Page 4 of 6 FirstFirst 123456 LastLast
Results 76 to 100 of 135

Thread: Radeon HD 7770 and 7750 Specs surface

  1. #76
    Xtreme Addict
    Join Date
    May 2007
    Location
    Romania
    Posts
    1,246
    You are not too shabby either We shall see tomorrow the real performance and o/c potential and stuff.

  2. #77
    Xtreme Cruncher
    Join Date
    Jun 2006
    Posts
    6,215
    We already know where will AMD position 7770 in their product stack. The point is that if they position it above 6790 and below 6850 then it can only be somewhere midway between these two and the spread is 10%. So 5% slower than 6850 is pretty much where it will land.I guess there will be games where it can be even faster than 6850 but overall that's the place where it will land. But when you count in OCing,TDP and price(after it settles a bit),it should be a good product. Sorry to hear you expected 5850/5870 killer... You will probably be disappointed by Pitcairn too with its performance versus 6970...

  3. #78
    I am Xtreme
    Join Date
    Dec 2007
    Posts
    7,750
    ^the old days the X7X0 were much slower than the X8X0 products because in the days for 4800 and 5800s the top of the line was not a X9X0, so now they have a few extra products, the gap between them is all very small and something is available for every price range.

    i could understand expecting the 7770 to beat a 5800, because i think we hoped the 7770 was faster than a 6870 and a 7870 faster than a 6970. but if thats not the case then maybe these will all be replaced with a bigger gpu in the next 10-14 months where each line gets another 15-25% faster.
    2500k @ 4900mhz - Asus Maxiums IV Gene Z - Swiftech Apogee LP
    GTX 680 @ +170 (1267mhz) / +300 (3305mhz) - EK 680 FC EN/Acteal
    Swiftech MCR320 Drive @ 1300rpms - 3x GT 1850s @ 1150rpms
    XS Build Log for: My Latest Custom Case

  4. #79
    Xtreme Guru
    Join Date
    Aug 2007
    Posts
    3,562
    Quote Originally Posted by informal View Post
    We already know where will AMD position 7770 in their product stack. The point is that if they position it above 6790 and below 6850 then it can only be somewhere midway between these two and the spread is 10%. So 5% slower than 6850 is pretty much where it will land.I guess there will be games where it can be even faster than 6850 but overall that's the place where it will land. But when you count in OCing,TDP and price(after it settles a bit),it should be a good product. Sorry to hear you expected 5850/5870 killer... You will probably be disappointed by Pitcairn too with its performance versus 6970...
    The spread between the HD 6790 and HD 6850 was closer to 15% at release.

    HOWEVER, this launch more than any other will come down to price. The HD 6790 currently retails for ~$120 before MIR while the HD 6850 is ~$150 before MIR. Even the GTX 560 muscles into this price point with MIRs. That's a seriously thin needle to thread so AMD needs to nail down their cost relative to performance in the current market. Not some mythological future market where they've discontinued everything else.

  5. #80
    Xtreme Enthusiast
    Join Date
    Apr 2010
    Posts
    514
    PowerColor Radeon HD7770



    http://www.neon24.de/PC-Hardware/Gra...70::27611.html

    • Chip clock: 1000MHz, Memory clock: 1250MHz
    • Chip: Cape Verde XT
    • Memory interface: 128-bit
    Stream processors: 896
    • Texture units: 16
    • Manufacturing process: 28nm
    • Maximum power consumption: not specified (operation), not specified (no load)
    • DirectX: 11.1
    • Shader model: 5
    • Construction: dual-slot
    • Dimensions: not specified
    • Interface: PCIe 3.0
    • Special features: Intel integrated 7.1 HD audio controller, CrossFireX

  6. #81
    Xtreme Cruncher
    Join Date
    Jun 2006
    Posts
    6,215
    That is strange. From here (heise.de) we have a bit different specs:
    Chiptakt: 1000MHz, Speichertakt: 1250MHz • Chip: Cape Verde XT • Speicherinterface: 128-bit • Stream-Prozessoren: 640 • Textureinheiten: 16 • Fertigung: 28nm • Maximaler Verbrauch: keine Angabe (Betrieb), keine Angabe (Leerlauf) • DirectX: 11.1 • Shader Modell: 5 • Bauweise: keine Angabe • Abmessungen: keine Angabe • Schnittstelle: PCIe 3.0 • Besonderheiten: AMD Eyefinity, integrierter 7.1 HD Audiocontroller, unterstützt CrossFireX

