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

TPU GPU review irregularities (question)

Joined
Sep 1, 2006
Messages
88 (0.01/day)
Processor AMD Phenom II X4 940 @ 3.6 (16x225)
Motherboard DFI LP DK 790FXB-M2RSH
Cooling CoolIT Domino A.L.C.
Memory OCZ Reaper HPC 8GB (4x2GB) DDR2 1066 (PC2 8500)
Video Card(s) XFX HD-487A-ZDDC Radeon HD 4870 (OC'd to 825/1000) | XFX HD-489A-ZDFC Radeon HD 4890
Storage 2x Seagate 320GB SATA2 (ST3320620AS)
Case LIAN LI PC-X500B
Audio Device(s) Creative X-Fi XtremeGamer 7.1
Power Supply PCP&C S75QB 750W | Thermaltake Pure Power Express 250W
Software Windows Vista Ultimate x64
like many i come around to TPU for video card reviews more than anything else. W1zzard you are a walking god in my GPU modding book but i need an explanation to the following. below is your recent Sapphire HD 4850 review numbers vs. your Zotac 8800 GTS 512 numbers in S.T.A.L.K.E.R. SHoC @ 16xx by 1xxx w/ 4xAA and 16xAF:





in comparison of these two reviews some numbers don't match up and i am curious as to why:

8500 GT 256MB
Sapphire HD 4850 review FPS: 8.2
Zotac 8800 GTS 512 review FPS: 22.8

--

7900 GTX 512MB
Sapphire HD 4850 review FPS: 26.7
Zotac 8800 GTS 512 review FPS: 78.7

--

HD 2900 XT 1024MB
Sapphire HD 4850 review FPS: 48.2
Zotac 8800 GTS 512 review FPS: 109.1

--

HD 3870 512MB
Sapphire HD 4850 review FPS: 48.6
Zotac 8800 GTS 512 review FPS: 110.2

--

8800 GT 512MB
Sapphire HD 4850 review FPS: 50.3
Zotac 8800 GTS 512 review FPS: 134.6

is this all due to the different screen resolution heights? i ask as i am gearing up to leave the wonderful world of 1280x1024 for 16xx+.

- Robert (pMr)dIj
 
Last edited:

ShadowFold

New Member
Joined
Dec 23, 2007
Messages
16,918 (2.83/day)
Location
Omaha, NE
System Name The ShadowFold Draconis (Ordering soon)
Processor AMD Phenom II X6 1055T 2.8ghz
Motherboard ASUS M4A87TD EVO AM3 AMD 870
Cooling Stock
Memory Kingston ValueRAM 4GB DDR3-1333
Video Card(s) XFX ATi Radeon HD 5850 1gb
Storage Western Digital 640gb
Display(s) Acer 21.5" 5ms Full HD 1920x1080P
Case Antec Nine-Hundred
Audio Device(s) Onboard + Creative "Fatal1ty" Headset
Power Supply Antec Earthwatts 650w
Software Windows 7 Home Premium 64bit
Benchmark Scores -❶-❸-❸-❼-
8800GTS uses a E6550 at 2.3ghz and the 4850 uses a 8400 at 3.6ghz
 
Joined
Sep 1, 2006
Messages
88 (0.01/day)
Processor AMD Phenom II X4 940 @ 3.6 (16x225)
Motherboard DFI LP DK 790FXB-M2RSH
Cooling CoolIT Domino A.L.C.
Memory OCZ Reaper HPC 8GB (4x2GB) DDR2 1066 (PC2 8500)
Video Card(s) XFX HD-487A-ZDDC Radeon HD 4870 (OC'd to 825/1000) | XFX HD-489A-ZDFC Radeon HD 4890
Storage 2x Seagate 320GB SATA2 (ST3320620AS)
Case LIAN LI PC-X500B
Audio Device(s) Creative X-Fi XtremeGamer 7.1
Power Supply PCP&C S75QB 750W | Thermaltake Pure Power Express 250W
Software Windows Vista Ultimate x64
8800GTS uses a E6550 at 2.3ghz and the 4850 uses a 8400 at 3.6ghz
yes you are correct:

Zotac 8800 GTS 512MB review uses:
Intel Core 2 Duo E6550 2.33 GHz
(Conroe, 2x 2048 KB Cache)

Sapphire HD 4850 512MB review uses:
Intel Core 2 Duo E8400 @ 3.6 GHz
(Wolfdale, 6144 KB Cache)

assuming you were pointing to the CPU as the cause of this --> if the CPU involved was a bottleneck in the test system we would see lower FPS on the E6550 system with the same cards rather than the reverse as above.

- Robert (PmR)DeathInJune
 

JC316

Knows what makes you tick
Joined
Jan 24, 2006
Messages
9,397 (1.41/day)
System Name Budget Gaming
Processor AMD FX6300
Motherboard Gigabyte 880GMA-USB3
Cooling Coolermaster Hyper 212+
Memory 8GB Ripjaws DDR3 1600
Video Card(s) HD7850 1GB
Storage 1TB Sata2
Display(s) Acer 24" LED
Case Generic black
Audio Device(s) Stock onboard
Power Supply FSP Aurum Gold 650W
Software Windows 7 Home Premium 64bit
In that case then, the 4850 should be even faster than the GTS. It's probably a driver issue or something like that. Stalker is a buggy game anyway.
 

W1zzard

Administrator
Staff member
Joined
May 14, 2004
Messages
27,049 (3.71/day)
Processor Ryzen 7 5700X
Memory 48 GB
Video Card(s) RTX 4080
Storage 2x HDD RAID 1, 3x M.2 NVMe
Display(s) 30" 2560x1600 + 19" 1280x1024
Software Windows 10 64-bit
we changed the stalker rendering settings at some point. also the cpu was upgraded, bios upgraded, xp fixes
 
Joined
Sep 1, 2006
Messages
88 (0.01/day)
Processor AMD Phenom II X4 940 @ 3.6 (16x225)
Motherboard DFI LP DK 790FXB-M2RSH
Cooling CoolIT Domino A.L.C.
Memory OCZ Reaper HPC 8GB (4x2GB) DDR2 1066 (PC2 8500)
Video Card(s) XFX HD-487A-ZDDC Radeon HD 4870 (OC'd to 825/1000) | XFX HD-489A-ZDFC Radeon HD 4890
Storage 2x Seagate 320GB SATA2 (ST3320620AS)
Case LIAN LI PC-X500B
Audio Device(s) Creative X-Fi XtremeGamer 7.1
Power Supply PCP&C S75QB 750W | Thermaltake Pure Power Express 250W
Software Windows Vista Ultimate x64
In that case then, the 4850 should be even faster than the GTS. It's probably a driver issue or something like that. Stalker is a buggy game anyway.
well the point isn't the results of the 4850 but rather the cards that appear in both reviews running drastically slower in the review of Sapphire's HD 4850 512MB as compared to the previous review of the Zotac 8800 GTS 512MB. just trying to figure out why this occurred.

we changed the stalker rendering settings at some point.
ok i can see this being the cause so that answers that. just had to be sure that there wasn't something between 1050px and 1200px that produced such a drop before i went out and got my new primary display. thanks!

- Robert (PmR)DIJ
 
Last edited:
Top