Results 1 to 21 of 21

Thread: Raqia@B3D: AMD Bulldozer Core Patent Diagrams

  1. #1
    Xtreme Addict
    Join Date
    Apr 2008
    Location
    Texas
    Posts
    1,663

    Raqia@B3D: AMD Bulldozer Core Patent Diagrams

    Raqia over at Beyond3D has an interesting post about Bulldozer patents: http://forum.beyond3d.com/showthread.php?t=54018. The interesting link from his post is this one:http://citavia.blog.de/. In the actual blog post, it has links to filed patents and patent applications concerning Bulldozer. This is fact, not speculation at this point. What can we gleam from this info?

    My try to predict AMD's bulldozer core µarchitecture

    by Dresdenboy @ 2009-04-15 – 10:37:15 am

    It's time for my first blog post after publishing my thoughts in several forums for years.

    I want to start with a graphic (first published on planet3dnow) showing what some of AMD's last year's patent applications contained as an exemplary MPU architecture. It's worth to note, that this architecture fits nicely to some rumor brought up by Charlie on the Inquirer.

    Another hint is this old AMD presentation, which mentions future developments like "Throughput Architecture" and "Cluster-based Multi-threading", although not explicitly stating its planned use. However, both sources tell us about some clusters. Now this is, what appeared in the patent applications:



    Additionally there are many other interesting bits hidden in many different patent applications (long numbers) and filed patents (shorter numbers):
    clustered multithreading with 2 int clusters with each of them having:
    2 ALUs, 2 AGUs
    one L1 data cache
    scheduler, integer register file (IRF), ROB
    (see 20080263373*, 20080209173, 7315935)
    a trace cache, not to make cheaper decoders but to quickly recover from a mispredicted branch (7197630 and many others)
    read port arbitration for a faster IRF (7315935)
    shared FPU supporting ADD, MUL, FMAC etc. and 64 or 128 bit max. operand width (20080263373)
    FPU may run in full bit or reduced bit modes to save power (20080209185)
    32 byte fetch, 4-way Decoder - multithreaded round robin or depending on queue saturation (20080263373, EP1244962)
    fine grained power management (token based, 20080263373) for optimal usage of given TDP/ACP
    a lot more speculation (data speculation, cache way prediction, see 7024537, 7028166 and many others)
    2 loads from L1 D$ per cycle per cluster (7502914)
    maybe 2 cycle effective L1 D$ latency instead of 4 thanks to replaying (7502914)
    possibly a shared L2 (7502914)
    loop detectors (7130991)
    dynamically scalable cache architecture to save power by switching off cache portions or levels (20080104324)
    AMD's turbo mode (running cores faster if others are less utilized, 7490254, filed 2005/08/02)

    Even if only some of these points will be true for Bulldozer, it will be a very interesting MPU.



    * Users registered on http://www.freepatentsonline.com/ may simply click on the links to open the pdf.
    Core i7 2600K@4.6Ghz| 16GB G.Skill@2133Mhz 9-11-10-28-38 1.65v| ASUS P8Z77-V PRO | Corsair 750i PSU | ASUS GTX 980 OC | Xonar DSX | Samsung 840 Pro 128GB |A bunch of HDDs and terabytes | Oculus Rift w/ touch | ASUS 24" 144Hz G-sync monitor

    Quote Originally Posted by phelan1777 View Post
    Hail fellow warrior albeit a surat Mercenary. I Hail to you from the Clans, Ghost Bear that is (Yes freebirth we still do and shall always view mercenaries with great disdain!) I have long been an honorable warrior of the mighty Warden Clan Ghost Bear the honorable Bekker surname. I salute your tenacity to show your freebirth sibkin their ignorance!

  2. #2
    Xtreme Mentor
    Join Date
    Nov 2005
    Location
    Devon
    Posts
    3,423
    Thanks for a link Mechromancer!

