• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

P45 vs x38/x48 for Crossfire

ViciousXUSMC

New Member
Joined
Jul 4, 2008
Messages
117 (0.02/day)
System Name Vicious's Rig
Processor Q6600 @ 3.6ghz @ 1.4v (Windows)
Motherboard Asus P5Q Deluxe
Cooling Thermalright Ultra 120 Extrme with S-Scyth fan
Memory 2x2gb Mushkin DDR2 800
Video Card(s) 2x Visiontek 4850 @ 675/1000
Storage 2x 640gb Western Digital
Display(s) 37" Westinghouse LVM37-W3
Case Antec P182 Special Edition
Audio Device(s) X-Fi Xtreme Music
Power Supply PC Power & Cooling 750w Silencer
Software Windows Vista x64 Ultimate
Benchmark Scores 3dmark06 18,582
I just built my new computer, and I feel I made a great build and all the right choices until just the other day I found one piece of information that really sort of upset me and gave me that ikky feeling in my stomach.

Based on this article: http://www.tweaktown.com/articles/1472/intel_p45_vs_x48_crossfire_performance/index.html

There are pretty much saying that due to the x8/x8 split for crossfire on the P45 it has a pretty big disadvantage compared to the x38/x48 chipsets with full x16/x16.

I built my entire system around the idea of that great sale for the 4850's at my local best buy only 150$ each so 300$ for the pair. I also loved the P35 chipset so figured P45 was a perfect choice and thus what I went with. (I normally do not even fathom the idea of dual card setups but this was a good exception)

I am just looking for confirmation that this article is correct from other experts because I cant see it being right. Maybe the numbers are true, but I feel its not due to the bandwidth of the pci-e slots and it has to be something else.


Sorry for the long thread but let me explain.

First of all I have done various testing in the past to determine when a slot will bottleneck a card because there is very little information that I can find on the net to determine this. I found that in my bosses computer with a P35 chipset and a 9600gt I was able to get about 3000pt in 3dmark06 with a 2x PCE-E 1.0 slot, and then increasing that to a 4x slot I got just short of 10,000 points. That shows a ton of bandwidth from just 2x to 4x. And that was 1.0 spec and its doubled for 2.0 spec.

So I cant reasonably conceive how a 8x 2.0 slot will hold back a 4850 as 8x 2.0 is 4x faster than that 10,000 mark. I know thats not a valid comparison since 3dmark06 is synthetic and also nearly half of it is cpu score, but I am leaving alot of room for error when a 4850 scores like 13,000 stock and it has room for "40,000" in a 8x slot.

The next common sense idea is that all the dual gpu core cards like the 9800GX2 or soon to be 4870X2 they are basically two cards together in one, and currently the fastest single slot is 16x. Thats basically allowing the bandwidth of that one card to be the same as splitting it into two cards each with 8x. I have not heard anything about a 16x slot getting even close to bottlenecking one of those cards.

Still tho with all that known I wont rest easy until I know for sure I made a mistake going with P45 or not. Since my goal was to finally have a high end setup that can play all my games at full 1080p, if that article is correct I really screwed up.

Here is my build: http://forum.desktopreview.com/showthread.php?t=234528

In short:
Q6600 @ 3.6ghz
4gb DDR2 @ 800mhz
2x 4850 675/1000
3dmark06 18,500

Its doing great for grid, CoD4, and even Crysis (well not so much Crysis I am getting artifacts of sorts supposedly crossfire related) but that extra 4 or 5 fps would really help in Crysis, but I dont know if I would be willing to sell and rebuild with a new mobo.
 

aCid888*

New Member
Joined
May 19, 2008
Messages
2,754 (0.48/day)
Location
In a state of flux...
The P45 is a solid chipset same as the P35, it wont make a whole lot of difference to your gaming experience/fps to only have 8x instead of 16x.

