fort Evil, Serbia

    PAK-FA, T-50: News #4 - Page 5 Empty Re: PAK-FA, T-50: News #4

    Post  PapaDragon Fri May 12, 2017 4:22 pm


    Engine nozzles aside, I believe that what we see here is final form of this aircraft visually speaking.

    Not a fan of new paintjob but it will save us from having to look for serial number on this particular airframe.
    Singular_Transform
    Singular_Transform


    Posts : 1032
    Points : 1014
    Join date : 2016-11-13

    PAK-FA, T-50: News #4 - Page 5 Empty Re: PAK-FA, T-50: News #4

    Post  Singular_Transform Sat May 13, 2017 12:42 pm

    GarryB wrote:pumping fuel through a radiator system to cool the electronics wont use "power".


    The AESA elements of the 5 radar systems will require cooling too... not really that big a deal... fluids are quite efficient at removing excess heat and heating fuel before putting it through the engine makes the engine more efficient... the whole concept of fuel in a jet engine is to produce heat... the more heat the better.
    The temperature of the fuel at supersonic speed is above 60 Celsius , the temperature of the air is above 120 Celsius.

    Either you evaporate fuel, or you use turbine power to drive compressors.

    The radar (regardless of AESA or PESA) using up a lot of cooling.

    To cool pesa is easier , it it not that sensitive the different temperatures, because you have only one(or two ) travelling wave tube,and that is it).
    Of course you have to keep it cool,otherwise the expected life dramatical decrease.


    BUT the ASEA require very homogenised cooling by position and by time, otherwise the performance of the radar degradate dramatically.


    The gas turbine performance bound by Carnot, means cooler fuel/air intake means higher efficient and performance. -> see Charge Air Coolers on cars/trucks (CAC)

    berhoum
    berhoum


    Posts : 119
    Points : 130
    Join date : 2015-05-19
    Age : 66
    Location : France

    PAK-FA, T-50: News #4 - Page 5 Empty PAK-FA, T-50: News #4

    Post  berhoum Sat May 13, 2017 3:13 pm

    PAK-FA, T-50: News #4 - Page 5 209560
    GarryB
    GarryB


    Posts : 40376
    Points : 40876
    Join date : 2010-03-30
    Location : New Zealand

    PAK-FA, T-50: News #4 - Page 5 Empty Re: PAK-FA, T-50: News #4

    Post  GarryB Sun May 14, 2017 10:39 am

    The gas turbine performance bound by Carnot, means cooler fuel/air intake means higher efficient and performance. -> see Charge Air Coolers on cars/trucks (CAC)

    Rubbish.

    Cool air going into the intake means denser air, which means more are for thrust, but colder fuel does not help at all.

    The point is to heat the air using burning fuel... hot fuel means it burns more efficiently... the whole point is expansion via heat...
    ATLASCUB
    ATLASCUB


    Posts : 1154
    Points : 1158
    Join date : 2017-02-13

    PAK-FA, T-50: News #4 - Page 5 Empty Re: PAK-FA, T-50: News #4

    Post  ATLASCUB Mon May 15, 2017 9:00 am

    berhoum wrote:PAK-FA, T-50: News #4 - Page 5 209560

    PAK-FA, T-50: News #4 - Page 5 209565
    avatar
    chicken


    Posts : 110
    Points : 115
    Join date : 2014-09-04

    PAK-FA, T-50: News #4 - Page 5 Empty Re: PAK-FA, T-50: News #4

    Post  chicken Mon May 15, 2017 1:37 pm

    berhoum wrote:PAK-FA, T-50: News #4 - Page 5 209560

    Is it just me or does the side bays look a little small?

    Sponsored content


    PAK-FA, T-50: News #4 - Page 5 Empty Re: PAK-FA, T-50: News #4

    Post  Sponsored content


      Current date/time is Wed Oct 30, 2024 11:38 pm