tRAS and Memory Stress Testing

Memory tRAS Recommendations

In past reviews, memory bandwidth tests established that a tRAS setting of 11 or 12 was generally best for nForce2, a tRAS of 10 was optimal for the nForce3 chipset, and a tRAS of 7 was optimal for the nForce4 chipset. The recently tested ULi 1695 was best at tRAS of 10.

Since this is our first review of a new ATI chipset, tRAS timings were first tested with memtest86, a free diagnostic program with its own boot OS that will boot from either a floppy disk or optical disk. Bandwidth of OCZ PC3200 Platinum Rev. 2, based on Samsung TCCD chips, was measured from tRAS 5 to tRAS 12 to determine the best setting.
Memtest86 Bandwidth
Sapphire PURE Innovation with Athlon 64 4000+
5 tRAS 2048
6 tRAS 2092
7 tRAS 2141
8 tRAS 2141
9 tRAS 2092
10 tRAS 2048
11 tRAS 2004
12 tRAS 1922

Two things stood out in our tRAS tests. First, the Sapphire ATI exhibits the best bandwidth at a tRAS setting of 7 or 8. We decided to use 7, since it is the same as the best setting for nForce4. Second, the bandwidth numbers on the ATI were a bit higher across the tRAS tests than what we saw with the ULi 1695. This suggests ATI having better optimized memory management - at least at this stage of development.

Memory Stress Test

Our memory stress test measures the ability of the Sapphire ATI to operate at its officially supported memory frequency (400MHz DDR), at the lowest memory timings that OCZ PC3200 Platinum Rev. 2 modules will support. All DIMMs used for stress testing were 512MB double-sided (or double-bank) memory. To make sure that memory performed properly in Dual-Channel mode, memory was only tested using either one dual-channel (2 DIMMs) or 2 dual-channels (4 DIMMs).

Stable DDR400 Timings - One Dual-Channel
(2/4 DIMMs populated)
Clock Speed: 200MHz
CAS Latency: 2.0
RAS to CAS Delay: 2T
RAS Precharge: 7T
Precharge Delay: 2T
Command Rate: 1T

Using two DIMMs in Dual-Channel 128-bit mode, the memory performed in all benchmarks at the fastest 2-2-2-7 timings at default voltage, which was the only memory voltage available.

Stable DDR400 Timings - 4 DIMMs
(4/4 DIMMs populated)
Clock Speed: 200MHz
CAS Latency: 2.0
RAS to CAS Delay: 2T
RAS Precharge: 7T
Precharge Delay: 2T
Command Rate: 2T

Tests with all four DIMM slots populated on the Sapphire required a 2T Command Rate with 4 DIMMs in two dual channels. This is the pattern seen on other top-performing Socket 939 boards. The Sapphire AMD had no problem running at a 200 CPU speed setting with 4 double-sided DIMMs. Since the Athlon 64 memory controller is on the processor, there were no real surprises in the memory stress tests. The Sapphire ATI is certainly competitive with the best Socket 939 boards in memory performance.

Overclocking: Sapphire PURE Innovation Test Setup
Comments Locked

52 Comments

View All Comments

  • QueBert - Saturday, July 30, 2005 - link

    I like the white, but not a big fan of red. As for it being the "first white motherboard we've seen" There was a really sweet looking one by I believe, Epox. Platinum colored PCB with blue and gold on the board. The color scheme of this Sapphire is different, and different is good. I hate green/red PCB's.
  • beorntheold - Saturday, July 30, 2005 - link

    Under Gaming Performance:
    "... If you keep in mind that the orange bar represents the same NVIDIA 6800 Ultra used to test the other boards in this review, you can clearly see that the Sapphire ATI is at or near the top in most game tests..."
    There is either an error in the graph or in the text - because the orange bar clearly says
    nV 7800 GTX.
  • Olaf van der Spek - Saturday, July 30, 2005 - link

    > This number is meaningless as far as hard disk performance is concerned as it is just the number of IO operations completed in a second.

    What exactly is meant here?
    Isn't more completed operations per second better?
  • Wesley Fink - Saturday, July 30, 2005 - link

    I meant that the operations per second is meaningless as a SPECIFICATION of hard drive performance. Yes, more operations per second is better, but you will never see iPeak ops/sec quoted as a specification.
  • roel - Saturday, July 30, 2005 - link

    And what about support for linux?
    Will it boot? Will it be fast as well?
    I'd like to know.

    roel
  • kevykev - Friday, July 29, 2005 - link

    What is with the stupid fish names though? How ridiculous.
  • TheInvincibleMustard - Saturday, July 30, 2005 - link

    Actually, that was one of the first things to catch my eye ... Not the fish, but the jaguar, with the silkscreened logo looking almost exactly the same as the automobile company (Google Images if you're curious). The actual leaping cat is positioned slightly differently between the two, but the similarity is remarkable.

    If I had better image manipulation on this machine I'm at, I'd whip up a side-by-side comparison to better illustrate (hehe) my point.
  • shoRunner - Friday, July 29, 2005 - link

    All i can say is drool, this definately looks like a very promising board, if they can fix the issue with the USB transfer rates(even without a fix its will still be very competative, how often do most ppl transfer huge amounts of data over USB?). What i still want to see is one of these boards with the integrate graphics and a DVI port.
  • Stas - Friday, July 29, 2005 - link

    Just as I expected: nVidia is whooped. ATi + AMD = Performance
  • Zebo - Friday, July 29, 2005 - link

    Looks great Wes..cept for memory is to close together and not staggard and may present cooling problems between the sticks...plus the board looks like green snot, much prefer blacks, reds and blues.

Log in

Don't have an account? Sign up now