Bottom Line: don't worry :)
 
Joined
May 15, 2007
Messages
773 (0.13/day)
System Name Daedalus | O'Neill | ZPM Hive |
Processor M3 Pro (11/14) | Epyc 7402p | i5 12400F |
Motherboard Apple M3 Pro | SM H11SSL-i | TUF B660M-E D4 |
Cooling Pure Silence | Noctua NH-U12S TR4 | Noctua NH-D14 |
Memory 18GB Unified | 128GB DDR4 | 16GB DDR4 |
Video Card(s) M3 Pro | RTX 4070FE (VM) | ARC A750 LE |
Storage 512GB NVME | ALOT of SSD's | 1TB NVME |
Display(s) 14" 3024x1964 | IPMI | 1440p UW |
Case Macbook Pro 14" | NZXT H510 Flow| BC1 Test Bench |
Audio Device(s) Onboard | None | Onboard |
Power Supply ~ 77w Magsafe | EVGA 750w G3 | HX1000i |
Mouse Razer Basilisk
Keyboard Logitech G915 TKL
Software MacOS Sonoma | Proxmox 8 | Win 11 x64 |

InnocentCriminal

Resident Grammar Amender
Joined
Feb 21, 2005
Messages
6,477 (0.93/day)
System Name BeeR 6
Processor Intel Core i7 3770K*
Motherboard ASUS Maximus V Gene (1155/Z77)
Cooling Corsair H100i
Memory 16GB Samsung Green 1600MHz DDR3**
Video Card(s) 4GB MSI Gaming X RX480
Storage 256GB Samsung 840 Pro SSD
Display(s) 27" Samsung C27F591FDU
Case Fractal Design Arc Mini
Power Supply Corsair HX750W
Software 64bit Microsoft Windows 10 Pro
Benchmark Scores *@ 4.6GHz **@ 2133MHz
I guess there are bored people, who just posts to get attention lol or to hassle other people :shadedshu

They don't even read the original post most of the time. ViciousXUSMC, you may find this review of the ASUS P5Q interesting, they compare it against these boards...

# Gigabyte GA-X48T-DQ6 (Intel X48, DDR3, F4 BIOS)
# Abit IX38 QuadGT (Intel X38, DDR2, "12" BIOS)
# Gigabyte GA-EP35-DS4 (Intel P35, DDR2, F3 BIOS)
# MSI P35 Diamond (Intel P35, DDR3, 2.2 BIOS)

What P45 board did you buy exactly? Hope it helps! Welcome to the forums btw. ;)
 
Last edited:
Joined
May 15, 2007
Messages
773 (0.13/day)
System Name Daedalus | O'Neill | ZPM Hive |
Processor M3 Pro (11/14) | Epyc 7402p | i5 12400F |
Motherboard Apple M3 Pro | SM H11SSL-i | TUF B660M-E D4 |
Cooling Pure Silence | Noctua NH-U12S TR4 | Noctua NH-D14 |
Memory 18GB Unified | 128GB DDR4 | 16GB DDR4 |
Video Card(s) M3 Pro | RTX 4070FE (VM) | ARC A750 LE |
Storage 512GB NVME | ALOT of SSD's | 1TB NVME |
Display(s) 14" 3024x1964 | IPMI | 1440p UW |
Case Macbook Pro 14" | NZXT H510 Flow| BC1 Test Bench |
Audio Device(s) Onboard | None | Onboard |
Power Supply ~ 77w Magsafe | EVGA 750w G3 | HX1000i |
Mouse Razer Basilisk
Keyboard Logitech G915 TKL
Software MacOS Sonoma | Proxmox 8 | Win 11 x64 |
I guess there are bored people, who just posts to get attention lol or to hassle other people :shadedshu

another test provided by google:

http://www.tomshardware.com/de/Graf...16-x8-x4-Crossfire,testberichte-239907-4.html

http://media.bestofmicro.com/D/P/71773/original/Crossfire-Einzelergebnisse-x8-vs-x16.jpg

I was asknig you a question and instead of responding in a civilised manner you accuse me of posting to get attention or trying to give you hassle?! What are you harping on about?

