| Products |
|
|
|
| Pricing |
|
|
|
| BIOS Frequency and Voltage settings (for overclocking) | |||
|---|---|---|---|
| ASRock X99 Extreme4 | Gigabyte X99-UD4 | MSI X99S Gaming 7 | |
| Base Clock | 90-300 MHz (0.1 MHz) | 80-333 MHz (0.01 MHz) | 91-300 MHz (0.05 MHz) |
| CPU Multiplier | 12x-120x (1x) | 12-80x (1x) | 12-80x (1x) |
| DRAM Data Rates | 800-2666 (200/266.6 MHz) | 800-2666 (200/266.6 MHz) | 1333-2666 (200/266.6 MHz) |
| CPU Vcore | 0.80-2.00V (1 mV) | 0.50-1.70V (1 mV) | 0.80-2.10V (1 mV) |
| VCCIN | 1.20-2.30V (10 mV) | 1.00-2.70V (10 mV) | 1.20-3.04V (1 mV) |
| PCH Voltage | 0.90-1.50V (25 mV) | 0.65-1.30V (5 mV) | 0.70-2.32V (10 mV) |
| DRAM Voltage | 1.00-1.80V (10 mV) | 1.00-2.00V (10 mV) | 0.60-2.80V (10 mV) |
| CAS Latency | 4-31 Cycles | 5-31 Cycles | 4-31 Cycles |
| tRCD | 5-31 Cycles | 1-31 Cycles | 4-31 Cycles |
| tRP | 5-31 Cycles | 1-31 Cycles | 4-31 Cycles |
| tRAS | 10-63 Cycles | 1-63 Cycles | 9-63 Cycles |
All three mid-priced X99-based motherboards reach the same 44 x 101 MHz base clock setting, though fractional differences in actual base clock could create the illusion of leadership. We call this a tie.
After seeing that none of the test samples were stable at the 125 MHz needed by our RAM to reach the highest DDR4 data rates, we scratched the test off our list and stuck to a 100 MHz strap. We found that Gigabyte’s X99-UD4 could potentially push a “locked” processor a little further, if Intel ever introduces one.
Not that we’d want to overclock with the X99-UD4, though. Going over the edge caused it to read a firmware error and reflash the chip with an old backup copy. If our hardware required new firmware to function (as sometimes happens after the introduction of a new CPU), we would have been stranded.
Our inability to reach DDR4-3000 at 125 MHz x 24, and Haswell-E’s inability to run 30x data rate ratios meant that we were stuck going backwards from 125 MHz to find each motherboard’s highest data rate, at 124x 24 for both the X99 Extreme4 and X99S Gaming 7. All of those re-flashes on the X99-UD4 eventually pushed us to try the board’s class-leading 106 MHz base clock in conjunction with the CPU’s highest 26.66x data rate multiple, for a grand total of 2832 MT/s.
The lack of a DDR3-2800 setting again forced us to use Haswell-E’s 26.66x maximum memory ratio to find each board’s peak bandwidth. This test was added to a previous platform after finding out that a brand not represented here today was using hobbled timings to win overclocking competitions, but Gigabyte was first to fall behind on this occasion.
|
|
|
|
|
|
|
- More, Less Or Just Different?
- ASRock X99 Extreme4
- ASRock X99 Extreme4 Software
- ASRock X99 Extreme4 Firmware
- Gigabyte X99-UD4
- Gigabyte X99-UD4 Software
- Gigabyte X99-UD4 Firmware
- MSI X99S Gaming 7
- MSI X99S Gaming 7 Software
- MSI X99S Gaming 7 Firmware
- How We Tested X99 Motherboards
- Results: 3DMark, PCMark And Sandra
- Results: 3D Gaming And Encoding
- Results: Adobe CC, Productivity And File Compression
- Results: Power, Heat And Efficiency
- Results: Overclocking
- Picking A Mid-Priced X99 Winner




Did something change?
12 Phases Six Phases Eight Phases what!
+1
Did something change?
LGA 2011 uses DDR3 while LGA2011-3 uses DDR4 so the sockets are different to prevent people from putting the wrong CPU in the wrong motherboard.
I think there were other changes but this is the big obvious one.