    If only half of that finds it way into Bulldozer it will be very interesting MPU. From the marketing slides projected performance of Int and Fp should be much better than current cores. Of course we don't know projected core speeds on that slides, still it looks good.
    I just wonder what Intel is planning for 2011.
    BTW I really like all these power savings patents in that list
    RiG1: Ryzen 7 1700 @4.0GHz 1.39V, Asus X370 Prime, G.Skill RipJaws 2x8GB 3200MHz CL14 Samsung B-die, TuL Vega 56 Stock, Samsung SS805 100GB SLC SDD (OS Drive) + 512GB Evo 850 SSD (2nd OS Drive) + 3TB Seagate + 1TB Seagate, BeQuiet PowerZone 1000W

    RiG2: HTPC AMD A10-7850K APU, 2x8GB Kingstone HyperX 2400C12, AsRock FM2A88M Extreme4+, 128GB SSD + 640GB Samsung 7200, LG Blu-ray Recorder, Thermaltake BACH, Hiper 4M880 880W PSU

    SmartPhone Samsung Galaxy S7 EDGE
    XBONE paired with 55'' Samsung LED 3D TV

  3. #3
    Xtreme Addict
    Join Date
    Apr 2008
    Location
    Texas
    Posts
    1,663
    Damn, AMD needed Bulldozer to come out in 2009! I hate so much it got delayed (most likely due to -$$$,$$$,$$$,$$$). It's a damn shame we have to wait until 2011 for this, but I suppose 6, 8, and 12 core K10.5 High-K CPUs will fill in the gap a little .
    Core i7 2600K@4.6Ghz| 16GB G.Skill@2133Mhz 9-11-10-28-38 1.65v| ASUS P8Z77-V PRO | Corsair 750i PSU | ASUS GTX 980 OC | Xonar DSX | Samsung 840 Pro 128GB |A bunch of HDDs and terabytes | Oculus Rift w/ touch | ASUS 24" 144Hz G-sync monitor

    Quote Originally Posted by phelan1777 View Post
    Hail fellow warrior albeit a surat Mercenary. I Hail to you from the Clans, Ghost Bear that is (Yes freebirth we still do and shall always view mercenaries with great disdain!) I have long been an honorable warrior of the mighty Warden Clan Ghost Bear the honorable Bekker surname. I salute your tenacity to show your freebirth sibkin their ignorance!

  4. #4
    Xtreme Mentor
    Join Date
    May 2008
    Location
    cleveland ohio
    Posts
    2,874
    bump I expect NOV 2010 if on 45nm.

    2011 probably like phenom II for 2009....
    HAVE NO FEAR!
    "AMD fallen angel"
    Quote Originally Posted by Gamekiller View Post
    You didn't get the memo? 1 hour 'Fugger time' is equal to 12 hours of regular time.

  5. #5
    Xtreme Addict
    Join Date
    Apr 2008
    Location
    Texas
    Posts
    1,663
    I found something interesting while looking for K10.5 block diagrams to compare with the one I posted above. Apparently in January, some Germans read the patent concerning features of Bulldozer and it pretty clearly states SMT capability in the core design! Read what they discussed months ago for yourself, particularly Dresdenboy's post (translated): http://translate.google.com/translat...D6&sl=de&tl=en

    Go to the second to last page of that forum post (they've kept it going for months!) and you'll find a link to Dresdenboys' blog with an updated Bulldozer diagram he made from viewing the patents.

    Here it is for you lazy folk:


    Now lets figure out what it all means...
    Core i7 2600K@4.6Ghz| 16GB G.Skill@2133Mhz 9-11-10-28-38 1.65v| ASUS P8Z77-V PRO | Corsair 750i PSU | ASUS GTX 980 OC | Xonar DSX | Samsung 840 Pro 128GB |A bunch of HDDs and terabytes | Oculus Rift w/ touch | ASUS 24" 144Hz G-sync monitor

    Quote Originally Posted by phelan1777 View Post
    Hail fellow warrior albeit a surat Mercenary. I Hail to you from the Clans, Ghost Bear that is (Yes freebirth we still do and shall always view mercenaries with great disdain!) I have long been an honorable warrior of the mighty Warden Clan Ghost Bear the honorable Bekker surname. I salute your tenacity to show your freebirth sibkin their ignorance!