Why didn't you also link to these additional tests in your first post?!

EDIT: God that TH review is hard to read with just tables of numbers.

As for the P45 CF issue a quick look on bit-tech shows that in CF (dual HD3870) the P45 boards are performing identically to the X38/X48 boards.

http://www.bit-tech.net/hardware/2008/06/20/msi-p45-platinum/12

(This reviews lists all of the P45 boards in the results).

It clearly shows that there is no perfomance drop from 16x16x to 8x8x.

This calls into question the validity of the tweaktown review (given that they also only tested one motherboard as well it doesn't help matters).
 
Last edited:
Joined
Nov 8, 2006
Messages
5,052 (0.80/day)
Location
Manchester, United Kingdom
Processor AMD FX 8320 @ 4GHz
Motherboard Gigabyte GA-990FXA-UD5 rev1
Cooling Corsair H70
Memory 4 x 4GB DDR3 Ripjawz 1600Mhz
Video Card(s) Sapphire Vapor-X AMD R9 280X
Storage 1 x 500GB Samsung Evo 850, 1 x 500GB Vrap Data Drive, 3 x 2TB Seagate, 1 x 1TB Samsung F1
Display(s) 3 x DGM IPS-2402WDH
Case Coolermaster HAF X
Audio Device(s) Onboard
Power Supply Coolermaster 1000W Silent Pro M
Mouse Logitech G502
Keyboard Logitech G510
Software Windows 10 Pro x64
Ok, firstly, MoA, I would like to point out that that review is borked big-style. Either I'm reading it wrong, or Call of Duty 4 is getting 700+ FPS, I've yet to see a card in the entire world that can do 700 FPS on CoD4 even at minimum settings.

And also, notice how the others are all within a few fps of each other? I think the biggest is about 9fps, honestly, that's not much when we are talking 80+ fps.

Second, that bit-tech review threw up something interesting, the P5Q performing better in CF then the rest? Hmm...
 

ViciousXUSMC

New Member
Joined
Jul 4, 2008
Messages
117 (0.02/day)
System Name Vicious's Rig
Processor Q6600 @ 3.6ghz @ 1.4v (Windows)
Motherboard Asus P5Q Deluxe
Cooling Thermalright Ultra 120 Extrme with S-Scyth fan
Memory 2x2gb Mushkin DDR2 800
Video Card(s) 2x Visiontek 4850 @ 675/1000
Storage 2x 640gb Western Digital
Display(s) 37" Westinghouse LVM37-W3
Case Antec P182 Special Edition
Audio Device(s) X-Fi Xtreme Music
Power Supply PC Power & Cooling 750w Silencer
Software Windows Vista x64 Ultimate
Benchmark Scores 3dmark06 18,582
Well I am very skeptical of that tweak town review, like I said it sort of defies what I have seen in my own testing, and even in there test results there is data that doesn't make too much sense. Heck I didnt see it myself but somebody posted on another forum about that article that originally they put nvidia drivers under the "test setup" section rather than the catalyst drivers....

I will go and read the links you guys have posted now and see if information from alternat sources seems to be consistently showing the same.

As for my board? Its the Asus P5Q Deluxe, my entire computer build is in the link I posted above.

edit: first link read the toms hardware one. It scared me at first confirming what tweaktown had but if you pay attention you will notice that the test was 8x/8x @ 1.0 spec vs 16x/16x 2.0 spec. In my case with a P45 being 2.0 spec it would have double the bandwidth of what they had in that test.

edit2: second link is the same 8x/8x 1.0 spec vs 16x/16x 2.0 spec I can definitly see a 4850 being slowed down in a 8x 1.0 slot... so I agree with those links. but with 2.0 spec and 8x lane is just as fast as an older 16x lane so I really cant image it being an issue. So the quest continues to verify these tweak town numbers.
 