12 Phases Six Phases Eight Phases what!
I see what you did there Intel
It has the features you need because your keyboard and mouse are USB 2.0, not "USB3".
The Firmware-Downgrade-Bug is present since P55 boards and I saw it on EVERY generation, be it 50, 60, 70, 80 or 90 series boards. Not necesserilly on every version of the chipset, e.g. I saw it on a P65 but not on a G61 chipset. But if you have the bug then rest assured they will NEVER fix it.
The P65 was especially nasty, it downgraded the bios even when actually nothing went wrong and all BIOS settings were "default". EG I once pressed Reset and voila, back to F3. Also happens if I disconnect from power, even if the system is actually already shut down. Only way to avoid this is to power down the system safely by the OS. But then you have to switch the system on at least every couple of days or it will downgrade to F3 immediatly after power on.
Gigabyte? No thanks.
(writing this from my old P35 gigabyte system, my last gigabyte in private use)
The Firmware-Downgrade-Bug is present since P55 boards and I saw it on EVERY generation, be it 50, 60, 70, 80 or 90 series boards. Not necesserilly on every version of the chipset, e.g. I saw it on a P65 but not on a G61 chipset. But if you have the bug then rest assured they will NEVER fix it.
The P65 was especially nasty, it downgraded the bios even when actually nothing went wrong and all BIOS settings were "default". EG I once pressed Reset and voila, back to F3. Also happens if I disconnect from power, even if the system is actually already shut down. Only way to avoid this is to power down the system safely by the OS. But then you have to switch the system on at least every couple of days or it will downgrade to F3 immediatly after power on.
Gigabyte? No thanks.
(writing this from my old P35 gigabyte system, my last gigabyte in private use)
Gigabyte also has true dual-BIOS motherboards at the high-end, where the forced downgrade can be disabled. You've probably just missed that little tidbit if you think all Gigabyte boards are like this one.
This is the worst board Gigabyte has sent in a long time, and the first one that I've had to recommend against buying. It's been a few years since I've had a Gigabyte board fail after continuous forced-downgrades.
Gigabyte also has true dual-BIOS motherboards at the high-end, where the forced downgrade can be disabled. You've probably just missed that little tidbit if you think all Gigabyte boards are like this one.
This is the worst board Gigabyte has sent in a long time, and the first one that I've had to recommend against buying. It's been a few years since I've had a Gigabyte board fail after continuous forced-downgrades.
Gigabyte obviously doesn't like that we told you any of this. But to not tell you would be to lie. Its PR are nice people, but I can't lie for them, so it looks like they'll probably pack up their toys and go home.
Gigabyte PR people might be nice people, but there's only so much one can do about pumping up poor hardware design.
I've had four Gigabyte motherboards, and I've had hardware compatibility issues with them all. I've had bios issues with three, where I've needed a new bios to support a new generation of processors, but the new version doesn't work well. The latest one was my x79-UP4 board, where it didn't like my EVGA GTX 680 with the F2 bios, and the F4 bios caused the screen saver and power saver features to never turn on, and it took their support nearly two weeks to get me a program that would allow back dating the bios to the F3, which finally worked. The audio plugs never detected when speakers were plugged in, so I couldn't use the on-board audio. Finally, after 9 months of annoyances, it stopped giving power to the USB ports and the PCIe x1 slots, so I lost access to pretty much everything. It was a slow, horrible death for a bad motherboard. I'm not buying a Gigabyte product again.
They don't need PR. They need to get their heads on straight and engineer their boards better, and test better. Testing is a basic necessity for any technology design. I am a systems admin for three test labs for enterprise level hardware and software. I have seen how many bugs we find and exterminate through our testing, and I have seen how it has affected the quality of our products as management decided to reduce the complexity of our testing. Above all, I have seen how it has affected our sales. Gigabyte has never been very good at their testing. If they want to compete, they need to get off their high horse and confront the criticisms for what they are: a chance to change. If they're pulling their products from being reviewed on this site because of these criticisms, then they have shown their attitude, and nobody should be buying from them anyway.
The response from APC on the power strip teardown shows how a successful company reacts to criticism. Gigabyte's reaction will show if they're up to the task of staying in business.