  7. #82
    Xtreme Member
    Join Date
    Jan 2011
    Location
    Poland
    Posts
    353
    http://preisvergleich.hardwareluxx.de/735500

    In the morning it was 896SP...now it is 640SP in description.
    i5 2500K (L041C124) @ 5GHz + Scythe Mugen 2 rev. B | ASRock P67 Extreme4 B3 UEFI L3.19 | ADATA 2x4GB DDR3 1600 | MSI Radeon RX 470 4GB | 2x Crucial m4 64GB SSD RAID 0, Seagate 7200.12 500GB, Samsung F4 EG 2TB | 24" HP LP2475w | EVGA SuperNOVA G2 750W | Fractal Design Define R3 | Windows 10 64 bit

  8. #83
    Registered User
    Join Date
    Jun 2009
    Posts
    52
    Quote Originally Posted by Lightman View Post
    First of all I meant architecture and therefor per shader performance, not transistor count per work done. If you want to compare it that way then with almost every step back you will see less transistors used (but more specialized) needed for particular task. You can't really add more features like improved tesselation, UVD encode block, DX11.1 extension, better caching, vastly improved compute and power gating (big transistor hog) with same transistor budget as older design.
    What is important is how well new architecture is aligned with todays tasks as well as manufacturing process.
    For me GCN is placed very well in both. Die sizes for Thaiti and Cape Verde are smaller than chips they replace and are much more efficient at dealing with todays tasks we expect them to run. Besides power consumption is just great on these parts compared to older gen!
    Even after u take those improvements into account, you're still talking about an increase from 1 billion to 1,5 billion transistors. That's a 500 million or a 50% increase in transistors while decreasing shaders at the same time. There is no doubt GCN shaders are bigger, way bigger than an average VLIW5 shaders. For one GCN shader, you could've crammed 1.5 to 2 VLIW5/VLIW4 in the same die area which would have offered better performance for gaming. The biggest advantage GCN has, is its compute abilities which offers more than twice the performance of VLIW. For this, they did sacrificed some gaming performance.
    It's not a gain-only architecture, they made a trade-off.

    Im pretty sure a die-shrink from 40nm to 28nm would resulted into the things u said in those last 2 sentences, not necessary thanks to GCN. If you look at the transistor budget for HD7970.. that would've fitted two and a half Barts.
    Check out the performance of HD6870 in crossfire and compare it with HD7970.

  9. #84
    Xtreme Mentor
    Join Date
    Nov 2005
    Location
    Devon
    Posts
    3,437
    Quote Originally Posted by bladerash View Post
    Even after u take those improvements into account, you're still talking about an increase from 1 billion to 1,5 billion transistors. That's a 500 million or a 50% increase in transistors while decreasing shaders at the same time. There is no doubt GCN shaders are bigger, way bigger than an average VLIW5 shaders. For one GCN shader, you could've crammed 1.5 to 2 VLIW5/VLIW4 in the same die area which would have offered better performance for gaming. The biggest advantage GCN has, is its compute abilities which offers more than twice the performance of VLIW. For this, they did sacrificed some gaming performance.
    It's not a gain-only architecture, they made a trade-off.

    Im pretty sure a die-shrink from 40nm to 28nm would resulted into the things u said in those last 2 sentences, not necessary thanks to GCN. If you look at the transistor budget for HD7970.. that would've fitted two and a half Barts.
    Check out the performance of HD6870 in crossfire and compare it with HD7970.
    GCN shader core is not much bigger than VLIW5 shader core. AMD themself said that. Everything else is. GCN has way more cache and they are R/W to start with. ROP, TMU, Memory interface, PCIe 3.0 interface, I/O for 6 displays, UVD, encode block, cross-bar, etc. They take way more transistors than shaders itself. I would guess in small GPU like HD7770 shader core will be less than 50% of die area. To see what I'm talking about just google Trinity die shoot
    Few extra points to think/read about:
    - HD6870 no DP support compared to HD6970/HD7970
    - same functionality chip implemented on 40nm process will usually have less transistors spend compared to 32nm or 28nm chip. In other words 28nm Juniper would have anywhere between 5% to 20% more transistors just to be efficiently implemented on new node.
    - more than 1tri/clk is very expensive logic wise (HD6970 already paid that price but HD7970 is even more refined / not necessary will apply to Cape Verde)