InnocentCriminal

Resident Grammar Amender
Joined
Feb 21, 2005
Messages
6,477 (0.93/day)
System Name BeeR 6
Processor Intel Core i7 3770K*
Motherboard ASUS Maximus V Gene (1155/Z77)
Cooling Corsair H100i
Memory 16GB Samsung Green 1600MHz DDR3**
Video Card(s) 4GB MSI Gaming X RX480
Storage 256GB Samsung 840 Pro SSD
Display(s) 27" Samsung C27F591FDU
Case Fractal Design Arc Mini
Power Supply Corsair HX750W
Software 64bit Microsoft Windows 10 Pro
Benchmark Scores *@ 4.6GHz **@ 2133MHz
^^

Hit the nail on the head with my link then.

:p

You should fill out your system specification if your User Control Panel.
 

ViciousXUSMC

New Member
Joined
Jul 4, 2008
Messages
117 (0.02/day)
System Name Vicious's Rig
Processor Q6600 @ 3.6ghz @ 1.4v (Windows)
Motherboard Asus P5Q Deluxe
Cooling Thermalright Ultra 120 Extrme with S-Scyth fan
Memory 2x2gb Mushkin DDR2 800
Video Card(s) 2x Visiontek 4850 @ 675/1000
Storage 2x 640gb Western Digital
Display(s) 37" Westinghouse LVM37-W3
Case Antec P182 Special Edition
Audio Device(s) X-Fi Xtreme Music
Power Supply PC Power & Cooling 750w Silencer
Software Windows Vista x64 Ultimate
Benchmark Scores 3dmark06 18,582

InnocentCriminal

Resident Grammar Amender
Joined
Feb 21, 2005
Messages
6,477 (0.93/day)
System Name BeeR 6
Processor Intel Core i7 3770K*
Motherboard ASUS Maximus V Gene (1155/Z77)
Cooling Corsair H100i
Memory 16GB Samsung Green 1600MHz DDR3**
Video Card(s) 4GB MSI Gaming X RX480
Storage 256GB Samsung 840 Pro SSD
Display(s) 27" Samsung C27F591FDU
Case Fractal Design Arc Mini
Power Supply Corsair HX750W
Software 64bit Microsoft Windows 10 Pro
Benchmark Scores *@ 4.6GHz **@ 2133MHz
Good man!

XD
 

mk_ln

New Member
Joined
Mar 7, 2006
Messages
287 (0.04/day)
Processor Intel Core i7-970 @ 4.02
Motherboard Gigabyte GA-X58A-UD5
Cooling Cooler Master Hyper 212+ (dual fans)
Memory 12GB Corsair Vengeance 1600MHz
Video Card(s) MSI GTX 465 TWIN FROZR II GE flashed to GTX 470 + PNY GTX470 (SLI)
Storage 1x Intel 80GB G2, 1x 320 series 160GB SSD, + 11 other drives totaling a bit over 12TB
Display(s) Samsung 2693 + ASUS VW266H
Case Silverstone TJ07-BW
Audio Device(s) Logitech Z5500
Power Supply Corsair HX850 - Modular
Software XP PRO x86, Vista Ult. x64, Win7 Ult. x64
the only issue i have with the bit-tech review is that it was using older cards not capable of saturating the bandwidth. supposedly these new cards ARE capable of doing that, but like ViciousXUSMC, i am skeptical about tweaktown's pcie performance review
 

MoA

New Member
Joined
Jun 24, 2008
Messages
27 (0.00/day)
yeah very strange bit-tech and tweaktown..
maybe both use different drivers, since there seems to be a big difference between:8.6 normal, 8.6 hotfix, 8.6 beta, 8.7 beta...

but tweaktown is using 4850 as tests which may request much more bandwith from the pciE especially with those 800SP, this is may be where the x16 and x8 shows the differences (maybe...)
 