  6. #6
    Xtreme Mentor
    Join Date
    May 2008
    Location
    cleveland ohio
    Posts
    2,874
    Quote Originally Posted by Mechromancer View Post
    I found something interesting while looking for K10.5 block diagrams to compare with the one I posted above. Apparently in January, some Germans read the patent concerning features of Bulldozer and it pretty clearly states SMT capability in the core design! Read what they discussed months ago for yourself, particularly Dresdenboy's post (translated): http://translate.google.com/translat...D6&sl=de&tl=en

    Go to the second to last page of that forum post (they've kept it going for months!) and you'll find a link to Dresdenboys' blog with an updated Bulldozer diagram he made from viewing the patents.

    Here it is for you lazy folk:


    Now lets figure out what it all means...
    :o this core is a modular monster.
    HAVE NO FEAR!
    "AMD fallen angel"
    Quote Originally Posted by Gamekiller View Post
    You didn't get the memo? 1 hour 'Fugger time' is equal to 12 hours of regular time.

  7. #7
    Xtreme Cruncher
    Join Date
    Jun 2006
    Posts
    6,173
    The diagram clearly illustrates the "clustered" core uarchitecture ,first presented by Fred Weber in 2005 at Analyst meeting.
    The BD design that will come in 2011(or late 2010) is a 2nd or maybe even 3rd iteration of the original idea that was born some 8 years ago in AMD.
    The latest iteration,judging by the patents submitted in '07 and '08 is practically a throughput clustered design with basic core being made of 2 (2 way!) int pipelines and one (4 way!) "super wide"-256bit- SSE/FP (a la Sandy Bridge AVX). The design should also have much more advanced power management features,with AMD's version of "Turbo" mode,many new tweaks to the branch prediction,IMC etc. There will be support for all previous and future x86 instruction set extensions(AVX and XOP/FMA4 included).
    There is some talk in the patents about "adapting" ISA capabilities of the cores(ie. cores can "learn" via firmware the, at first unsupported, instruction set extensions-this one may be slated for 2012 refresh of the BD cores).
    All in all,Dresdenboy has done amazing job at detailing the patents so far. All we can do is wait for the AMD's next big Hammer to land(on the opponents ).
    Last edited by informal; 08-22-2009 at 05:41 PM.

  8. #8
    Xtreme Addict
    Join Date
    Apr 2007
    Posts
    2,477
    Quote Originally Posted by Mechromancer View Post

    Now lets figure out what it all means...
    AM3s like the 965 being priced lower and lower per month then in 2012 something new maybe? I'm just going off of recent AMD history. I hope not but I'm going off of what a recent article posted on here that 965 could be around w/out anything new until mid 2010
    i3 2100, MSI H61M-E33. 8GB G.Skill Ripjaws.
    MSI GTX 460 Twin Frozr II. 1TB Caviar Blue.
    Corsair HX 620, CM 690, Win 7 Ultimate 64bit.

  9. #9
    Xtreme Cruncher
    Join Date
    Jun 2006
    Posts
    6,173
    Bulldozer for desktop is Q1 2011,but looking at AMD's excellent Deneb/Istanbul launches(pushed forward both server parts),BD may land even in late Q4 2010 IF GF meets its 32nm SOI goals.
    Until that happens we may see higher clocked Denebs and hex core on desktop.Maybe even a new revision of Deneb with 1MB of L2 per core done @ 32nm to tide them over until Bulldozer comes(sort of 32nm practice run for both GF and AMD).