    You have a valid point in that GCN sacrifices some gaming performance compared to VLIW5/4 in yesterdays and todays games, but future engines or even current BF3/Civ5 are geared more towards what GCN and Fermi offer.
    Lets see what small GCN chip can do in few hours Maybe it's not that bad, or thought I expect price to be too high in relation to older generation initially ...
    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

  10. #85
    Xtreme Addict
    Join Date
    May 2007
    Location
    'Zona
    Posts
    2,346
    Quote Originally Posted by Lightman View Post
    GCN shader core is not much bigger than VLIW5 shader core. AMD themself said that. Everything else is. GCN has way more cache and they are R/W to start with. ROP, TMU, Memory interface, PCIe 3.0 interface, I/O for 6 displays, UVD, encode block, cross-bar, etc. They take way more transistors than shaders itself. I would guess in small GPU like HD7770 shader core will be less than 50% of die area. To see what I'm talking about just google Trinity die shoot
    Few extra points to think/read about:
    - HD6870 no DP support compared to HD6970/HD7970
    - same functionality chip implemented on 40nm process will usually have less transistors spend compared to 32nm or 28nm chip. In other words 28nm Juniper would have anywhere between 5% to 20% more transistors just to be efficiently implemented on new node.
    - more than 1tri/clk is very expensive logic wise (HD6970 already paid that price but HD7970 is even more refined / not necessary will apply to Cape Verde)

    You have a valid point in that GCN sacrifices some gaming performance compared to VLIW5/4 in yesterdays and todays games, but future engines or even current BF3/Civ5 are geared more towards what GCN and Fermi offer.
    Lets see what small GCN chip can do in few hours Maybe it's not that bad, or thought I expect price to be too high in relation to older generation initially ...
    Saved me from having to type that much.
    Would like to add a few things, scheduling and power gating.

    Edit- RV770 and RV740 ALUs took up some 30-40% of the die area, give or take a little depending on exactly what you consider part of the ALU.
    Last edited by LordEC911; 02-14-2012 at 05:39 PM.
    Originally Posted by motown_steve
    Every genocide that was committed during the 20th century has been preceded by the disarmament of the target population. Once the government outlaws your guns your life becomes a luxury afforded to you by the state. You become a tool to benefit the state. Should you cease to benefit the state or even worse become an annoyance or even a hindrance to the state then your life becomes more trouble than it is worth.

    Once the government outlaws your guns your life is forfeit. You're already dead, it's just a question of when they are going to get around to you.

  11. #86
    Xtreme Mentor
    Join Date
    Feb 2008
    Posts
    2,671
    I really dont like that the shader count was cut down from 800 on the previous *770s to 640 on these, and I didnt want to believe it before. It will be interesting to see how these perform, but I doubt that they are going to bring any significant performance increase in the <Ł150 price range.

  12. #87
    Xtreme Addict
    Join Date
    May 2003
    Location
    Hopatcong, NJ
    Posts
    1,078
    that thing is going to be so bandwidth starved.... but in any case, a much needed and significant stepup from 5770/6770
    Should be fine for playing at 1080p with no AA, unless they plan on using more shader-based post processing AA/effects like Morphological AA

  13. #88
    Xtreme Guru
    Join Date
    Aug 2007
    Posts
    3,562
    Quote Originally Posted by Miwo View Post
    that thing is going to be so bandwidth starved.... but in any case, a much needed and significant stepup from 5770/6770
    Should be fine for playing at 1080p with no AA, unless they plan on using more shader-based post processing AA/effects like Morphological AA
    A much needed and significant step up from the HD 5770 was the HD 6790....a card that was released for $149.

  14. #89
    Xtreme Mentor
    Join Date
    Jun 2008
    Location
    France - Bx
    Posts
    2,601

  15. #90
    Xtreme Addict
    Join Date
    Jan 2007
    Location
    Brisbane, Australia
    Posts
    1,264
    Quote Originally Posted by radier View Post


    Enjoy!
    Less bw than 5770/6770.. Lower tdp bracket. almost doesnt seem like a replacement. Rather a class beteen x770 and x670

  16. #91
    Xtreme Addict
    Join Date
    Sep 2010
    Location
    Australia / Europe
    Posts
    1,310
    well the thing about this card is that it envelope should in theory be enough to put into HTPC's it should be rather easy to make low profile versions of this card. whilst that was not possible for its predecessors.

    correct me if I was wrong on that Mr Rodent "SKYMTL"

  17. #92
    Xtreme Addict
    Join Date
    Mar 2006
    Location
    Saskatchewan, Canada
    Posts
    2,207
    Quote Originally Posted by mAJORD View Post
    Less bw than 5770/6770.. Lower tdp bracket. almost doesnt seem like a replacement. Rather a class beteen x770 and x670
    Could kind of be like what happen with the 5870 --> 6870. No increase in performance, just a naming casualty/performance casualty.