Last edited:

InnocentCriminal

Resident Grammar Amender
Joined
Feb 21, 2005
Messages
6,477 (0.93/day)
System Name BeeR 6
Processor Intel Core i7 3770K*
Motherboard ASUS Maximus V Gene (1155/Z77)
Cooling Corsair H100i
Memory 16GB Samsung Green 1600MHz DDR3**
Video Card(s) 4GB MSI Gaming X RX480
Storage 256GB Samsung 840 Pro SSD
Display(s) 27" Samsung C27F591FDU
Case Fractal Design Arc Mini
Power Supply Corsair HX750W
Software 64bit Microsoft Windows 10 Pro
Benchmark Scores *@ 4.6GHz **@ 2133MHz
The Bit-Tech review is relatively old compared to the TweakTown.
 
Joined
Feb 20, 2008
Messages
4,628 (0.79/day)
Location
where everyone wants to be
System Name Everchanging
Processor AMD Ryzen 9 5900X
Motherboard ASUS ROG Crosshair Dark Hero
Cooling Arctic Cooling Liquid Freezer 2 420mm
Memory 2x16GB Corsair DDR4 3600 CL16
Video Card(s) eVGA RTX 3090 Ti FTW3
Storage Samsung 980 Pro 256GB, Samsung 980 Pro 2TB
Display(s) 2xSamsung 28" 4k HDR 144Hz
Case Fractal Meshify 2 XL
Audio Device(s) fiio K9 to Hifiman Sundara's via 4.4mm balanced cable
Power Supply Seasonic Prime TX 850w
Mouse Corsair Harpoon Wireless RGB
Keyboard Corsair K70 RGB PRO
Software Windows 11 x64
From fitseries3 experience, running a 3870x2 on a PCI-E 1.x 4x lane didn't seem to affect performance too much, so where as yes with crossfire in dual PCI-E 2.0 16x compared to the 8x, you do technically have double the speed, I'm hesitant to believe that it makes too much of a difference, maybe with the newer 4870 and coming 4870x2, but for the 4850 and lower, probably not as much.
 

InnocentCriminal

Resident Grammar Amender
Joined
Feb 21, 2005
Messages
6,477 (0.93/day)
System Name BeeR 6
Processor Intel Core i7 3770K*
Motherboard ASUS Maximus V Gene (1155/Z77)
Cooling Corsair H100i
Memory 16GB Samsung Green 1600MHz DDR3**
Video Card(s) 4GB MSI Gaming X RX480
Storage 256GB Samsung 840 Pro SSD
Display(s) 27" Samsung C27F591FDU
Case Fractal Design Arc Mini
Power Supply Corsair HX750W
Software 64bit Microsoft Windows 10 Pro
Benchmark Scores *@ 4.6GHz **@ 2133MHz
Thinking of it, I doubt many cards actually utilise the PCIe bus to it's fullest. It probably just gives the cards more headroom if they ever needed. I don't even think the AGP bus is fully utilised now, even with a HD3850.
 

HTC

Joined
Apr 1, 2008
Messages
4,601 (0.79/day)
Location
Portugal
System Name HTC's System
Processor Ryzen 5 2600X
Motherboard Asrock Taichi X370
Cooling NH-C14, with the AM4 mounting kit
Memory G.Skill Kit 16GB DDR4 F4 - 3200 C16D - 16 GTZB
Video Card(s) Sapphire Nitro+ Radeon RX 480 OC 4 GB
Storage 1 Samsung NVMe 960 EVO 250 GB + 1 3.5" Seagate IronWolf Pro 6TB 7200RPM 256MB SATA III
Display(s) LG 27UD58
Case Fractal Design Define R6 USB-C
Audio Device(s) Onboard
Power Supply Corsair TX 850M 80+ Gold
Mouse Razer Deathadder Elite
Software Ubuntu 19.04 LTS
Someone please correct me if i'm wrong but i think that the x8 VS x16 only comes into play with high resolutions and very high details: i'm talking 1920 / 1200 and bigger resolutions.

The reason this is so, is because the x8 bandwidth isn't fully utilized @ lower resolutions but it starts to reach it's peak @ 1920 / 1200 with high details, and it becomes bottlenecked.