  10. #10
    Xtreme Cruncher
    Join Date
    Apr 2008
    Location
    Ohio
    Posts
    3,119
    Reading this, gets me all kinds of excited . I keep thinking that the Phenom II line is just the "tide me over" that AMD is using to really square away this new design. This might be why the Phenom line is just getting minor tweaks here in there. Kinda like Intel did with the Core arch. Presott was getting old quick, but Intel was able to string it along until "WAP" Core duo hit like a bomb. I have high hopes and in time we'll see.
    ~1~
    AMD Ryzen 9 3900X
    GigaByte X570 AORUS LITE
    Trident-Z 3200 CL14 16GB
    AMD Radeon VII
    ~2~
    AMD Ryzen ThreadRipper 2950x
    Asus Prime X399-A
    GSkill Flare-X 3200mhz, CAS14, 64GB
    AMD RX 5700 XT

  11. #11
    Xtreme Mentor
    Join Date
    May 2008
    Location
    cleveland ohio
    Posts
    2,874
    Quote Originally Posted by informal View Post
    The diagram clearly illustrates the "clustered" core uarchitecture ,first presented by Fred Weber in 2005 at Analyst meeting.
    The BD design that will come in 2011(or late 2010) is a 2nd or maybe even 3rd iteration of the original idea that was born some 8 years ago in AMD.
    The latest iteration,judging by the patents submitted in '07 and '08 is practically a throughput clustered design with basic core being made of 2 (2 way!) int pipelines and one (4 way!) "super wide"-256bit- SSE/FP (a la Sandy Bridge AVX). The design should also have much more advanced power management features,with AMD's version of "Turbo" mode,many new tweaks to the branch prediction,IMC etc. There will be support for all previous and future x86 instruction set extensions(AVX and XOP/FMA4 included).
    There is some talk in the patents about "adapting" ISA capabilities of the cores(ie. cores can "learn" via firmware the, at first unsupported, instruction set extensions-this one may be slated for 2012 refresh of the BD cores).
    All in all,Dresdenboy has done amazing job at detailing the patents so far. All we can do is wait for the AMD's next big Hammer to land(on the opponents ).
    if it's a cluster will improve any single threaded or just fall off a cliff ???...

    I'd love to see all cores loaded from one thread being weaved threw each core hehe.
    HAVE NO FEAR!
    "AMD fallen angel"
    Quote Originally Posted by Gamekiller View Post
    You didn't get the memo? 1 hour 'Fugger time' is equal to 12 hours of regular time.

  12. #12
    Xtreme Addict
    Join Date
    Apr 2008
    Location
    Texas
    Posts
    1,663
    From the patent:
    SUMMARY

    Various embodiments of an apparatus for executing branch predictor directed prefetch operations are disclosed. The apparatus may include an instruction cache, a fetch unit, and a branch prediction unit. According to one embodiment, the branch prediction unit may provide an address of a first instruction to the fetch unit. The fetch unit may send a fetch request for the first instruction to the instruction cache to perform a fetch operation. In response to detecting a cache miss corresponding to the first instruction, the fetch unit may execute one or more prefetch operation while the cache miss corresponding to the first instruction is being serviced. The branch prediction unit may provide an address of a predicted next instruction to the fetch unit. The branch prediction unit may predict the address of the next instruction based on the predicted outcome of various branches in the instruction stream. The fetch unit may send a prefetch request for the predicted next instruction to the instruction cache to execute the prefetch operation.

    In one embodiment, in response to detecting a cache miss corresponding to the predicted next instruction, the fetch unit may send a prefetch request for the predicted next instruction to a next level of memory, e.g., an L2 cache. If a cache hit is detected in the next level of memory, the fetch unit may store prefetched instruction data corresponding to the predicted next instruction in the instruction cache. In other embodiments, the fetch unit may store prefetched instruction data corresponding to the predicted next instruction in a prefetch buffer. The prefetch request may be sent to other parts of the memory hierarchy of the system. For instance, if a cache miss is detected in the L2 cache, the prefetch request may be sent to an L3 cache or main memory until the instruction data is found. In other embodiments, the prefetch operation may be aborted if a cache miss is detected in the L2 cache.

