Socket 754 Roundup: Comparing Generation 2
by Wesley Fink on May 28, 2004 5:57 PM EST- Posted in
- Motherboards
Abit KV8 PRO: Overclocking and Stress Testing
FSB Overclocking Results
| Front Side Bus Overclocking Testbed | |
| Default Voltage | |
| Processor: | Athlon 64 3200+ 2.0GHz |
| CPU Voltage: | 1.5V (default) |
| Cooling: | AMD Stock Athlon 64 Heatsink/Fan |
| Power Supply: | Antec TruePower 430W |
| Maximum OC: (Standard Ratio) |
248FSB x10 2480MHz (+24%) |
| Maximum FSB: (Lower Ratio) |
No Ratios Available |
This late 3200+ achieved about the same outstanding overclock on the Abit K8V PRO that we saw in our review of the K8N Neo Platinum. The Abit reached 248FSB at the stock 10X multiplier and default voltage. However, to even test the stock 10X multiplier, we had to force the Abit to 10X with the free Clock Generator utility from www.cpuid.com. This 24% overclock at stock multiplier suggests that our late 3200+ is coming from improved yields on the AMD production line for Athlon 64. If this is really the case, it is very good news for overclocking enthusiasts.
We were not able to test the highest FSB that could be achieved by lowering the multiplier, since no ratio adjustments were available on the Abit. Multipliers on Athlon 64 chips are unlocked at ratios below the shipping multiplier. This is primarily a result of the Cool'n'Quiet feature according to AMD, and AMD refers to this as Athlon 64 processors being "downclockable". FX chips are completely unlocked, and can be set at either higher or lower ratios than stock. Abit tells us that they will add ratios to a future BIOS, and they are working on a BIOS revision to correct the problem with the incorrect boot ratio. By using Clock Gen, we were able to verify that the KV8 PRO could reach 275 1:1 at an 8X multiplier with a DDR550 DIMM.
Memory Stress Test Results:
The memory stress test is very basic, as it tests the ability of the Abit KV8 PRO to operate at its officially supported memory frequency (400MHz DDR), at the best performing memory timings that our Mushkin PC3500 Level 2 or OCZ PC3500 Platinum Ltd Modules will support. Memory stress testing was conducted by running RAM at 400MHz with 2 DIMM slots filled. However, please keep in mind that Socket 754 is single-channel and will still operate at top speed with just one DIMM.| Stable DDR400 Timings - 2 DIMMs (2/2 DIMMs populated) |
|
| Clock Speed: | 200MHz |
| Timing Mode: | N/A |
| CAS Latency: | 2.0 |
| Bank Interleave: | N/A |
| RAS to CAS Delay: | 2T |
| RAS Precharge: | 10T (10T for Best Performance)* |
| Precharge Delay: | 2T |
| Command Rate: | 1T |
*Several memory tests have shown that memory performs fastest on the nVidia nForce and VIA K8T800 chipsets at a TRas (RAS Precharge) setting in the 9 to 13 range. We ran our own Memory Bandwidth tests with memtest86 with TRas settings from 5 to 15 at a wide range of different memory speeds. The best bandwidth was consistently at 9 to 11 at every speed, with TRas 10 always in the best range at every speed. The memory bandwidth improvement at TRas 10 was only 2% to 4% over TRas 5 and 6 depending on the speed, but the performance advantage was consistent across all tests. Since best performance was achieved at 2-2-2-10 timings, all Athlon 64 benchmarks were run at a TRas setting of 10.
The Abit KV8 PRO was completely stable with 2 DIMMs at the best performing settings of 2-2-2-10 at default speed. Higher overclocks could be achieved with 1 DIMM compared to 2 DIMMs, but at default speed, 1 or 2 DIMMs were reliable at the same aggressive 2-2-2-10 timings.
Abit had no adjustments at all for Bank Interleave or Command Rate, but the Command Rate is likely set to 1T. We have been discovering that Command Rate is very important to performance of Athlon 64 Dual channel memory and it should be set to 1T if possible for best performance.
Facebook
Twitter
RSS
30 Comments
View All Comments
intercollector - Saturday, May 29, 2004 - link
I'm a little surprised to see why the MSI K8N didn't get the gold compared to the Epox. Both seem almost identical in every way, except that the K8N seems to include Firewire. Shouldn't this feature make it surpass the Epox board?The only downside of the MSI board seems to be the limit of a 300 max FSB, which is probably fine for 99.99% of overclockers.
Klaasman - Saturday, May 29, 2004 - link
#7-Thanks for link, but my KV8 Pro still wont boot when selecting "fixed" in bios setup.
Why wouldn't my board have the pro chip? Manufacturing screw up maybe?
Wesley Fink - Friday, May 28, 2004 - link
#6 -Bank Interleaving is not an option in any of the BIOS' tested here. Many current BIOS enable Bank interleaving by default. Where it is an option, we definitely enable the best interleaving option available and list what we set in the memory chart. We are not ignoring this option.
bigtoe33 - Friday, May 28, 2004 - link
If you are looking for the latest Abit KV8 pro bios have a look here. http://www.bleedinedge.com/download/bios/abit%20am...multi support and PCI lock inc. if your pro board won't lock the pci bus with this bios then your board may not have the pro chipset.
Myrandex - Friday, May 28, 2004 - link
Well after finishing the article, I was wondering why none of the boards are run with bank interleaving on? Doesn't it increase memory performance for the ones that support it?Myrandex - Friday, May 28, 2004 - link
On the KV8 spec page, it states:Onboard USB 2.0/IEEE-1394 8 USB 2.0 ports supported by nF3-250
No FireWire
Should be K8T800 Pro instead of nF3-250.
Myrandex - Friday, May 28, 2004 - link
Klaasman - Friday, May 28, 2004 - link
What revision of KV8 Pro were you using and what bios version?Nobody else with a recently purchased KV8 can get the locks to work. How come your board does?
Aikouka - Friday, May 28, 2004 - link
Anandtech should have looked at the problem that the Chaintech VNF3-250 has with it's RAID and installing an OS, and the problem where the board refuses to boot from SATA if you enable RAID on any IDE HDDs. People've said they've been able to circumvent it, but I haven't got it to work yet, and Chaintech is worthless when it comes to customer service. I received an automated response about 6 or 7 days after my initial submission on their website (they have no US phone number.) And I still have not received an email from a representative yet.RyanVM - Friday, May 28, 2004 - link
If there's so little variation in system performance between these, why not look at other aspects like USB throughput/CPU utilization, IDE/SATA throughput, ethernet throughput/CPU utilization, etc.Ace's Hardware just did a great article showing that the rather crummy components being used these days on cheaper motherboards have a pretty large impact in performance in those areas.
http://www.aceshardware.com/read.jsp?id=65000298