I remember reading about this, a while back (don't recall from where), and this is what i understood of it.
 
Last edited:
Joined
Aug 29, 2004
Messages
967 (0.14/day)
Location
Danville IL
Processor I7 4770k
Motherboard ASUS z87 pro
Cooling Corsair a50
Memory 4x4 gig Gskil aries ddr3 1866
Video Card(s) Gigabyte r280x windforce
Storage intel 520 120 gig ssd
Display(s) 24 inch Asus IPS
Case Cool Master
Audio Device(s) realtek onboard
Power Supply hiper 880
Software win 7 ult
Ive done extensive testing on 2-2900xts and x38-975x-p35 chipsets in crossfire- 2-4 fps in crysis difference in all 3 chipsets The single card solution is the best, crossfire at best gives 30-35% increase on second card but the card isnt 65-70% off price. Dual gpu seems to do the best for now. ALLEN
 

ViciousXUSMC

New Member
Joined
Jul 4, 2008
Messages
117 (0.02/day)
System Name Vicious's Rig
Processor Q6600 @ 3.6ghz @ 1.4v (Windows)
Motherboard Asus P5Q Deluxe
Cooling Thermalright Ultra 120 Extrme with S-Scyth fan
Memory 2x2gb Mushkin DDR2 800
Video Card(s) 2x Visiontek 4850 @ 675/1000
Storage 2x 640gb Western Digital
Display(s) 37" Westinghouse LVM37-W3
Case Antec P182 Special Edition
Audio Device(s) X-Fi Xtreme Music
Power Supply PC Power & Cooling 750w Silencer
Software Windows Vista x64 Ultimate
Benchmark Scores 3dmark06 18,582
I agree allen337 for the most part, but I only decided to get 2 of them because of the great deal I got. 25% off all visontek cards at bestbuy means they costed 150$ each so 300$ for the pair and I knew that 2 4850's would beat a single 4870.
 

Tatty_Two

Gone Fishing
Joined
Jan 18, 2006
Messages
25,793 (3.88/day)
Location
Worcestershire, UK
Processor Rocket Lake Core i5 11600K @ 5 Ghz with PL tweaks
Motherboard MSI MAG Z490 TOMAHAWK
Cooling Thermalright Peerless Assassin 120SE + 4 Phanteks 140mm case fans
Memory 32GB (4 x 8GB SR) Patriot Viper Steel 4133Mhz DDR4 @ 3600Mhz CL14@1.45v Gear 1
Video Card(s) Asus Dual RTX 4070 OC
Storage WD Blue SN550 1TB M.2 NVME//Crucial MX500 500GB SSD (OS)
Display(s) AOC Q2781PQ 27 inch Ultra Slim 2560 x 1440 IPS
Case Phanteks Enthoo Pro M Windowed - Gunmetal
Audio Device(s) Onboard Realtek ALC1200/SPDIF to Sony AVR @ 5.1
Power Supply Seasonic CORE GM650w Gold Semi modular
Mouse Coolermaster Storm Octane wired
Keyboard Element Gaming Carbon Mk2 Tournament Mech
Software Win 10 Home x64
For all I have read (and thats a lot) on tests using modern cards (3850/3870 onwards) 8 x 8 across the board is only a minor 5% decrease in performance against 16 x 16 and very much better than say the P35 at 16 x 4 which seems to sit around at 15% hit, sometimes more. Just set the PCI-E bandwidth from 100 to 110mhz and that makes up for the 5% straight away! :rockout:
 