    In one embodiment, if a cache hit is detected corresponding to the predicted next instruction, the fetch unit may send a next prefetch request for a subsequent predicted instruction to the instruction cache to execute a next prefetch operation. The fetch unit may obtain the address of the subsequent predicted instruction from the branch prediction unit. After servicing the cache miss corresponding to the first instruction, the fetch unit may stop executing prefetch operations and resume execution of fetch operations to the instruction cache.
    BRIEF DESCRIPTION OF THE DRAWINGS

    FIG. 1 is a block diagram of one embodiment of an exemplary front end of a processor core;

    FIG. 2 is a flow diagram illustrating a method for performing instruction fetch and prefetch operations, according to one embodiment;

    FIG. 3 is a block diagram of one embodiment of the processor core of FIG. 1; and

    FIG. 4 is a block diagram of one embodiment of a processing unit including multiple processing cores
    .
    I can't find a site that has these 4 diagrams referenced above, but what is up with FIG. 4? What are "multiple processing cores"? SMT maybe?

    I don't want to get my own hopes up, but this CPU sounds like it will definitely hit us with a ton of bricks like Intel's Core uarch did.

    Thank you Informal for your added info. I honestly was not all that excited about Bulldozer, but this has rekindled the fire.
    Core i7 2600K@4.6Ghz| 16GB G.Skill@2133Mhz 9-11-10-28-38 1.65v| ASUS P8Z77-V PRO | Corsair 750i PSU | ASUS GTX 980 OC | Xonar DSX | Samsung 840 Pro 128GB |A bunch of HDDs and terabytes | Oculus Rift w/ touch | ASUS 24" 144Hz G-sync monitor

    Quote Originally Posted by phelan1777 View Post
    Hail fellow warrior albeit a surat Mercenary. I Hail to you from the Clans, Ghost Bear that is (Yes freebirth we still do and shall always view mercenaries with great disdain!) I have long been an honorable warrior of the mighty Warden Clan Ghost Bear the honorable Bekker surname. I salute your tenacity to show your freebirth sibkin their ignorance!

  13. #13
    Xtreme Addict
    Join Date
    Apr 2008
    Location
    Texas
    Posts
    1,663
    I just posted an update and continuation of this story on the news page: http://www.xtremesystems.org/forums/...d.php?t=232784
    Core i7 2600K@4.6Ghz| 16GB G.Skill@2133Mhz 9-11-10-28-38 1.65v| ASUS P8Z77-V PRO | Corsair 750i PSU | ASUS GTX 980 OC | Xonar DSX | Samsung 840 Pro 128GB |A bunch of HDDs and terabytes | Oculus Rift w/ touch | ASUS 24" 144Hz G-sync monitor

    Quote Originally Posted by phelan1777 View Post
    Hail fellow warrior albeit a surat Mercenary. I Hail to you from the Clans, Ghost Bear that is (Yes freebirth we still do and shall always view mercenaries with great disdain!) I have long been an honorable warrior of the mighty Warden Clan Ghost Bear the honorable Bekker surname. I salute your tenacity to show your freebirth sibkin their ignorance!

  14. #14
    Xtreme Mentor
    Join Date
    Jul 2008
    Location
    Shimla , India
    Posts
    2,624
    As i said on the news page this is a very good core if true. The core is a cluster and reminds me of semi modular forum of the nehalem arc. The arc. is off course something new all together. The k10/10.5 is not so complex dont know about nehalem tough. The cores seems to have higher core inter dependency and communication. The sandy bridge seems more of a evo of the nehalem/clarkdale than a full new arc.

  15. #15
    Xtreme Member
    Join Date
    Apr 2005
    Location
    London, UK
    Posts
    261
    One of the AMD's server PR people said, that AMD does not have any plans for Hyperthreading like core system. So that means Bulldozer will not have any kind of HT implementation.