    Yeah, the bandwidth seems way too low. That is lower than a 4870 which is close to 4 years old and could perform very similarly to this card if those specs are true.
    Core i7 920@ 4.66ghz(H2O)
    6gb OCZ platinum
    4870x2 + 4890 in Trifire
    2*640 WD Blacks
    750GB Seagate.

  18. #93

  19. #94
    Xtreme Member
    Join Date
    Jun 2002
    Location
    Sweden
    Posts
    182
    Prices in Sweden: Sapphire 7770 149 USD, 7750 109 USD, excluding VAT.

  20. #95

  21. #96
    Xtreme Enthusiast
    Join Date
    Jun 2010
    Posts
    588
    GIGABYTE GV-N56GOC-1GI GeForce GTX 560 $164.99 with Rebate
    MSI N560GTX-M2D1GD5 GeForce GTX 560 $159.99 with Rebate
    ASUS ENGTX560 DC/2DI/1GD5 GeForce GTX 560 $174.99 with Rebate
    EVGA 01G-P3-1460-KR GeForce GTX 560 $174.99 with Rebate

    HD 7770 Slower than GTX 560 (20%-25%)
    and the ing Problem is AMD Priced this card HD 7770 $160!!!
    So how the hell AMD Price this card In this way?
    GTX 560 cost you between $160 and $175 and this card is Faster and better than HD 7770

    Now the question is it worth buying this card HD 7770 ?
    Simply not worth....Because you have an alternative (GTX 560) at the same price and better performance.
    WOOOOOF

  22. #97
    Xtreme Member
    Join Date
    Jul 2009
    Location
    NY
    Posts
    225
    Tha Last Meal, I cannot believe you are still not banned, you are such an obvious shill, it's not even funny. Your posting history has been either positive news for nvidia or negative news for AMD, until I called you out on it, after which you started posting news, that didn't really make sense, just to blend in.

    So what ,that after rebate GTX is cheaper, for GTX 560 AMD has 6870 which it handily beats for less money !!! 6870 also beats this new card in raw performance, but this card has features, that 6870 simply does not, that's why it's currently at a premium.

    Go shill elsewhere.
    My Heatware
    Originally Posted by some guy on internet
    That's your problem right there. Just forget about how things look on paper as that's irrelevant.

  23. #98
    Xtreme Cruncher
    Join Date
    Jun 2006
    Posts
    6,215
    It's no secret that price needs adjustment . IMO Techpowerup has summed it up nicely in the conclusion of their review. Card @ stock is slightly slower than 6850,around ~4% across all resolutions, while it offers around 15% perf. improvement when OCed(matching stock 5850 on average). It has great power characteristics too,both idle and load. AMD just needs to lower the price to ~6850 level and this card will sell good.

  24. #99
    Xtreme Enthusiast
    Join Date
    Apr 2010
    Posts
    514
    Quote Originally Posted by ~CS~ View Post
    Tha Last Meal, I cannot believe you are still not banned, you are such an obvious shill, it's not even funny. Your posting history has been either positive news for nvidia or negative news for AMD, until I called you out on it, after which you started posting news, that didn't really make sense, just to blend in.
    You are not the judges of the people

    http://www.xtremesystems.org/forums/...ing&highlight=

    http://www.xtremesystems.org/forums/...zed&highlight=

    http://www.xtremesystems.org/forums/...970&highlight=

    http://www.xtremesystems.org/forums/...1st&highlight=

    http://www.xtremesystems.org/forums/...!!)&highlight=
    Last edited by cold2010; 02-15-2012 at 03:10 AM.

  25. #100
    Xtreme Enthusiast
    Join Date
    Nov 2009
    Posts
    526
    Quote Originally Posted by informal View Post
    It's no secret that price needs adjustment . IMO Techpowerup has summed it up nicely in the conclusion of their review. Card @ stock is slightly slower than 6850,around ~4% across all resolutions, while it offers around 15% perf. improvement when OCed(matching stock 5850 on average). It has great power characteristics too,both idle and load. AMD just needs to lower the price to ~6850 level and this card will sell good.
    Maybe they are limiting demand...

Page 4 of 6 FirstFirst 123456 LastLast

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
  •