Ketxxx

Heedless Psychic
Joined
Mar 4, 2006
Messages
11,521 (1.75/day)
Location
Kingdom of gods
System Name Ravens Talon
Processor AMD R7 3700X @ 4.4GHz 1.3v
Motherboard MSI X570 Tomahawk
Cooling Modded 240mm Coolermaster Liquidmaster
Memory 2x16GB Klevv BoltX 3600MHz & custom timings
Video Card(s) Powercolor 6800XT Red Devil
Storage 250GB Asgard SSD, 1TB Integral SSD, 2TB Seagate Barracuda
Display(s) 27" BenQ Mobiuz
Case NZXT Phantom 530
Audio Device(s) Asus Xonar DX 7.1 PCI-E
Power Supply 1000w Supernova
Software Windows 10 x64
Benchmark Scores Fast. I don't need epeen.
The article is right and wrong. Some stay steadfast in the belife that 16x CF on PCIE is better than 8x CF on PCIE 2.0. The reality is PCIE 2.0 offers advantages over PCIE. PCIE 2.0 in 8x CF offers the same bandwidth as PCIE @ 16x CF, while PCIE 2.0 also offers more power via the slot (125w I belive over 75w on PCIE) as well as having various subroutine and packet optimisations PCIE does not have. So, in summary PCIE 2.0 has over PCIE;

PCIE 2.0 has the same bandwidth @ 8x as PCIE has @ 16x
PCIE 2.0 offers more power via the slot than PCIE does
PCIE 2.0 has various subroutine and packet optimisations PCIE does NOT have.

So, is PCIE 2.0 beter than PCIE even when on a P45 PCIE 2.0 runs @ 8x in CF and on an X38\48 the PCIE runs @ 16x in CF? Yes. Due to the above mentioned, 8x CF on PCIE 2.0 at least makes it equal to PCIE x16 in terms of bandwidth, and in other areas PCIE 2.0 is noticably superior on a subroutine level.
 

Jansku07

New Member
Joined
May 24, 2008
Messages
171 (0.03/day)
Location
Finland
Processor 4000X2 @ 2,1Ghz w/ AC Freezer
Motherboard MSI K9AGM3-FD
Memory 2GB DDR2
Video Card(s) 3650 w/ 256MB GDDR3
Storage WD 320GB - 7200RPM
Display(s) DELL m991 /1024x768/ 85hz
Case resonating metal case
Power Supply FSP 350W -> 16A@12V
Software Vista Home Premium x86 SP2 FIN
Benchmark Scores oh noeees.
Look @ this. X48 vs x38 vs p45 vs p35. Testsetup: Hd4850CF + Q6600 @ 3,0 gHz. http://www.legionhardware.com/document.php?id=761&p=0
part of the conclusion said:
... For the most part the X48 is just 5% faster than the cheaper P45 chipset, while this chipset held roughly the same margin over the older X38 chipset. Therefore as we see it, X38 owners really would need to purchase an X48 motherboard to see any real performance gains, while the P45 is a massive step up from the P35 chipset when talking Crossfire performance.
 

Ketxxx

Heedless Psychic
Joined
Mar 4, 2006
Messages
11,521 (1.75/day)
Location
Kingdom of gods
System Name Ravens Talon
Processor AMD R7 3700X @ 4.4GHz 1.3v
Motherboard MSI X570 Tomahawk
Cooling Modded 240mm Coolermaster Liquidmaster
Memory 2x16GB Klevv BoltX 3600MHz & custom timings
Video Card(s) Powercolor 6800XT Red Devil
Storage 250GB Asgard SSD, 1TB Integral SSD, 2TB Seagate Barracuda
Display(s) 27" BenQ Mobiuz
Case NZXT Phantom 530
Audio Device(s) Asus Xonar DX 7.1 PCI-E
Power Supply 1000w Supernova
Software Windows 10 x64
Benchmark Scores Fast. I don't need epeen.
Thats talking about the P45. I was specifically comparing PCIE vs. PCIE 2.0, clearly 2.0 is superior. Any performance deficite the P45 currently has can easily be put down to immature chipset drivers and the BIOS needing slight refinement. Hell, theres only been one driver Intel has churned out for the P45 currently its that new and you get an unknown error while trying to install the P45 chipset drivers, you have to run it again to install everything, and who knows if that unknown error is overcome the 2nd time around or if that particular part of the driver is still bugged.
 
Top