  16. #16
    I am Xtreme FlanK3r's Avatar
    Join Date
    May 2008
    Location
    Czech republic
    Posts
    6,804
    I heard, AMD have some samples "Orochi" to time...but nothing more
    ROG Power PCs - Intel and AMD
    CPUs:i9-7900X, i9-9900K, i7-6950X, i7-5960X, i7-8086K, i7-8700K, 4x i7-7700K, i3-7350K, 2x i7-6700K, i5-6600K, R7-2700X, 4x R5 2600X, R5 2400G, R3 1200, R7-1800X, R7-1700X, 3x AMD FX-9590, 1x AMD FX-9370, 4x AMD FX-8350,1x AMD FX-8320,1x AMD FX-8300, 2x AMD FX-6300,2x AMD FX-4300, 3x AMD FX-8150, 2x AMD FX-8120 125 and 95W, AMD X2 555 BE, AMD x4 965 BE C2 and C3, AMD X4 970 BE, AMD x4 975 BE, AMD x4 980 BE, AMD X6 1090T BE, AMD X6 1100T BE, A10-7870K, Athlon 845, Athlon 860K,AMD A10-7850K, AMD A10-6800K, A8-6600K, 2x AMD A10-5800K, AMD A10-5600K, AMD A8-3850, AMD A8-3870K, 2x AMD A64 3000+, AMD 64+ X2 4600+ EE, Intel i7-980X, Intel i7-2600K, Intel i7-3770K,2x i7-4770K, Intel i7-3930KAMD Cinebench R10 challenge AMD Cinebench R15 thread Intel Cinebench R15 thread

  17. #17
    Xtreme Mentor
    Join Date
    Nov 2005
    Location
    Devon
    Posts
    3,423
    Quote Originally Posted by FlanK3r View Post
    I heard, AMD have some samples "Orochi" to time...but nothing more
    About time!

    If they are really lucky, they can probably validate and correct design within 12-14 months. That would mean Q3/Q4 2010 launch which is most we can count on. But bear in mind AMD is doing new arch + new node at the same time so it might be not so smooth for them ... TIME WILL TELL
    RiG1: Ryzen 7 1700 @4.0GHz 1.39V, Asus X370 Prime, G.Skill RipJaws 2x8GB 3200MHz CL14 Samsung B-die, TuL Vega 56 Stock, Samsung SS805 100GB SLC SDD (OS Drive) + 512GB Evo 850 SSD (2nd OS Drive) + 3TB Seagate + 1TB Seagate, BeQuiet PowerZone 1000W

    RiG2: HTPC AMD A10-7850K APU, 2x8GB Kingstone HyperX 2400C12, AsRock FM2A88M Extreme4+, 128GB SSD + 640GB Samsung 7200, LG Blu-ray Recorder, Thermaltake BACH, Hiper 4M880 880W PSU

    SmartPhone Samsung Galaxy S7 EDGE
    XBONE paired with 55'' Samsung LED 3D TV

  18. #18
    Brilliant Idiot
    Join Date
    Jan 2005
    Location
    Hell on Earth
    Posts
    11,085
    I hope these continue the trend of liking the cold.
    heatware chew*
    I've got no strings to hold me down.
    To make me fret, or make me frown.
    I had strings but now I'm free.
    There are no strings on me

  19. #19
    I am Xtreme FlanK3r's Avatar
    Join Date
    May 2008
    Location
    Czech republic
    Posts
    6,804
    chew* roger ...and performance more better than PII
    ROG Power PCs - Intel and AMD
    CPUs:i9-7900X, i9-9900K, i7-6950X, i7-5960X, i7-8086K, i7-8700K, 4x i7-7700K, i3-7350K, 2x i7-6700K, i5-6600K, R7-2700X, 4x R5 2600X, R5 2400G, R3 1200, R7-1800X, R7-1700X, 3x AMD FX-9590, 1x AMD FX-9370, 4x AMD FX-8350,1x AMD FX-8320,1x AMD FX-8300, 2x AMD FX-6300,2x AMD FX-4300, 3x AMD FX-8150, 2x AMD FX-8120 125 and 95W, AMD X2 555 BE, AMD x4 965 BE C2 and C3, AMD X4 970 BE, AMD x4 975 BE, AMD x4 980 BE, AMD X6 1090T BE, AMD X6 1100T BE, A10-7870K, Athlon 845, Athlon 860K,AMD A10-7850K, AMD A10-6800K, A8-6600K, 2x AMD A10-5800K, AMD A10-5600K, AMD A8-3850, AMD A8-3870K, 2x AMD A64 3000+, AMD 64+ X2 4600+ EE, Intel i7-980X, Intel i7-2600K, Intel i7-3770K,2x i7-4770K, Intel i7-3930KAMD Cinebench R10 challenge AMD Cinebench R15 thread Intel Cinebench R15 thread

  20. #20
    I am Xtreme FlanK3r's Avatar
    Join Date
    May 2008
    Location
    Czech republic
    Posts
    6,804
    whau, news in XBitlabs

    http://www.xbitlabs.com/news/cpu/dis...echnology.html

    UPDATE: AMD has contacted X-bit labs claiming that it has not announced any simultaneous multi-threading technologies for Bulldozer processors. Still, there are other multi-threading implementations that may still be supported.

    Advanced Micro Devices announced during Hot Chips conference that its next-generation code-named Bulldozer microprocessors will feature a multi-threading technology (SMT) which would be akin to Intel Corp.’s well-known HyperThreading.

    AMD did not reveal many details about its multi-threading capability and only said that its Bulldozer processors would support it in 2011. Still, it is rather likely that AMD’s approach may be somewhat different compared to Intel’s HT and may even be of the same kind like Sun Microsystems’ simultaneous multi-threading feature than supports execution of four threads on one physical core.


    etc...
    ROG Power PCs - Intel and AMD
    CPUs:i9-7900X, i9-9900K, i7-6950X, i7-5960X, i7-8086K, i7-8700K, 4x i7-7700K, i3-7350K, 2x i7-6700K, i5-6600K, R7-2700X, 4x R5 2600X, R5 2400G, R3 1200, R7-1800X, R7-1700X, 3x AMD FX-9590, 1x AMD FX-9370, 4x AMD FX-8350,1x AMD FX-8320,1x AMD FX-8300, 2x AMD FX-6300,2x AMD FX-4300, 3x AMD FX-8150, 2x AMD FX-8120 125 and 95W, AMD X2 555 BE, AMD x4 965 BE C2 and C3, AMD X4 970 BE, AMD x4 975 BE, AMD x4 980 BE, AMD X6 1090T BE, AMD X6 1100T BE, A10-7870K, Athlon 845, Athlon 860K,AMD A10-7850K, AMD A10-6800K, A8-6600K, 2x AMD A10-5800K, AMD A10-5600K, AMD A8-3850, AMD A8-3870K, 2x AMD A64 3000+, AMD 64+ X2 4600+ EE, Intel i7-980X, Intel i7-2600K, Intel i7-3770K,2x i7-4770K, Intel i7-3930KAMD Cinebench R10 challenge AMD Cinebench R15 thread Intel Cinebench R15 thread

  21. #21
    Xtreme Enthusiast
    Join Date
    Oct 2007
    Location
    Singapore
    Posts
    967
    throw more MAC and ALU into the core pls
    also not to forget vertex units for multimedia enhanced apps
    IPC increment is hard to achieve within short time, unlike Intel, has a lot of resource to throw in to fine tune the IPC performance. AMD has no extra resources to do this research so focus on more units in a core.
    same goes with branch prediction hit rate, shorten the branch a little bit and shorten the re-build time in the brunch if it was a miss.
    Main Rig:
    Processor & Motherboard:AMD Ryzen5 1400 ' Gigabyte B450M-DS3H
    Random Access Memory Module:Adata XPG DDR4 3000 MHz 2x8GB
    Graphic Card:XFX RX 580 4GB
    Power Supply Unit:FSP AURUM 92+ Series PT-650M
    Storage Unit:Crucial MX 500 240GB SATA III SSD
    Processor Heatsink Fan:AMD Wraith Spire RGB
    Chasis:Thermaltake Level 10GTS Black

Bookmarks

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •